Download - Loudspeaker Builder

Transcript
Page 1: Loudspeaker Builder

LoudspeakerBuilder.ca - (Thiele-Small Parameters)

But still the most common cause of speaker failure is simple abuse; cranking it up beyond its power rating while asking the speaker to produce frequencies lower than it's frequency rating. So be sure to take into account the suggested usable frequency range and the Xmech parameter in conjunction with the power rating of the speaker to avoid such failures.

The Thiele-Small Q's - The control of those peaks at resonant frequency (Fs) is done with the speaker's suspension (spider, surround) balanced off against the opposing force of the voice coil and magnet. The measurements used in describing the control (dampening) the movement of the speaker's suspension are the Qms, Qes and Qts. If the manufacturer does this right they can often put the resonate frequency of the speaker outside; either above or below its intended frequency range use which helps results in a more "flat" frequency response speaker.

Q - The relative damping or system losses of a loudspeaker in an enclosure. The ratio of stored to dissipated energy

Qms - Is the measurement of control from the mechanical suspension system at resonance (Fs) which include the spider and the surround. They allow and control the movement of the speaker cone.

Qes - Is the measurement of control from the electrical suspension system at resonance (Fs) which include the voice coil and the magnet; the "engine" of the speaker.

Qts - The opposing forces from the mechanical and electrical suspensions acting against each other is the total Q of a speaker in free air at resonance (Fs).

Qmc - The Q of a speaker in a sealed box considering only the mechanical resistance.

Qec - The Q of a speaker in a sealed box considering only electrical resistance.

Qtc - Put a speaker into an enclosure and you then change how that speaker will act due to the resistance of the air pressure inside of the enclosure. When the speaker cone moves in or out the air pressure within the enclosure will put a resistance on its movements. The size and type of enclosure you build will depends upon the Qtc value you desire.

To decide upon your loudspeakers enclosure size you will need some loudspeaker software or good math skills and the math formulas; I prefer the software approach. In the software programs you will be asked to enter the required Thiele-Small parameters about the drivers such as the Q's, Fs, Vas, etc, some also ask enclosure type, number of drivers, etc. Then for the program to calculate the enclosure size it will want to know the Qtc value you want. The Qtc value you choose is a personal preference. A value of 0.707 is what most designers aim for, it will give you the flattest frequency response (accurate sound reproduction) and the lowest possible F3 (widest usable frequency range). Some people may not like this sound and want to enhanced base response so they may aim for 0.8 or higher.

Page 2: Loudspeaker Builder

In general high quality accurate loudspeakers Qtc are around 0.707, while loudspeakers that are designed to enhance the base may range from 0.8 to a max of 1.1. The more you move away from 0.707 anything over that will slowly start to sound boomy and unnatural and the base response will become more restricted.

If you want loud clean base go with larger drivers in larger enclosures, don't overwork a smaller speaker and try to increase t

A.N. Thiele and Richard H. Small defined most of the relationships and terms we now use to describe what happens in a speaker and between a speaker and a particular enclosure's type and size. Their work has become the standard for speaker measurement criteria and is known as the Thiele-Small parameters. All speaker manufactures use the Thiele-Small parameters in describing their products which allow you to do a direct characteristics comparison of different speakers as well as give your the necessary information for designing the crossover network and enclosure.

The most commonly used Thiele-Small parameters are listed below:

EBP - Is used loosely to decide what type of enclosure will be best for any given speaker. It is calculated by dividing the Fs by the Qes. A result closer to 100 is usually best suited for an vented enclosure while an EBP closer to 50 will usually require a closed box design. This is just the "rule of thumb", some well designed high quality system violate this rule so use the EBP as a guide if the speaker manufacturer doesn't make a recommendation.

Fs - This is the free-air resonant of a speaker; it's the frequency that the speaker wants to vibrate at. This is a result of the weight of the moving parts (cone, etc) in balance with the stiffness of the speaker's suspension. At a speaker's Fs the speaker will over emphasize (make louder) that frequency and cause crossover points to change due to impedance variances. For accurate sound reproduction these frequency peaks must be controlled (kept flat).

Fb - The enclosure resonance (bass reflex).

Fc -The enclosure resonance (sealed enclosure systems).

Fp -Is the free-air resonant (Fs) frequency of a passive radiator.

F3 - Is the frequency where the response (loudness) is down from the reference level by 3 dB. Anything below this frequency is often too quiet to be useful; so F3 help defines a speaker's useful range. Look at the graph down below at the Qtc 0.707. The reference frequency is 90 dB, look along the line and look where it crosses the 87 db. The F3 for this speaker in a box with a Qtc of 0.707 is aprox 45 Hz

Lv - Is the length of the speaker enclosure's port.

Pe - Is defined as the maximum continuous (RMS) power-handling capability of a speaker.

Power Handling - Is rated on how much power a speaker can handle without causing damage. The most important consideration is the speakers ability to get rid of excessive heat. Factors that effect this

Page 3: Loudspeaker Builder

include magnet and voice coil size and their ability to handle heat, venting, and the adhesives used in voice coil construction.

Mechanical factors are also considered, such as the power required to cause;

1. The coil to hit the back plate or come out of the gap

2. The cone buckling from too much outward movement.

3. The spider bottoming on the top plate.

A speaker that has a "flat" response is a speaker that usually reproduces sounds accurately. A speaker that doesn't have a "flat" response is said to "color" (distort) the sound it reproduces. Some people actually prefer certain forms of "coloring" over the sound of an accurate "reference" type speaker. Home theater center channel speakers often sound best if the midrange frequencies are slightly louder, this will tend to improve the on screen dialogue; others prefer a loudspeaker with louder lower frequencies (boom box type). Even so it's generally better to have the loudspeakers response as linear (flat) as possible and if more base or midrange is desired you may add the distortion with an equalizer or the tone controls on the amplifier.

This graph below is what a perfectly flat loudspeaker response chart would look like; it has the same amplitude (loudness) across every frequency. If nothing extra is added or removed (coloring) the loudspeaker will accurately reproduce the original sound.; thus this is what most loudspeaker builders will aim for even though it's impossible to achieve.

With some knowledge and skill obtaining a response like the one below is about as good as it gets. A loudspeaker is considered linear (flat) if it stays within a range of about two dB's from highest to lowest. The frequencies outside of the blue lines have become too quiet and won't

Page 4: Loudspeaker Builder

affect the overall sound of the loudspeaker so they are not considered within the loudspeakers frequency range.

Their are many variables that influence the amplitude / frequency response that a loudspeaker will produce; these variables include the choice of loudspeaker parts, type of enclosure, cabinet material, crossover network type, crossover slope, the materials / furniture in the room and the room's size and shape. To compensate for room acoustics an equalizer or the tone controls on the amplifier can be used to "color" the sound produced by the loudspeaker.

Why Use A Speaker Crossover Network?

No one single driver is capable of reproducing the entire sound bandwidth faithfully. Because of this we divide the task among two or more drivers to get better sound quality. A typical three-way loudspeaker consists of a woofer, midrange and a tweeter. Each one of these components is designed to do a certain bandwidth of the sound spectrum. The woofer typically reproduces all the sounds under around 500Hz, the tweeter may take all the frequencies above 6000Hz and the midrange does everything in between. By dividing up the frequencies onto three different speaker components the overall sound quality is dramatically improved over that of a single component loudspeaker.

The crossover network is responsible for dividing up the sound bandwidth into specific frequencies and sending them to the proper speaker component; this is done by the use of coils, capacitors and resistors. Coils filter out higher frequencies and capacitors filter out the lower frequencies.

 

 

Speaker Crossover Network Design

Page 5: Loudspeaker Builder

This is without a doubt the most complex and difficult part of building your own loudspeakers. No matter what you do or how much you learn, or how much money you spend you will never build yourself anything near to a perfect passive crossover network. Capacitors and inductors just don't have what it takes to build the "brains" necessary to have a world class perfect crossover network. If you seek absolute perfection look into Active (Electronic) crossover networks, for the rest of us passive is usually good enough.

You really don't need any crossover network "brains" for a single driver type loudspeaker since there is no division of the f requencies between different drive units. Add a tweeter with the driver and things change a little, you now have a two-way design and the minimum amount of "brains" now required is a simple capacitor (High-Pass Filter). The capacitor is used to protect the tweeter from the lower frequencies (which can damage a tweeter) and still allow the higher frequencies to pass on through to the tweeter.

But this design can be improved upon; the mid-woofer is not a good choice for reproduction of high frequency sounds (that's what the tweeter is for, right!). So let's make life easier for the mid-woofer by not forcing it to try and reproduce sounds that it's not designed for. By adding a Low-Pass filter (coil) in series with the mid-woofer the loudspeaker's sound quality can be improved by limiting the high frequencies that are sent to the mid-woofer.

Let's take the crossover network to the next level by dividing up the frequencies previously sent to the mid-woofer. The midrange frequencies will now be sent to a midrange speaker and the low frequencies will be sent to a woofer, the high frequencies are still sent to the tweeter.

The midrange driver now only gets the frequencies in the middle of the bandwidth; for this a Band-Pass filter is used. A Band-Pass filter uses both a capacitor and a coil (inductor). The capacitor removes the higher frequencies and the coil removes the lower frequencies leaving only the midrange frequencies to pass to the midrange driver.

Page 6: Loudspeaker Builder

 

 

Speaker Crossover Network Points And Useable Frequency Range

This is the frequency range for which the speaker should have satisfactory sound reproduction. Different manufacturers use different techniques for determining what they believe is acceptable sound quality. Most of the methods used by the different manufactures are recognized as "acceptable" and it's "useful advertised" range rating depends upon who is rating it. Unfortunately many speakers are used to produce frequencies in ranges in which they are not capable of reproducing accurately due to manufactures overly optimistic speaker ratings

It is important to match up components to each other that will allow a good amount of frequency overlap between them. Don't set the crossover points too close to the "edges" of a speakers frequency range specs. Just because the manufacturer says that their woofer can go up to 1,500Hz doesn't make that a good place for the cut off frequency. At 1,500Hz that speaker is working very hard and rapidly becoming very inefficient, off axis coverage may suffer and in general the speaker is not going to be at it's best. That woofer would probably sound much better if you limit it to around 800Hz or so. Where the best cut off frequency is unique to each speaker; it is never ever found at or near the speaker's extremes. So be generous in allowing for overlap between the woofer, midrange and the tweeter.

This overlap is easy to obtain with a three-way and becomes more difficult with two-way loudspeaker. If you insist on going with a two-way then you will have to make some compromises. But then again, lately some manufactures are offering some outstanding components which are redefining what a two-way speaker can do.

Still the basic strategy is to get components that can easily do their job which unfortunately is usually the more expensive speakers.

Some other reasons why I like having a lot of overlap capacity between your tweeter, midrange and woofer are:

Page 7: Loudspeaker Builder

Crossover Networks need some "space" to do their crossing over; wherever you set the crossover point allow at least 300Hz on each side

Crossover points can and do vary due to impedance variances, better to be on the safe side and allow something for it.

Off axis coverage (sound quality) may suffer if you push a speaker to it's frequency limits.

 

 

 

Speaker Crossover Network Slopes

The perfect crossover network slope would look something like this; for this I choose a two-way 1800 Hz crossover. ALL frequencies below 1800 Hz go to the mid-woofer and EVERYTHING above 1800Hz is sent to the tweeter. This "perfect" crossover network is referred to as the "Brick Wall" filter and it is pure fantasy as far as passive networks go. Active crossover networks do work this way but they are very expensive to build and take up a lot of space because of the multiple amplifiers and other electronic components.

Below is a more realistic crossover slope for a passive crossover network. You may notice that the frequencies around the crossover point (1800 Hz) are now being shared between two separate drivers. The big problem this has as compared to the Brick Wall crossover is that now two separate drivers often of different size, shape and material are reproducing the same frequencies. Those frequencies will interact with one another causing distortion that affects the loudspeakers accuracy; this is referred to as Wave Interference.

Page 8: Loudspeaker Builder

Passive networks use capacitors and inductors which work by gradually removing (attenuated) the unwanted frequencies. How fast these unwanted frequencies are effected (attenuated) by the filter depend upon the design of the crossover. First order ones are the simplest, but they are also the slowest acting. As you increase the order of the crossover you will increase the ratio (speed) at which the unwanted frequencies are removed and lessen wave interference.

 

 

Crossover Network Filters (first order)

First order crossovers are often chosen over higher order designs because to their relative simplicity and response which is predictable due to the fact that first order crossovers are less affected by impedance variations than the other higher order crossovers. Due to the slow acting speed usage in a three-way system isn't recommended. Be aware that extra frequency space is needed around crossover points. The phase shift on a first-order crossover is 90 degrees which isn't a desirable network trait.

High-Pass Filter (first order) - The capacitors capacitive reactance causes an increases of resistance as the frequencies lowers. This increase of resistance to lower frequencies filters out the lower frequencies and allow the higher ones to pass. Used on tweeters.

Low-Pass Filter (first order) - The coils inductive reactance causes an increases of resistance as the frequencies rises. This increase of resistance to higher frequencies filters out the higher frequencies and allow the lower ones to pass. Used on woofers

Page 9: Loudspeaker Builder

Band-Pass Filter (first order) - The coil and capacitor filters out the lower and higher frequencies allowing only the mid range frequencies to pass. Used on midrange drivers.

Crossover networks work in a gradual way, they work by adding in frequency related resistance's which are used to filtering out certain frequencies, First order 6 db per octave crossovers will give you about a 75% reduction in power per octave (100Hz) and about a 94% reduction at 2 octaves. So at a crossover point of 400 Hz, the frequencies at 500Hz are still at 25% of full power and at 600Hz it's at 6% of full power. If a faster acting crossover is needed then you will need to build a second or third order filter. This is important when it comes to tweeters because they are much more sensitive and are more likely to fail if subjected to frequencies below what they were designed to handle.

 

 

Crossover Network Filters (second order)

Second order filters act twice as fast as a first order filter. At one octave (100Hz) above the crossover frequency, power at this frequency to the component at will be reduced by 93.75.% (12 dB level). The first order reduced the power by only 75% at one octave above the crossover frequency.

On second order crossovers (two way) the phase shift is 180 degrees (reversed polarity) and is known as phase reversal. In a two-way system basically the tweeter and woofer are firing out of phase with one another. This is easily corrected by reversing the wires on the tweeter. For a three-way second order system the midrange driver wires are usually wired in reversed leaving it "out of phase" (but not really) with the tweeter and driver. Doing this allows all the drivers to operate in phase with one another.

Page 10: Loudspeaker Builder

 

 

 

More Stuff On Crossover Network Phase Shifts

Higher order crossovers also have some phase shift; Third order (18db/octave) two-way crossover are at 270 degrees (-90 degrees), Fourth order (24dB/octave) two-way crossovers is at 360 degrees (No phase shift).

Even order (2,4,6, etc) filter are more desirable over odd ordered filters because even ordered networks will give you a multiple of either a 180 degree or 360 degree shift in phase. 360 degrees it's the same as 0 degrees so no phase shift there, to fix the ones at 180 degrees simply reversing polarity on the tweeter will put the drivers back into phase with one another.

 

 

 

Loudspeaker Impedance

Page 11: Loudspeaker Builder

The crossover network crossover points are calculated based upon the assumption of a consistent impedance load from each speaker; the problem is a speaker has a voice coil in it. Put DC current threw a wire coil and you will have a purely resistive resistance (after the initial build up); this type of resistance is a constant never changing resistance. Change DC to AC and things change, the consistently changing (alternating) direction of current flow causes a different type of resistance to occur in the coil of wire which is known as inductive reactance (XL). Inductive reactance doesn't resist current flow, it resists the change of current flow and it is a function of frequency; the higher the frequencies (Hz) the more reactance (more resistance).

That's bad for speaker builders because that's exactly how your speakers work, they make sound by oscillating at different frequencies. The frequencies sent to a speaker coil can range from a low of 20 Hz to a high of 40,000 Hz; so your speakers resistance and the crossover points will change according to the frequencies it's being asked to reproduce.

 

 

 

Methods Used To Control Speaker Impedance

Some of the methods that we can use in controlling the variance in the impedance peaks and the resulting crossover point shift of a speaker include;

Ferro Fluid - It's part liquid and part magnet. Used in the voice coil gap it effectively controls the impedance peak at resonance mechanically, also it dramatically increased the amount of power the speaker could handle due to better cooling. Sounds too good to be true? Ya it was; the problem was the fluid slowed down the speaker's response and made them "sluggish". A relatively new, very light less restrictive ferro fluid is now being used very successfully in some tweeters.

By Design - Design the speaker so the resonant frequency (Fs) is well above or below the speakers intended use. This way a manufacturer can put the Fs where it does the least amount of interference.

Zobel Filter - Used to compensate for the rise in impedance at high frequencies due to inductance reactance in the voice coil; All crossover calculations assume a purely resistive load across all the drivers so a change in the speakers impedance due to inductive reactance will negatively affect the way the crossover network works.

 

 

 

Special Filters (and other stuff)

Page 12: Loudspeaker Builder

Zobel Filter - Is used to "control" the variances in the speaker's impedance produced in the relative larger voice coils of a woofer and some midrange drivers. Tweeters may also benefit form a zobel filter; but if your tweeter has ferro fluid in it you won't need a zobel filter

L-pad - In general, as a speaker becomes larger it usually becomes more inefficient. Tweeters tend to be very efficient, midrange speakers are less efficient than tweeters and woofers tend to be the least efficient. The reason for this is that a woofers has to move a lot of air, for every octave drop in the audio spectrum the driver has to be capable of moving 4 times more air to keep the same output level (loudness). Tweeters don't have to move much air so efficiency is easily maintained. Woofers that are capable of 19 Hz and below are very inefficient compared to woofers that only go to 35 Hz.L-pads are often used on both the midrange and woofer to attenuate (reduce) their power level (loudness) to match the woofers efficiency (loudness). L-pads are variable control and thus easily adjusted after its installed

Attenuation Circuit - Also used to match the output levels of different drivers; not as easily adjustable as the L-pad. Tweeters are generally more efficient (higher sensitivity) than woofers, this could result in a overly loud tweeter as compared to the woofer. This circuit will attenuate (reduce) the tweeter output. You can also attenuate the tweeter by simply adding only the first resistor in series with the driver. Problem is that you just added more load to your amplifier, using the two resistors as shown won't increase the circuits impedance, only the impedance going to the tweeter.

Series-Notch Filter - Even though speakers are rated at a nominal impedance the actual impedance varies with frequency due to inductive reactance. The greatest amount of inductive reactance (change of a speakers impedance) occurs at the speaker's resonance (Fs) frequency. These impedance peaks causes the crossover network to shift their crossover points. This filter prevents this by controlling the impedance at the resonant frequency. It is used mainly on the

Page 13: Loudspeaker Builder

tweeter or midrange drivers and not on the woofer because of cost and size and the added resistance problem of the inductor

Parallel-Notch (trap) Filter - Are designed to remove broad peaks in the frequency response of a driver, a bit more complicated than the series-notch filter. Uses a capacitor, inductor (coil) and a resistor in series with the driver and parallel to one another.

Types of Speaker Enclosures

Infinite Baffle - Is defined as an enclosure that contains a larger volume of air than the Vas of the driver. You don't need to build a enclosure for this type of loudspeaker. You simply mount them in the wall and use the wall cavity as the enclosure or if were talking about car speakers the trunk becomes the enclosure. In general these enclosures have the least amount of sound coloration and they are inefficient because the rear energy which was never intended to be heard by the listener is "wasted". The important thing to remember is to have the front and rear of the speaker isolated into different air spaces so the air from the front of the speaker can't interact with the air at the rear of the speaker to avoid phase cancellation.

Sealed Enclosures (acoustic suspension) - Just like the name implies this is a sealed box with a speaker mounted in it. This type is preferred due to the simplicity of it's design which promotes a smooth frequency response, excellent cone control which translates into accurate sound reproduction. The only downside to this design is that it does this all at the expense of low efficiency. The volume of air within a sealed enclosure is less than the Vas of the driver; the air trapped in the enclosure helps control the movement of the cone somewhat like a shock absorber controlling the springs on a car.

Page 14: Loudspeaker Builder

Vented Enclosures (ported, bass reflex, tuned, or tuned ported) - This is a much harder design to "get it right" with (unless you got some good loudspeaker software). In the first two designs the wave at the front of the speaker was used while the wave at the rear of the speaker was "thrown away". Vented designs use this "thrown away' energy to enhance the overall sound of a loudspeaker (if properly designed) with better efficiency and deeper bass extension from sound produce in the port. This is done by tuning the ports (length and diameter) so the rear wave pressure will produce a lower tuned frequency in the port. Put the port somewhere close around the woofer, preferably just below it (within one foot). The direction the ports face (rear or front) is a personal design choice, and if the need arises smaller multiple ports can be used in place of a larger single port. Variations of this design include the use of multiple rear chambers tuned to different frequencies and passive radiators.

 

Passive Radiator - This is essentially a unpowered driver which replaces the port on a ported loudspeaker. It has the advantage over the port of less distortion and it allows the loudspeaker to act with precision of a sealed enclosure without the loss of efficiency.

 

Isobaric - In this design the speakers face one another, one of them is in the enclosure. It's also wired out of phase; the speaker in the enclosure is wired normally and the other one's wires get flipped around. This will allow them to move together in the same direction at the same time. This design may be used when enclosure size must be small. This enclosure can be half the size of the regular enclosure.

Dipole - This is for a "effect" rear surround sound home theater type of loudspeaker which "hides" the source of the sound very well. In this design the drivers are fired out of phase with one another. As a result their is a "nil" or dead zone due to phase cancellation in the space between the loudspeakers which unfortunately also kills a lot of the loudspeakers bass. This type of loudspeaker will give the desired effect only within a defined space and need special considerations in their placement. Not intended to be used as a general loudspeaker.

Page 15: Loudspeaker Builder

 

Bipole-This type differs from dipolar in that both drivers fire in phase with one another "spraying" sound everywhere. Loudspeaker placement isn't as critical as with the dipole, with a much wider listening space area and no loudspeaker bass reduction due to phase cancellation.

Bandpass - A combination of a sealed and a ported enclosure. Combining the two together will result in louder bass output than what the sealed enclosure alone would put out. This type of enclosure is not used as true subwoofer because the frequencies that come out of the port are a limited to a very narrow bandwidth. This design is often used to enhance the lower frequencies of smaller bookshelf type of loudspeakers.

 

Selecting The Best Drivers For Your Enclosure

This is very important and its worth repeating (Qtc 0.707 stuff); you must match your drivers to the enclosure and I am not just talking about the size of circular hole in the front of the speaker enclosure. Most drivers are now designed to be used in a specific type of enclosure with a specific volume (size). So when deciding upon your drivers you have to take into consideration the type of enclosure (ported, sealed) and the enclosure size recommended by the manufacturer. If their recommendations are not available you will have to take the other information that is given such as Fs, SPL, Vas, Qms, Qts, etc and figure that out for yourself. You can do that by using various complex mathematical equations or by simply plugging in the information about the driver into any of the many loudspeaker building software programs that are available.

Okay, you now know to match up your speaker drivers to the enclosure size and type recommended; you have to do that with the midrange speaker also if it is not-rear sealed . The reason why you build an enclosure within an enclosure for a non-rear sealed speaker is that you'll want to isolate the air spaces behind the driver and midrange from each other to prevent them from "pushing" on each other. The volume of air displaced by a large woofer moving in and out can destroy the much smaller midrange driver and even if it doesn't, they will interfere with one another causing distortions which is not a good thing to have happen.

Page 16: Loudspeaker Builder

Don't forget to add the total volume of the midrange enclosure to the speakers enclosure's total volume. Don't worry about building a separate enclosure for the tweeter; every tweeter that I have ever seen is rear sealed so the driver's air pressure won't affect it.

Two or Three Way Speakers

Has to do with the numbers of ways the frequencies are divided (the crossover network); and generally to the amount of transducers (speakers) used;

A Two-Way loudspeaker uses a driver and a Tweeter. In the Two-way design the work of reproducing the total sound spectrum is divided between the two components. The driver may take the frequencies between 45Hz-2400Hz; the tweeter then must be able to handle the frequencies from 2.4Mhz and up. Often a subwoofer is needed to enhance to lower frequencies because of the design limitations of the driver. This is because most drivers can't do a lot of different frequencies all at once well. A driver that can get low as 20Hz usually won't sound good at 1,500Hz. And most tweeters that can reach 30,000Hz or above often won't operate below 2,000Hz.

So most two-way loudspeaker are a compromise between the low and high frequency. But due to new advances in speaker technology a new generation of mid/woofers and tweeters have come along that have much wider frequency ranges than ever dreamed of before. Because of this quality two-way loudspeakers can now begin to compete with three-way loudspeaker systems.

A Three-Way loudspeaker uses a woofer, midrange and a tweeter and is still considered superior over the Two-way design. The workload of reproducing the sound is divided among the Three components. The woofer may handle between 20Hz to 800 Hz, the midrange handles the frequencies between 800Hz to around 5000Hz. The tweeter then is left to deal with everything above 5000hz. This design has no compromises like that needed in the two-way system. Three-way or higher is the only way to go for a true full range loudspeaker. Some people have taken this to the extreme and added more than three speaker components making a four or five-way loudspeaker!

The big problem with the "more drivers are better" idea is in trying to design an acceptable crossover network. Two-way crossovers are easy, three-way are much more difficult. Four-way and above higher order crossovers are usually only attempted by the "Pros" or hard core loudspeaker builders due to the "huge pain in the ass to design successfully" factor involved.

Loudspeaker Dampening

In every loudspeaker sound wave are produced at the rear of the speaker. Those waves will have their energy absorbed into the enclosure causing the enclosure to resonate (vibrate). This interaction between the standing sound waves and the enclosure is what causes the audible coloration and smearing of the loudspeaker's sound. In extreme conditions where the enclosure has no dampening materials the energy from the standing waves can approach the output of the drive unit itself. To control these waves we dissipate their energy through frictional losses with

Page 17: Loudspeaker Builder

any material which offers high resistance to the sound waves such as a damping sheet or acoustical dampening fiber fill.

Another very effective way to control enclosure resonance is through internal bracing. The best way of bracing is to design it into the enclosure from the start and not just add some cross bracing latter. Better quality cabinets use diagonal or planar bracing which bonds all four sides of the enclosure together. A simple often used method is to add bracing which is basically a board with a hole cut in it. The hole space in the center of the board are needed to allow air movement in the enclosure.

The loudspeakers shape also is also a factor in limiting enclosure resonance. A cube is the absolute worse shape (unless your designing a sub woofer) that a loudspeaker could be due to the standing waves produced inside the enclosure. In general try to pattern your loudspeakers to the golden rule ratio of (Heigh 2.618 : Width 1.618 : Depth 1) This ratio will minimize these standing waves and the resulting enclosure resonance and distortion. You don't have to match the ratio exactly, but the closer the better.

 

Speaker Box Enclosure Mass Loading

Adding weight such as dry sand or lead shot into a special compartment at the base of the enclosure will benefit the loudspeaker by increase it's effective mass. By adding mass to the loudspeaker structure it will act as an efficient absorber of vibrational energy. The heavier an object is the more energy it takes to make it vibrate; so weight is a resonance dampener. By experimenting with the amount of the fill material it is possible to fine tune the enclosure's tonal qualities to better match your own tastes or to help control some unpleasant sounding resonance produced by your loudspeaker enclosures. If you are determined to extract the absolute best performance that your loudspeakers can deliver, it is recommend that you mass load your speakers.

 

Speaker Box Enclosure Material

Most speaker builders would agree that medium density fiberboard (MDF) or medium density overlay (MDO) board is the product of choice. This is due to it's relative low cost, and because the stuff is strong. Don't use particle board or cheap quality plywood. Higher quality plywood that have no voids in the laminations are acceptable; you can spot poor quality by checking for missing pieces of laminates that have fallen out along its edges. Three quarters of an inch is the minimum board thickness that should ever be used because a strong and stiff enclosure is needed to minimize enclosure resonance.

 

Page 18: Loudspeaker Builder

Speaker Alignment within the Box Enclosure

The tweeter, midrange and the woofer should be kept in a straight vertical line with one another; this is done to keep the sounds in-phase with the listeners ear. The woofer sits nearer to the floor and the tweeter should be near ear level with the midrange in between the two. Mount all the drivers close together; the sound quality is usually improved due to less phase shifting and better imaging.

 

Make the Speaker Box Enclosure Air Tight

Make the enclosure airtight (except for the port of course). Leaks will show them self as unwanted air noise and in more extreme cases the leaks could effect the proper operation of the drivers and change the port frequency of base reflex loudspeakers. Use caulking to seal around the inner joints and any other place that air could escape.

 

Yeh, Size Does Matter

It's only logical that if you want big sound with deep clean base you need big speakers. I have read claims by some builders that their 6.5 inch based or smaller loudspeaker can match or exceed the sound quality and quantity of a much larger driver based loudspeaker.

The laws of physics say higher frequencies don't require much air movement, but for every octave drop in the audio spectrum the driver has to be capable of moving 4 times more air to keep the same output level (loudness). Simply put, deep base requires you to move a lot of air, something a small driver cannot do so don't fight the real facts, just accept them.

So now your considering using a 15 inch driver in your loudspeakers but are concerned that your loudspeakers will have too much bass and they will sound unnatural; don't worry, they won't. They will sound very similar to those quality 6.5 inch speakers at low to mid volume levels. They will just be more capable of delivering much louder clearer bass than the 6.5 inchers when asked. They will only sound boomy if your enclosure is too small; remember keep the Qtc to 0.707.

About the author - by Ralph Calabria

When I asked Lou to write this article, I was well aware of his knowledge of woodworking skills, both in furniture making and speaker building. Lou has been a woodworker for some time now, and the fine craftsmanship seen in his work is testimony. Lou's many speaker-building projects include: a 3-way tower design using an ACI AC-10 woofer, Focal 5N313 mid, and Scan Speak D2905/9000 tweeter; a design of

Page 19: Loudspeaker Builder

the NHT 1259 that looks more like a decorative pedestal than a subwoofer; and a budget 2-way using a Vifa 5" mid and a SEAS 1" tweeter. His furniture-building credentials include a bunk bed made from cherry, an armoire, and a mahogany nightstand.

Introduction

Speaker building is not just an exercise in audio engineering; it is also a construction project. For the beginner, obtaining the proper woodworking tools can therefore be as daunting a task as driver selection. This article will attempt to shed some light on basic woodworking tools. Each tool will be viewed primarily from a speaker building perspective, although any do-it-yourself project may be substituted. I will try to look at:

Features and Functions - variations, features, and functions typically available. Speaker Building Uses - how useful is this tool in our pursuit? Add-Ons - related accessories that make this tool work even better. Usage Tips - making the most of this tool for our projects. Cost - always a factor!

Of course woodworking tools are not limited to speaker building. These same tools can be used to build speaker stands, stereo cabinets, room treatments and traps, even complete sound rooms. For simplicity, however, the term "speaker building" will be used throughout this article to denote all of the above possible uses.

This article will not attempt to cover all aspects of tool usage or woodworking in general; nor will it address tools and techniques necessary for building exotic shapes and compositions. It is an article geared toward the beginner, and basic rectangular boxes are the assumed goals.

Since I'll be merely scratching the surface on tool usage, it's essential that readers take the time to read and understand each tool's user's manual. Taking the time to understanding the fundamentals of each tool, how it works, what its limitations are, and most important of all, how to use the tool safely, is the first step to building a successful project.

There is no way for me to cover all details of each tool. Manufacturers are constantly upgrading their wares with better features and innovations. My goal is, therefore, to cover most of the important points that I consider relevant to speaker building. As always, there will be exceptions to every rule, so keep that in mind.

Tool prices provided are approximate and reflect the market as of this writing. Prices are intended to give readers a feel for the relative and absolute costs of the tools. Some brands and models may cost significantly more or less than others. All prices are in US Dollars and do not include any taxes, shipping or handling charges.

Page 20: Loudspeaker Builder

I've tried to avoid using brand names except in the URL section. The intent is to concentrate on the characteristics of the tool, not the actual commercial offerings. Woodworking magazines regularly review tools and are the best source for up-to-date side-by-side comparisons.

A Note on Safety and Maintenance

All tools are potentially dangerous. They should be treated with a healthy dose of respect and care. Always read and understand the manufacturer's directions on both safety and usage before using a tool. Always use appropriate safety equipment including eye (safety glasses), respiratory (mask), and hearing protection (earplugs). Always maintain the proper work environment, and give your tool and work your undivided attention. Remember - safety first !

Proper care and maintenance is required on all tools. Refer to the manufacturer's directions for such information. Keeping tools in tip-top shape not only maintains tool life and maximizes performance, it also helps prevent accidents and potentially dangerous operating conditions.

Motors & Horsepower

This section is a very simple look at power tool motors. I won't go into great details - just enough to guide you through some of the numbers you'll likely see when shopping for power tools.

There are two basic types of power tool motors: induction motors and universal motors. The induction motor is typically found on large tools and can deliver fairly consistent and reliable power. It is characterized by its single speed and large cylindrical case. Induction motors are usually used on drill presses, table saws, jointers, and other floor standing machinery. Changing speed, when allowed, usually means moving belts between pulleys of varying sizes. By comparison, the universal motor is small and loud. Its size and weight makes it ideal for handheld tools like portable drills and routers. With no load, the universal motor usually spins at a rather high rate. Under load this rate drops dramatically, and the motor heats up quickly.

Both types of motors are rated in horsepower, but the actual numbers can be deceiving. For example, a typical 115 VAC induction motor in a contractor table saw drawing 14 amps may be rated at 1.5 Hp. The same 14 amp in a router with a universal motor might be rated at 3 Hp. Suffice it say that universal and induction motors should not be compared with each other. When comparisons of any sort must be made, make them when they are of the same type, and use the motor's current draw for comparison, not the horsepower rating. Avoid making judgements based on small differences in current or horsepower draw. Large differences in horsepower or current rating are more likely to be accurate on a relative basis. For example, a 3 Hp router is likely to be more powerful than a 1.5 Hp router, though not necessarily twice as powerful.

Page 21: Loudspeaker Builder

Induction motors come in various flavors. The best are Totally Enclosed Fan Cooled (TEFC) motors. Most large power tools from respectable companies come with such motors. Less expensive tools often come with Drip-Proof motors. These offer less protection from the environment than TEFC motors. In some applications, explosion-proof motors are required. Typical uses include exhaust fans used in handling combustible gases.

Grades of Tools

What differentiates a consumer grade tool from a professional grade tool aside from price? Are these differences worth the extra cost? Let's take the easy question first and identify some differences between consumer and professional tools. Typically, professional tools:

Are designed for long life under tougher environmental conditions. This may include motors capable of withstanding higher temperatures, use of better dust seals, bearings and lubricants, and stronger, more durable construction.

Deliver better overall performance, including motors that provide smoother power with less vibration.

Tend to have more accessories, options and features. May use more advanced and expensive materials such as light-weight alloys or

use of cast instead of stamped metal parts. Are designed with serviceability in mind.

Should the hobbyist invest in costlier tools? This is a personal question, and the answer will vary with each individual. Variables to consider include the predicted amount of actual tool usage, the care one puts into proper maintenance, the desired accuracy from each tool, the desired work efficiency, and of course cash flow, to name but a few. Consider such variables before spending any money. In many instances, a pro-sumer or consumer level tool is all that's needed. In other instances, professional tools are a worthwhile investment.

Note: In the various tools discussed below, my comments assume a "decent" level of tool quality. Poor quality tools usually lead to poor results no matter how hard one tries. This does not imply top-notch tools, but rather, just good enough, relative to the user's expectation. This is obviously a tricky statement to make, and impossible to quantify in a general sense.

Table Saw

It's easy to understand why so many consider the table saw to be the indispensable saw in the shop. With a table saw, one can make an abundant number of cuts, most with a high degree of precision. Jigs and accessories can make the table saw invaluable.

Page 22: Loudspeaker Builder

Features and Functions

I group table saws into three categories: the bench saw, the contractor saw, and the cabinet saw. The cabinet saw is the largest and most accurate. It usually comes with a full-size polished cast iron top, a totally enclosed base, a large and accurate fence, a large 230 or 460 VAC single or three phase TEFC motor, a 10 to 12 inch blade capacity, and plenty of working surface. It is used primarily by professionals, and by hobbyist standards, is quite costly.

On the other extreme is the bench table saw. This is usually a small device designed to be clamped onto a benchtop or mounted onto an optional floor stand. It operates off a regular outlet (~115 VAC in North America) and consequently is limited in motor size. The blade is usually 8 to 10 inches in diameter, and the fence may need some tweaking to get that great cut. Bench saws are great for the hobbyist with limited space or needs. Because of their portability, bench saws are also a great professional job-site saw.

Between the cabinet saw and the bench saw is the contractor's saw. Like the bench saw, these saws usually operate from regular household outlets. They also offer some of the added size, weight, and stability of the cabinet saw. The contractor's saw typically has a cast iron surface with an open frame. The typical blade size is 10 inches. Fence quality varies, though after-market fences are usually available as upgrades.

There are many qualities that makes one saw better than another. Here are a few in no implied order:

Cast iron top - a well-made saw is ground and polished flat. It will take lots of abuse, and with proper care, will last a lifetime. Stamped steel or aluminum, while lighter, do not provide the mass needed for stability and accuracy.

A straight, easily adjustable fence - the job of the fence is to keep a parallel surface with respect to the blade. Some believe that the fence should be slightly off parallel - the rear being further from the blade than the front, thus ensuring that the rear will not touch what's already been cut. No matter what you believe, it's important never to let the rear of the fence be closer to the blade than the front. Some fences have micro-adjustments that make small incremental fence changes easier to make. Some fences lock only on the front rail, while others lock on both the front and rear rail. Most of the quality fences lock only in the front.

TEFC motor - for 115 VAC, the most you'll get is about a 1.5 Hp induction motor. Anything larger is usually wired for 230 VAC operation. Universal motors are not common in table saws and are not a good choice.

Speaker Building Uses

Page 23: Loudspeaker Builder

Table saw uses should be fairly obvious. Some typical uses are:

Cutting large sheet goods down to size. Cutting dadoes and grooves. This is usually done for nice, strong, clean, tight

fitting joints. Cutting miters for corners and other "odd" angled panels. One or more jig or

accessory may be needed to perform these tasks accurately.

Add-Ons

Here are some useful accessories:

One potentially very useful and important safety accessory (when not standard equipment) is a magnetic switch. A magnetic switch consists of a pushbutton on/off switch that engages power via a relay. It operates like a regular switch except when power is interrupted. Should a partial or full power failure occur, the relay opens and power is cut to the motor. When power comes back on, be it a split second or hours later, the relay does not re-engage until it is turned on again. This prevents sudden restarts caused by brown-outs, thus preventing dangerous kickback.

A better after-market fence can dramatically improve the performance and accuracy of a saw. Not all saws can accept after-market fences, so check availability before buying.

Dust collection does wonders to air quality in a shop. A table saw is capable of throwing an enormous amount of dust into the air. Dust collection removes most of the small and troublesome particles. Some saws come with built-in collection hookups, while others require some user hack.

Hold-downs are a great safety and precision addition to a saw. Most hold-downs clamp onto the fence and provide downward pressure to the tabletop and sideways pressure to the fence. Should kickback occur, the hold-downs limit the amount of travel on the stock, preventing it from going airborne.

Stand-mounted rollers placed on the side and rear of a table saw can effectively extend its capacity when cutting large sheet goods. These rollers are not expensive and normally have many other uses around the shop.

A panel cutter is a jig that rides in the saw's T-slot, which provides right-angle cuts for large sheet. This jig differs from the standard T-slot miter gauge mainly in size and purpose. While the miter gauge allows you to set any cut angle, the panel cutter is usually made just for right angles. Most woodworkers make their own panel cutters.

A tapering jig forms a wedge between the stock and the fence resulting in an angled cut. This jig is useful for making panels with non-parallel sides. A speaker with a sloped baffle might require such a panel on each side of the cabinet.

Zero clearance insert - The insert is a plate through which the blade rises on the table. A zero clearance insert starts out as a solid piece of material through which the blade is slowly raised. The result is an insert that has a minimum amount of

Page 24: Loudspeaker Builder

clearance all around the blade, thus maximizing support for the wood as it is cut. This is easily made in the shop from scrap hardwood.

Usage Tips

Many of the items listed below are usually mentioned in the user's manual:

Use the anti-kickback device supplied with your saw whenever possible. Use the splitter supplied with your saw whenever possible. Splitters are usually

temporarily removed or displaced when cutting dadoes or grooves. Be sure to reattach them after completing the cut.

Use a zero clearance insert when working with small items that may require extra support near the blade.

Always wear eye and ear protection. Never stand in line with the blade. Never force stock through the blade; let the blade do the cutting. Wear a full face shield (in addition to eye protection) when cutting particle board

or similar material. Flying chips can strike the exposed skin on your face. Use dust collection when cutting medium density fiberboard (MDF). MDF

releases a lot of very fine dust when cut. A respirator can keep this dust out of your lungs, but won't keep the dust from coating your shop surfaces.

The blade should rotate down into the stock meaning that any tearout occurs on the underside. It is, therefore, good practice to put the good side of the stock up. If tearing on the bottom is a problem, try adding another layer of scrap stock on the underside and cut through both layers. Masking tape along the cut line can also be used so long as it does not damage the stock when it is removed.

Cost

Bench saws: usually under $200 Contractor's saw: $500 to $1,000 Cabinet saw: $1,000 and up

Router

The router is a portable, multi-purpose power tool. Functionally, it's a very simple device, but with the proper jigs and/or attachments, a router can perform a variety of tasks. For example, routers can cut dadoes, mill moldings, cut biscuit slots, create dovetails, drill holes, join edges, and even plane surfaces flat.

Features and Functions

There are fundamentally three types of routers - fixed base, plunge base, and D-handle:

Page 25: Loudspeaker Builder

Fixed-base routers are the most basic. They usually have a pair of handles near the base of the unit. The extension of the bit (the depth of cut) is set before use. Depending on the make and model, the fixed base router may be easier to use when mounted in a router table.

D-handled routers are similar to fixed-base routers except that the router is controlled mainly via a D-shaped handle. These are probably the least common of the three types.

Plunge routers are the most popular and allow the bit to plunge into the wood, hence their name. The plunge depth is adjustable, often in very small increments. This is clearly the most useful of the three types, but also the most expensive. Most high power routers are plunge models, and not all plunge routers are suitable for table mounting.

Routers fall into two groups based on their power. Low power units range from just under 1 Hp to about 2 Hp. High power routers range from 2.5 Hp to about 3.5 Hp. From a current draw basis, all routers typically draw from 10 to 15 amps. High power routers tend to be bigger and heavier. From a capability standpoint, the rated power of a router does not, by itself, define what it can do, or how quickly it can do the job. However, advanced features, options, and accessories are more likely to be available on larger, pricier, and more powerful models.

Some features that differentiate routers from one another include:

Page 26: Loudspeaker Builder

Soft start - reduces the kick produced when the router starts up. In applications where the bit must make contact with the stock when the router is turned on, this feature helps reduce the likelihood of gouging the stock.

Variable speed - a typical router can spin a bit (with no load) upwards of 20,000 rpm. This can be very dangerous with larger diameter bits since the wider the bit, the faster the perimeter of the tips travel at a given rpm. Variable speed is a "must-have" when using large diameter bits. Slowing the bit speed also reduces friction and thus reduces the likelihood of burning the stock from friction.

Collet size - router bits come in 1/4 and 1/2 inch diameter shanks. The larger shank offers more stability and is therefore used with larger bits. Smaller bits are offered either in 1/4 inch shank only or in both sizes. Some routers only have a 1/4 inch collet, thereby limiting bit selection while others provide both 1/4 inch and 1/2 inch collets.

Micro-adjustment - Some routers make fine depth adjustments very simple and accurate. The micro-adjustment feature is very handy when attempting to get precise cut depths.

Speaker Building Uses

There are many important uses for a router. Some typical ones are:

Cutting circular holes for drivers in a baffle. Milling cabinet edges. This is usually a roundover or chamfer. Cutting grooves and dadoes inside cabinets to accept shelf bracing. Cutting rabbets to cabinet side panels for panel joinery.

Add-Ons

Because the router is so versatile, there are many accessories available. Many can be made in the shop.

A circle-cutting jig is a popular homemade jig due to its simplicity. The concept is straightforward, and can be made in a variety of ways. One example is to use a board that has a router mounted on one end, and pegs placed some distance away from the router's bit. The assembly's peg is inserted into a hole in the stock to be routed and acts as the center of a the circle. By varying the distance between the router and the peg, the radius of the circle that the router will mill is changed. Commercial circle cutters may be mounted to the router in place of the base plate or may be mounted to some other point on the router's lower section. Radius adjustments should be easy to make, and may contain a fine-adjustment feature for precise settings. This is a "must-have" accessory for speaker building.

An edge guide is useful when milling dadoes and grooves that are parallel to one edge of a panel. The guide follows the edge of the panel, maintaining a parallel and repeatable cut. Some edges can also ride in an existing groove, thus allowing one dado or groove to be cut parallel to another. Never try to cut a dado or groove freehand. Always use a guide or follow a straight edge.

Page 27: Loudspeaker Builder

For some users, a router table provides a more secure and stable platform for routing. A router table holds a router upside-down with the router mounted from the underside. The bit sticks out of an opening through the table. The depth of cut is preset before milling. It's not practical, or safe to plunge up from below while the stock is sitting above the bit. The stock is moved over the router in full view of the bit. The benefit of the router table is in the added stability provided by the table, more precise control, better visibility of the bit, and increased contact area between the stock and router. Many router tables have fences, dust collection attachments, and clear polycarbonate plastic shields that cover as much of the bit as possible for safety. More advanced tables provide features found in large shapers such as a T-slot. Router tables can be stand-alone or built as part of a table saw surface. Many of the parts needed for a home-made router table are readily available at woodworking stores.

Usage Tips

Here are some tips for using a router:

Never try to remove too much material at once. It's better to use multiple passes, each removing a little more material. How much is too much? This is partly a matter of experience and partly a matter of common sense. If the stock offers a lot of resistance, or if your bit clogs up with sawdust forcing you to stop, you may be trying to remove too much stock at once. Also, check the grain of the stock and see if it's likely to tear with the bit rotation. You can minimize tearout by removing less stock per pass. As a rule, I like to make at least two passes on each cut, with the last one removing a minimal amount of material to clean up any aberrations left over from previous cuts.

Whenever possible, move the router in the direction such that the bit's rotation is against the direction of motion. This forces the bit into the wood instead of allowing the bit to pull the router along. The user's manual should diagram this since it's one of the most important points to remember when using a router. Make sure you understand this.

For our uses, this translates to the following:

If you're milling the inside edge of a driver hole (e.g., milling a recess for the driver flange with a rabetting bit), move the router in a clockwise direction inside the cutout.

If you're milling the exterior of the box (e.g., milling a roundover on the baffle edges), go counterclockwise.

When milling end grain, it's easy to tear the wood at the edges. One simple way to avoid this is to clamp scrap wood of the correct size to the end of the stock and mill it with the stock. This causes any tearing to show up on the scrap, not your stock.

When using a router to mill a round driver's opening in a baffle, do all your baffles together. Setup is crucial, so take your time, set it up right, and then do all baffles the same way.

Page 28: Loudspeaker Builder

If you have spare stock, mill one or more extra baffles. These need not be full size - just large enough to hold the driver's opening. The purpose of these spare baffles is to let you test your setup when you have multiple routing steps. Use the spares when proceeding from one step to the next to test your setups without risking your real baffles.

If you're flush-mounting the driver, be sure to allow some extra depth for the mounting gasket.

When cutting the hole for the driver's basket, make sure not to cut through without taking some precautions. The circle cutting attachment is mounted to the center of the driver's opening and will be loose when it's cut free of the baffle.

Two ways to solve this problem are: o Mount some scrap on the underside of the baffle temporarily to secure the

center waste to the rest of the baffle. Remove it after you've routed the circle and separated the center waste from the baffle.

o Don't mill all the way through the baffle, leaving about 1/16 inch holding the center in place. When you're done routing, cut through the remainder with a jig saw or hand saw (a keyhole saw works well). The cut may not be perfect, but since it's on the inside of the baffle, it's not visible and won't affect performance.

If you wish to bevel the inside of a driver opening, use a pattern-following bit. This bit has a bearing that rides along the inside rim, and removes stock evenly around the circle.

When possible, get 1/2 inch bits instead of 1/4 inch bits. They do tend to cost more than 1/4 inch bits (all else being equal) but are more stable.

Carbide bits are a worthwhile investment. Wood products such as Medium Density Fiberboard and particle board eat High Speed Steel (HSS) bits. There are many advanced bit designs including anti-kickback features, low friction Teflon ® coating, and modular construction. Invest in these bits if your budget allows.

Cost

Professional high performance, high power plunge router: $200 to $300 Professional fixed base low power router: $100 to $200 Consumer router: under $100

Portable Circular Saw

The circular saw is probably one of the two most basic and common power tools around (the other being the drill). It is the ultimate fast cutting tool - portable, capable, flexible and affordable. It can be used in the shop or on a job site, and may be used for rough as well as highly accurate cuts.

Features and Functions

Page 29: Loudspeaker Builder

Circular saws tend to look more or less the same. Yet there are differences in their designs.

The majority of circular saws on the market today have side-mounted motors. These are the familiar saws with the blade directly in-line with the motor shaft and housing. Power is typically delivered from the motor to the blade via reduction gears. Less common are worm-drive saws often used by professionals needing higher torque and a narrower saw profile. The motor is mounted behind the saw blade, parallel to the direction of cut, and connected via worm gears.

The grip on a saw is either on top or behind the motor. Professionals often prefer the latter design since it gives them more reach. Either design offers a second grip or knob on the front of the saw for two-handed operation.

Many saws come in both right and left-handed versions. The depth of cut is controlled either via a drop foot or a pivot foot. The pivot foot

is easier to use, but changes the angle of the handle relative to the stock. Some advanced saws have an electronic brake to quickly stop the blade when

power is cut. This can be a valuable safety feature as well as a time saver when making many consecutive cuts.

For keeping dust in check, some saws have adjustable dust ports that can be aimed away from the work or even attached to some sort of dust collection system.

Speaker Building Uses

For serious DIYers, the circular saw can be seen as a poor man's table saw. But like its bigger and more expensive brother, this saw can make excellent cuts with the right jigs. Even dadoes are possible, though they require much more care in setting up the cuts. Cutting small stock is difficult with this saw, so use a hand saw for safety.

Add-Ons

I tend to think of the circular saw as a straight cutting tool without a fence. The flat blade of this saw dictates a straight cutting path, yet the saw is really free to go anywhere, thereby risking binding and kickback. So, the most important accessory to this saw is a straight edge. A straight edge can be anything from a pre-made commercial unit with built-in clamps, to a piece of plywood with a factory edge. The important thing is that the saw has a surface to ride against, thus keeping it in a straight path.

Usage Tips

Page 30: Loudspeaker Builder

One notable difference between a table saw and a circular saw is the direction of the blade's rotation. As mentioned earlier, the table saw blade rotates down into the stock. The blade on a hand-held circular saw rotates up from the bottom and through the stock. This means that tearout occurs on the top side of the stock. You should, therefore, cut your stock with the good side down. To avoid tearout on the top side, use masking tape along the cut line (assuming the material will not be damaged by tape). Adding a second layer of scrap stock on the top side may be done if the layers are properly supported and secured.

To cut a straight line, use the aforementioned straight edge. Clamp the straight edge down as solidly as possible. Some saws have accessories that make ripping parallel sides easy. Check the manufacturer's catalog for these accessories.

Cost

Professional heavy duty, high performance saw : $150 to $200 Professional saw : $100 to $150 Consumer saw : under $100

Jig Saw

Functionally, the jig saw is very similar to the circular saw except that it cuts much more slowly and can cut curves. This saw is sometimes referred to as a sabre saw. Jig saws are very portable, relatively quiet compared to their larger circular bladed cousins (important to apartment dwellers !), and safer in many respects. What they lack in cutting speed, they make up in flexibility.

Features and Functions

There are several types of blades, not all of which are supported by all vendors. The universal blade is the most common. It has a straight shaft with a hole at the mounting end. The hook mount is supported by Porter Cable, and as its name implies, has a 90 degree hook for mounting. The 'T' or tang or bayonet mount has a mounting system shaped like a small sword. Each blade type does the job well. The main differences are in the ease of changing blades. Whether such ease is worth the distinction of one make and model over another is a personal matter. If you change your blade often, a one-step, toolless changing system can be a time saver. But if you're a rare user of a jig saw, it's probably not worth much as a unique differentiating feature. What is important is that any tool needed to change a blade (if any) be readily available.

Page 31: Loudspeaker Builder

Simple and inexpensive jig saws have a single reciprocating stroke. More advanced saws offer two strokes - reciprocating and orbital. The reciprocating motion is a simple down and up stroke. This offers the smoothest but slowest cut. Orbital motion causes the blade to push down in a straight motion, but pull back up with the blade angled forward. Since jig-saw blades cut on the upstroke, this has the effect of cutting more aggressively. Saws that offer orbital motion usually allow more than one orbital setting. This setting varies the aggressiveness of the up-stroke.

Simple jig saws have only an on/off switch. Costlier models allow control of the stroke speed either by the amount of trigger pressure (as in a drill), by a dial setting, or both. If a dial setting is used, it should be mounted in an easy-to-reach place so that speed adjustments can be made during the cut. Some models allow a number of fixed speeds, but offer no continuously variable settings.

Jig saws remove waste slowly and do not tend to throw large amounts of dust. However, since they cut on the up stroke, dust usually accumulates on the cut line. Some saws have small blowers that push this dust aside while you cut. Others provide dust collection hookup.

Almost all jig saws offer a base that tilt up to 45 to the left and right. Usually a 90 stop is provided, as well as possibly other common angles. At least one manufacturer provides a rigid non-tilting base. Fixed-base jig saws offer a more stable platform and remove the possibility of accidental tilting.

Most jig saws have a top-mounted D-shaped handle. This handle may be placed almost on top of the blade assembly or it may be placed further back on the body. Another grip style is the body grip, where the unit's body is used to hold the saw. This grip places the hand closer to the work surface, usually offering more accurate control of the saw.

Blades - Some are made for specific uses such as wood or metal cutting. Blade materials and teeth specifications also vary with application and cost. If the jig saw is your primary tool, invest in quality blades. If this is an occasional tool more likely to be used for rough work, you can be less selective with your blades.

Speaker Building Uses

There are several ways to think of the jig saw. For someone on a limited budget, it can be the one and only power tool. With care, it can cut holes, curves, straight lines, and even bevels. Its accuracy is almost entirely a function of the user; and any resulting ragged edges can often be sanded down.

Where one's budget is not an issue, the jig saw is more useful supplementing other power tools. Its ability to cut curves and bevels make it ideal for roughing out holes for surface-mounted drivers, especially those requiring odd shapes.

Add-Ons

Page 32: Loudspeaker Builder

Jig saw accessories are few, and tend to be very manufacturer specific. Below are some possible accessories. If any of these are important to you, be sure to check with the manufacturer or dealer before you buy:

Dust collection hookups (when not standard equipment) Edge guide to cut lines equidistant from an existing edge Circle cutting guide similar to that used on a router but made specifically for a

particular jig saw Rail cutting system where the saw rides on a clamped rail to cut straight lines Anti-splintering device that keeps the stock down just in front of the blade

Usage Tips

Blade support and deflection - A jig saw blade is mounted only at the top and may be influenced in its path by the stock. Saws provide varying levels of blade support either from the rear of the blade, from the side or both. The goal is to keep the blade as straight and stable as possible. Cutting tight curves, especially on thick stock, will always place strain on the blade. The greatest strain and deflection will occur at the tip (bottom) of the blade furthest from the saw body, and will tend to cause the blade to bend towards the inside of the curve. Be aware of this and cut slowly to minimize blade deflection.

Starting a hole - One of the jig saw's greatest assets is its ability to cut inside an enclosed region such as cutting a rough opening for a driver in a baffle. There are two basic ways to start the cut. The simplest is to drill a hole large enough to insert the blade of the jig saw. If a hole cannot be drilled, the jig saw can be used to cut a slit into the stock. This is done by leaning the saw forwards at a steep angle such that the tip of the blade is barely making contact with the stock. With the motor running, slowly tip the saw back towards its normal position, thus pushing the blade down into the stock. Do this slowly. The drawback of this technique is that it will not work in tight spaces.

Cutting straight line - A jig saw can be used to cut a straight line with the use of a straight edge guide. A dedicated guide made specifically for the jig saw is best but any straight edge clamped to the stock may be used. Straight cuts are a good use of orbital strokes.

Cutting curves - The sharper the curve, the less likely an orbital stroke will prove suitable. When in doubt, use a reciprocating stroke t.

· To help limit splintering - Use masking tape and draw your cut line directly on the tape. This is no panacea, but should help. Test first to make sure the glue from the tape will no stain the stock you're using.

Cost

Professional high performance jig saw: $100 to $175 Consumer jig saw: under $100

Drill

Page 33: Loudspeaker Builder

The other highly popular home power tool, the drill, is an incredibly simple tool that should require little explanation to anyone interested in DIY.

Features and Functions

Cordless or corded - While the cordless drill has replaced the corded drill in many applications, a corded drill is still a valuable tool in any shop. Corded drills still offer much more power and torque than their battery-powered counterparts. They also tend to be cheaper and you never have to worry about dead batteries... just power failures! For the hobbyist on a limited budget, a basic corded drill is a good investment whose only drawback is its umbilical to the wall. The more advanced hobbyist with lots of uses around the shop and home will probably find the cordless drill more useful. Ultimately, your possible uses will dictate what to buy. Heavy vs. medium vs. light duty, features, and portability are probably the key considerations, though by no means the only ones.

Keyed or keyless - Traditional drills require a key to tighten a bit or shaft in the chuck. Keyless chucks only require hand tightening. For the most part, keyless chucks have become standard equipment on most cordless and many corded drills. From my experience, a quality keyed chuck will always hold tighter than a keyless chuck. In some applications, this may be important. For the audio hobbyist, this is not likely to be an issue except in some extreme cases.

Built-in clutch - A drill with a built-in clutch becomes a very useful screwdriver. Clutches are usually adjustable over a wide range of torque settings, typically by turning a ring near the chuck. The clutch allows the drill shaft to rotate only when pressure is applied and only up to the chosen torque setting before disengaging. This is very useful for driving applications - clutches can prevent the overtightening of screws or the stripping of screw heads.

Variable speed and reversible - Variable speed reversible (VSR) is standard today. It may not be possible to settle for less! The reverse direction can be used for extracting screws or for removing stubborn drill bits from their hole. The reversing switch should be easily reached by the hand on the trigger.

Voltage and torque - Cordless drills are usually listed by their operating voltage. Voltages are usually multiples of 1.2 V - the voltage of a single Nickel-Cadmium cell. Most of the units less than 12 volts are better suited for general home use. For serious hobby use, go with 12 V or higher. Torque is sometimes advertised along with voltage. Most cordless drills have a two speed range that trade speed for torque. Typically, torque is not that big a factor when comparing drills of similar voltages.

Page 34: Loudspeaker Builder

Handle location - There are two types of handle arrangements for drills - pistol grip and T grip. Each has its proponents and detractors. The T grip provides better balance while holding the drill. Users who are physically weaker will likely find this shape more comfortable especially when drilling in an awkward positions. In my opinion, the pistol grip provides better directional force and control when drilling and driving. The uneven weight distribution is countered by the support provided by the drill bit biting into the stock, while the angle of the pistol grip allows more force to be exerted. If you are unsure when buying, try both styles, but remember to try them with real bits and real wood. Simply holding it for balance doesn't mean much since it's the act of drilling that actually is important. Try different drilling positions if possible, such as holding it sideways or overhead.

Kits - Cordless drills are often available in kit form. This is largely due to the need for a second battery while the first charges. Kits usually contain two batteries, a charger, drill and case. If you must buy a cordless drill, a kit is usually worth the investment if cost is not an issue. If funds are tight, most drills are also sold with a single battery and charger.

Brakes - Some drills stop on a dime via an electric brake. This can be a very useful feature when doing many consecutive operations. For normal use, it's a nice feature but not a necessity.

Speaker Building Uses

The obvious use in speaker building is drilling holes (pre-drilling is a good idea to avoid splitting stock), and driving screws. Do not use portable drills with bits intended for the drill press. For example, it's possible to mill very accurate circles with an adjustable one-armed T-shaped circle cutter. However, this is strictly a drill press tool and using it on a hand held drill is very dangerous.

Add-Ons

1/4 inch hex extenders - Most driver and socket bits are very short. An extender chucks into the drill while its female end magnetically holds the bit firmly in place.

Drill/driver units - There are several jigs on the market that allow the user to switch from drilling to driving very quickly. They vary in design, price and usability. Each typically allows the user to drill a hole with a countersink, then switch to the driver without having to touch the chuck.

T-square - The visually challenged may want a small T-square or other right angle tool to guide the drilling angle.

Usage Tips

If available, choose the correct speed setting. The slower setting gives more torque.

Excessively high drill speeds can burn wood and dull bits unnecessarily. · Reversing the bit direction makes bit extraction easier.

Page 35: Loudspeaker Builder

Countersink screws to hide the screw heads. Start with a low setting when you're unsure of the torque needed to drive a screw

into a particular stock.

Cost

Professional high performance, high voltage cordless drill kit : $200 to $250 Professional high quality cordless drill kit : $150 to $200 Professional high performance corded drill : under $100 Consumer corded drill : under $50

Saw Blades

I've decided to place saw blades in its own category in order to highlight the possible choices. The simplest shop can get by with a single all-purpose combination blade. On the other extreme, the professional shop might have many highly specialized bladed in addition to the general purpose ones. For the speaker builder, it's possible to get by with one blade, but having one or two special blades can be very handy.

Blades come in various sizes. The two most common sizes are 10 inches for table saws, and 7-1/4 inches for hand held circular saws. Both are typically made for a 5/8 inch arbor thus the smaller 7-1/4 inch blade can be used in a 10 inch table saw though the reverse does not apply.

Features and Functions

Stack dado head blades - One very important accessory for a table or radial arm saw is the dado blade. Whereas a typical circular saw blade cuts a path as wide as its teeth allow (the kerf), a dado blade cuts a wide path through the stock. The stack dado head cutter usually consists of two full size blades sandwiching a number of chipper blades. The chippers are usually not full blades and their job is to remove the stock between the two edge blades. By varying the number of chipper blades and their widths, the width of the cut is changed. The typical stack dado head set is 6 to 8 inches in diameter.

Wobble dado blades - The wobble dado blade performs the same function as the stack dado blade but has two very important differences:

1. The wobble dado is a single unit, not a number of blades that are stacked to perform the task. Wobble blades usually have one or two blades that are tilted to carve out the full width of the cut. The side affect of this action is that the bottoms of the dadoes cut are not flat; but are curved. Furthermore, the lack of full contact side blades means the edges of the dado are not often as crisp as that made by a stack dado head cutter.

2. The width is continuously adjustable. It is not limited by the width of the chipper blades. Thus, fine tuning the width is much simpler with a wobble dado than with a stack dado blade.

Page 36: Loudspeaker Builder

Thin kerf blades - Thinner blades remove less material as they cut, and therefore require less power to use and produce less dust. They are also not as stiff as their regular kerf counterparts, so they must be more carefully designed to avoid warping with heat buildup. Blades often have expansion slits to avoid such warping.

Ripping blades - Blades made specifically for ripping tend to have fewer teeth (24 is typical on a 10 inch blade). This allows more room between the teeth for waste removal leading to a faster cut.

Crosscut blades - In order to avoid tearing the wood fibers during a crosscut operation, these blades tend to have many teeth (60 to 80 is not unusual for a 10 inch blade). The high number of teeth means there's less space between teeth for material removal. These blades cut more slowly but more cleanly.

Combination blades - To strike a balance between ripping and crosscutting, a combination blade typically sports 40 to 50 teeth. As expected, it is slower than a ripping blade while ripping, and rougher on crosscuts than a crosscutting blade. If your budget only allows for a single high quality blade, a carbide tipped combination blade would be a good investment.

Other specialty blades - The blades I've mentioned above are the more common ones. For special applications, there are blades for cutting all types of materials including Corian ® , plexiglass, non-ferrous metals, plastics, laminates, decking, plywood, etc.

Carbide - Standard blades are made from steel. These will dull very quickly especially when cutting composite materials such as particle board or medium density fiberboard. Carbide blades are a worthwhile, if not a necessary investment.

Speaker Building Uses

The better the blade, the better the cut. Dado blades are very useful for milling rabbets or grooves. A shelf brace inside a speaker or stand could easily be made to fit snugly by milling a dado on all side panels in which the shelf sits.

Add-Ons

Blade stabilizers - The typical table saw clamps the blade on either side of the blade, near the center spindle. A blade stabilizer consists of two round metal plates that clamp the blade further away from the center of the blade thus reducing blade chatter. There are three major side affects to be aware of when using such a stabilizer:

1. The stabilizer moves the blade away from its regular position relative to the fence. This means all distance markings are now wrong.

2. The slot in the table saw insert may no longer match the blade position. A custom insert can be made from some hardwood much the same way a zero clearance insert is made.

Page 37: Loudspeaker Builder

3. The maximum blade height is reduced.

Dado blade shims - Shims may be plastic or metal, and come in various thicknesses. They are inserted between dado head chippers to fine tune the cutting width.

Usage Tips

Follow regular maintenance on your blades as you should with all your tools. Remove any burned-on residue and never use a dull or broken blade.

Because dado blades remove much more material than conventional blades, feed your stock more slowly. As with a router, make multiple passes, removing a little more each time rather than trying to remove too much material at once.

Cost

Professional high performance specialty blade : can be over $100 · Professional high performance 10 inch blades : $40 to $60 Typical 10 inch blades : under $40 High quality 8 inch stack dado blades : under $150

Other Saws

Before we continue through other essential tools, let's take a brief side trip to visit other saws. Depending on your interest, skill, and resources, some of these saws might serve you very well in many uses. Needless to say, this list could easily get as large as the rest of this article, so I am only listing a few select items:

Radial arm saw - The radial arm saw is a very useful and flexible saw. It excels at crosscutting stock at various angles. It can also be used to rip stock by rotating the head 90 degrees. For an only saw in a speaker builder's shop, this is not the tool to get.

Chop or miter saw - I am actually referring to several types of saws here. The simplest miter saw provide a simple chop action and is usually capable of providing a miter in the range of +/- 45 degrees to the left and right by rotating the saw head relative to the saw fence. The more sophisticated compound miter saw allows the head to tilt, thus cutting two angles at once. For larger cutting capacity, the sliding compound miter saw has a rail system on which the saw head rides. This allows the head to extend forward, increasing the cutting capacity that would otherwise be limited by the blade diameter. This latter miter performs many of the functions of the radial arm saw and can be quite costly.

Band Saw - Some woodworkers consider this saw to be the most useful of all saws. The band saw can rip, crosscut, and re-saw. It can cut straight lines as well as curves. Best of all, it is very safe since kickback is non-existent. One thing it cannot do is mill dadoes. For the speaker builder, this is not the first saw to get, but for the woodworking junkie, it is a highly desirable and indispensable tool.

Page 38: Loudspeaker Builder

Hack Saw - The interchangeable blade make this the perfect low cost tool for cutting all kinds of materials.

Soldering Iron

While this article is about woodworking tools for the DIYer, the soldering iron is a necessary tool when considering soldering your own crossovers and performing mods to your existing equipment.

Features and Functions

Grip style - soldering irons are either pencil shaped or pistol shaped. For control, I personally prefer the pencil shape, but this is obviously a matter of personal taste and preference.

Power (watts) - hobby soldering irons are sold for many purposes. They range from low power devices such as those for electronics to high power units for use with stained glass. It's important to use an iron of the correct wattage to avoid excessive or insufficient heat. For speaker building, a small iron suitable for soldering crossovers and other electronic parts would be in the 30-watt range. Some irons have selectable temperature settings, which, while not necessary, might be of use to the more experienced user.

Tips - Irons typically have replaceable tips. Tips are either steel plated or copper. Irons featuring steel tips cost more than copper tipped irons, but are well worth the extra expense. Copper tips degrade and lose their shape quickly making them almost useless for fine work. The cost of standard fixed-wattage steel tipped irons is low enough to make them worth the investment even to the infrequent user.

Speaker Building Uses

The soldering iron is used in building crossovers and other electrical assemblies.

Add-Ons

Stand - avoid burning your bench top by getting a decent stand to hold the hot iron.

Sponge - wipe the iron's tip with a wet sponge prior to each use. A sponge made for soldering irons is preferred over a household sponge.

Clip-on heat sink - use these between the soldering area and any heat-sensitive components to prevent heat damage.

De-soldering tool - vacuum tools are popular. Solder wicks also work well at absorbing and removing molten solder.

Extra tips - these come in different shapes and sizes. Select the size and shape depending on your use.

Page 39: Loudspeaker Builder

Usage Tips

Be sure to use the right type of solder for your work. Check the solder's label before using. Never use acid core solder for electronic work.

· Always make sure the tip has reached its operating temperature. Check by touching some solder to see if it melts quickly.

Make sure the physical connection is secure before the solder is applied. The solder should not be used as a mechanical means by which the joint is held.

I like to get the iron, solder and parts together at the same time and place. (By "part", I mean wire, lead, pad or anything that requires soldering). This instant merging will melt the solder onto the components at the right spot. Removing the solder and iron quickly reduces the chance of excessive heat buildup which can damage sensitive parts. In essence, my goal is to use the iron to melt the solder onto the parts. I have seen some directions that instruct the user to heat the target parts first, and then melt the solder onto the part using the part to heat the solder. While this method is preferred for high heat applications such as stained glass and plumbing, it can easily overheat sensitive electronic parts.

The secret to great soldering - practice, practice, practice!

Cost

Professional high quality soldering iron station: upwards of several hundred dollars

· High quality 30 watt iron plated pencil iron with simple stand : under $50 Consumer grade 30 watt soldering iron : under $20

Other Tools

Here's a small list of other tools, both large and small, some essential, others optional. Much can and has been written about each, and readers are encouraged to further research tools that interest them:

Dust collector - This is a stationary item specifically designed to move large amounts of air. They are usually powered by an induction motor and are best for collecting dust from table saws, jointers, and other large floor standing power tools.

Shop Vacuum - A shop vacuum is a portable device often capable of wet and dry use. It uses a small, noisy, high-speed motor and does not have the volume capacity of a dust collector. However, it is readily available at home centers and department stores, and the typical hobby shop can benefit greatly from this tool for both cleanup and dust collection.

Clamps - No matter how many you might have, you won't have enough! Pipe clamps are best suited for large projects. Bar clamps exist in both heavy duty (similar to pipe clamps) as well as light duty. Lengths vary from about 12 inches to several feet. C-clamps are extremely handy and secure, but can be a hassle to use. One-handed clamps are much friendlier. These usually have some sort of

Page 40: Loudspeaker Builder

squeezable handle that quickly move the jaws onto the stock. Band clamps are straps that wrap around an item to hold everything together. Right angle clamps hold parts together at right angles, allowing fasteners to be inserted. This is a very handy device to use on boxes.

Glues - regular (yellow) woodworking glue is the most common for use on wood and wood products. Gaining popularity as an all-purpose adhesive is polyurethane glue. This glue expands as it dries thus requiring some care in handling. Stock to be glued should be securely held to avoid being moved by the glue's expansion. Epoxy is always a favorite for gluing just about anything. And then there are the construction adhesives. Their thick consistency makes them suitable for filling gaps. In all cases, avoid glue fumes, and don't expose components such as drivers to these fumes to avoid chemical reaction damage.

Fasteners - Particle board and drywall screws are popular for holding composite materials together. T-nuts are often used to fasten drivers to baffles, though I've had no problems with drywall screws. Pre-drill and countersink to avoid tearing the stock when driving in screws. Avoid placing fasteners too close to the edge of wood products to avoid splitting the material.

Framing square - Most basic boxes consist of right angles, and a large framing square is the simplest way to get this with a decent amount of accuracy.

Buying Tools

Here are some beginner's tips to buying tools:

Before buying any particular class of tool (router, drill, etc), make sure you need this tool. Budget your purchase and return on investment with other possible tools that may perform the same function. Do a little research on each tool in your price range. This article should provide you with some basic points of interest, but ultimately, each model will have its own features and functions to consider.

Consider your working environment - Do you have neighbors that might object to the noise? Do you have allergies that might be exasperated by wood dust?

Are there outside sources of tool time such as those in a community center or adult education program?

What other uses for each tool might you have aside from your hobby?

Like all things, it often makes good financial sense to buy things on sale. Aside from the year-end holidays, tool sales can often be found around Father's Day. Many regional or national chains also have frequent sales. Mail order suppliers are common in this business, and many offer very competitive prices. Be sure to check their return policies and shipping charges. Buying from a local dealer may appeal to those looking for more personal service. Many mail order and retail outlets have web sites and advertise in leading woodworking publications.

Other factors that often affect a purchase include brand loyalty, product quality, and availability of replacement parts and accessories. Do some research on the

Page 41: Loudspeaker Builder

manufacturer, their service department, dealers, and repair history if these items are important to you.

Recently, refurbished tools have become increasingly popular. These tools are usually sold by the manufacturer through outlet stores or through dealer special purchases. Such tools usually have a manufacturers' warranty, but be sure to check the warranty duration.

Internet Resources and Links

Here are some URLs relating to tools:

(Disclaimer: this information is provided as a service to the reader, and does not represent an endorsement of any kind of the companies represented by the URLs.)

Tool manufacturers

Aeg Bosch Delta International DeWalt Fein Jet Makita Powermatic Porter Cable Ryobi Skil

Mail order tool sources

Tool Crib of the North Trendlines Woodworkers Warehouse Constantine's Woodcraft The Woodworkers' Store Highland Hardware Garrett Wade Lee Valley

Publications

Taunton Press

Closing Thoughts

Page 42: Loudspeaker Builder

I've tried to keep the focus of this article in the speaker building domain. The experienced hobbyist will no doubt notice the vast amount of missing information. It's possible to write an entire book on each tool, and many such books already exist. Readers wanting to know more are encouraged to visit their local library, bookstore, or Internet sites.

As a woodworker, I feel obligated to mention that power tools are not absolutely necessary for speaker building (or most other DIY projects). Craftsmen with hand tools have practiced the art of fine woodworking long before the advent of power tools. Some of the finest furniture made today is still made with hand tools. Hand tools are capable of performing the same cutting, shaping and drilling operations with no electrons present. And like all tools, the only limitation is one's experience, patience and imagination.

Hand tools offer some significant advantages over their powered counterparts. Outfitting a beginner's shop is usually cheaper than with power tools. Other benefits are the lower dust levels, the absence of loud motors, the smaller floor and bench space for tools, and more "feel" for the material. While this article was not intended to cover hand tools, many resources exist for those interested in learning more. Readers are encouraged to consider hand tools in their woodworking craft.

Louis Lung ([email protected])

B1 Buffer Preamp

Nelson Pass

Introduction

Side A

So here we are in the New Millennium, and thanks to Tom Holman and THX we’ve got lots of gain in our electronics. More gain than some of us need or want. At least 10 db more.

Think of it this way: If you are running your volume control down around 9 o’clock, you are actually throwing away signal level so that a subsequent gain stage can make it back up.

Routinely DIYers opt to make themselves a “passive preamp” - just an input selector and a volume control.

What could be better? Hardly any noise or distortion added by these simple passive parts. No feedback, no worrying about what type of capacitors – just musical perfection.

Page 43: Loudspeaker Builder

And yet there are guys out there who don’t care for the result. “It sucks the life out of the music”, is a commonly heard refrain (really - I’m being serious here!). Maybe they are reacting psychologically to the need to turn the volume control up compared to an active preamp.

I suppose if I had to floor the accelerator to drive 55 mph, maybe I’d think the life was being sucked out of my driving. Then again, maybe I like 55. Nice and safe, good gas mileage…

Is impedance matching an issue? Passive volume controls do have to make a trade-off between input impedance and output impedance. If the input impedance is high, making the input to the volume control easy for the source to drive, then the output impedance is also high, possibly creating difficulty with the input impedance of the power amplifier. And vice versa: If your amplifier prefers low source impedance, then your signal source might have to look at low impedance in the volume control.

This suggests the possibility of using a high quality buffer in conjunction with a volume control. A buffer is still an active circuit using tubes or

transistors, but it has no voltage gain – it only interposes itself to make a low impedance into a high impedance, or vice versa.

If you put a buffer in front of a volume control, the control’s low impedance looks like high impedance. If you put a buffer after a volume control, it makes the output impedance much lower. You can put buffers before and after a volume control if you want.

The thing here is to try to make a buffer that is very neutral. Given the simple task, it’s pretty easy to construct simple buffers with very low distortion and noise and very wide bandwidth, all without negative feedback.

There are lots of different possibilities for buffers, but we are going to pick my favorite:

Side B

Figure 1 shows the full schematic of the B1 buffered passive preamp.

There are two channels shown with a common power supply. Supply parts in common are numbered from 1 to 99. Parts in the right channel are 100 to 199, and the left channel is 200 to 299.

With the exception of R1, all the resistors are ¼ watt – I used RN55D types, but you can use whatever you like. C1 and C2 are big electrolytic types, with a nominal rating of 15000 uF at 25 Volts. C100, 200, 101 and 201 are high quality film capacitors. You can use these values or substitute in your favorites.

Page 44: Loudspeaker Builder

For C1 and C2 I used Digikey P6890-ND. The value is not critical, and you can use as low as 1,000 uF at 25V. C3, C100 AND C200 are 1 uF metallized polypropylene film capacitors (Digikey BC2076-ND) C101 and C201 are 10 uF polyester film capacitors (I used Axon 10 uF metallized polypropylene from Orca Design). Feel free to use any comparable types.

D1 is a generic 1N914 type diode, and D2 is a generic LED for indicating power the board.

All of the transistors are N channel JFETs. The stock parts are 2SK170’s, LSK170’s or 2SK370’s, and you can use substitutes having Idss between than 5 and 10 milli-Amps and transconductance numbers from 5 to 30 milli- Siemens.

The potentiometers are linear taper at 25 Kohm, but again you can easily use higher or lower values as you like.

The buffer uses an external power supply from 18 to 24 Volts DC. You can power it with batteries, but most convenient is an external regulated supply running off the wall. The preamp typically draws fewer that 0.02 Amps, so current is not much of an issue. A regulated supply is better, but the circuit is pretty good at ignoring noise on the supply and minor fluctuations.

The design uses RCA input and output connectors, and a DPDT switch for selecting one of two inputs. You are, of course, free to use a switch with many more inputs.

For a fact this circuit can be easily constructed with perf-board and point-topoint wiring. I know, however, that many people won’t start a project like this without a circuit board. The Gerber artwork is posted at www.passdiy.com and I have arranged to have finished PC boards available at cost around the time you read this.

Side C

Let’s talk about what some of these parts do.

The input switch selects one of the two input signals, routing it to the top (clockwise position) of potentiometers P100 or P200. A divided input signal appears on the wiper. This signal goes to the Gates of Q100 and Q200 through a resistor R102, R202 and capacitor C100, C200. As a practical matter, the input impedance of this preamp is determined by this volume control potentiometer. A 25 K Ohm pot gives 25 K Ohm input impedance.

R102 and R202 are there to prevent parasitic oscillation with the very wide bandwidth JFETs. C100 and C200 are there because the Gate of the JFETs needs to be set at ½ the DC voltage of the power supply – a voltage delivered to the buffer inputs by R2, R3, and C2 through R103 or R203.

D1 performs the service of drawing down this DC voltage with the power supply when the power is turned off, otherwise C2 may hold a charge for a long enough time to give you a turn-on thump when powered back on. By the way, the time constant of R2, R3, and C2 are long

Page 45: Loudspeaker Builder

enough that it takes a minute or two for the circuit to reach normal operating values, so don’t get excited if there’s no sound for a few seconds when you turn it on.

R1 and C1 filter noise coming from the external supply.

Q100 and Q200 are JFETs operated as follower transistors. The Source pins of these transistors follow the voltage at the Gates. The input impedance of the Gate is exceedingly high – many millions of ohms, and the output impedance at the Source pin is about 50 ohms.

Q101 and Q201 are constant current sources formed by simply attaching the Gate pins of the JFETs to the Drain pins. They provide without loading them down or creating significant distortion.

The best performance generally comes from matching the Idss of Q100 and Q101, and also Q200 and Q201. The Idss is simply the current that flows through the JFET when the Gate and Source are grounded and +10 volts or so is applied to the Drain. Often when you buy JFETs you can get them in Idss grades. I use GR or BL grades for this project.

The Source DC voltage of the JFETs Q100 and Q200 is about the same as the Gate DC voltage (1/2 the supply voltage), and the output from the Source needs to have the DC voltage removed by C101 or C201, leaving the AC output signal, which passes through another safety resistor R104 or R204.

Resistors R100, R200, R105, R205 are there to prevent the various potential thumps from switching inputs and turn-on transients.

Side D

The circuit of Figure 1 has quite good performance specifications.

Figure 2 shows the distortion at 1KHz. Below a volt, the distortion comes in at about .0007%, and about the time your amp is clipping, it measures about .003%

Figure 3 shows the .0007% distortion waveform at 1 Volt and 1 Khz.

Figure 4 shows the distortion vs frequency. It does not alter over the audio band.

Page 46: Loudspeaker Builder

Figure 5 shows a 100 KHz square wave at 1 volt at the output. The bandwidth of the preamp is -3dB at about 700 KHz. On the bottom end, the low frequency roll-off is about 1.5 Hz into a 10 Kohm load and about 0.3 Hz into 47 Kohm.

Side E

So how does it sound? Well, no suckage here. I’ve noticed that simple nofeedback circuits have tremendous clarity if the circuit has wide bandwidth and really low distortion.

At the moment I’m driving a pair of Lowther DX55’s with some passive baffle-step correction (6 dB loss there) and an F3 amp with only 12.5 dB voltage gain. The preamp is fed by an Xono phono stage with a low output cartridge (Grado Statement).

Page 47: Loudspeaker Builder

The sound is extremely neutral without being clinical - just about what you were looking for when you were thinking about a passive preamp. A teensy bit of second harmonic and no noise at all.

There’s just enough gain. If you were using any other power amp, you’d get 8 to 18 dB more gain, and would be able to break your lease or the speaker, or both.

Do I feel like the pedal’s to the metal and I’m only doing 55?

No - I’m listening as loud as I want and I sleep soundly at night, knowing that I’m not throwing away signal with my volume control.

Side F

Have you noticed that they’re putting out some LP’s with not just 2, not just 4, but with 6 sides holding the content of a CD?

It’s Sweeeet. Two cuts to a side, and when you look at the grooves you can see the land area between them. Right now I’m listening to “High Fidelity Lounge”, vol 4, Side F.

Bride of Zen

Nelson Pass

Introduction

This is the second installment of a trilogy of construction projects centered around the performance obtainable from absolutely minimalist circuitry. Part one described the Zen amplifier, a 10 watt single-ended class A power amplifier using a single MOSFET gain stage. In this piece we will examine its preamplifying mate, also a single gain stage MOSFET circuit.

A lot of what I might say about this design has already been expressed in the Zen amplifier piece, and will not be repeated here. It is enough to say that I continue to be very intrigued by the sound of very simple topologies, single-ended in particular, and the use of MOSFET gain devices in realizing them. I assert that simpler circuits sound better. Anyone who disagrees is welcome to that opinion, and I wish them well listening to op amps.

As with the Zen amplifier, the purpose here is to explore a design which pushes simplicity to the limit. I want you to build it and experience not only the satisfaction of creating something tangible and functional, but also the pleasure of subjectively high performance audio.

As before, the simplest possible gain circuit has one stage, with a single transistor. Of necessity for linear operation, it must be operated in single-ended class A, and because we want both a high input impedance and voltage gain, the only useful connection is common-source, where the input signal is presented to the Gate of the MOSFET, and the output signal appears at the Drain.

Page 48: Loudspeaker Builder

Quite a few people have told me that they understood the concept of single-ended Triode amplifiers, and had not imagined that such a thing could be done with transistors. Conceptually, you can do almost anything with transistors that you could with tubes, although you do not always get the same results. The converse will not be true until a reliable supply of positrons becomes available to run through a P channel tube.

I did not find it practical to build a single gain stage tube amplifier, but the relaxed conditions under which a preamp functions makes a tube version of this circuit workable. Simply adjust the voltage and bias values for a tube instead of the MOSFET we will use here. I leave that as an exercise for the reader.

Requirements

Due to its extreme simplicity, the Zen amplifier’s interfaced needs to be indulged a bit, both at the output and also at the input. With respect to its output, it is designed for an 8 ohm load with an efficiency of 94 dB or greater. I have driven the Zen into Thiel 3.6 loudspeakers, 88 dB efficient at 2.3 ohms, but the result is not pretty. With a pair of Klipsch speakers it will throw off its mild mannered disguise and behave like a party animal.

The Zen amp input makes a preamp do a little more work than usual. Depending on the specific input values, its input impedance is typically 4.75 KOhm, and it needs about 2.5 volts input to drive it to full output. Its gain is about 15 dB, which is 10 dB less than most power amps and it requires more signal from the preamp for equivalent levels.

The point of a preamp circuit here is to function as a gain circuit with a high input impedance, a low output impedance, and a volume control. This preamp has a 50 KOhm input impedance, variable gain to about 15 dB, and will deliver the 2.5 volts at .1% distortion or less.

In the Zen amplifier, the need to linearly swing a relatively large amount of current into a load of varying impedance gave us a good reason to employ negative feedback around the MOSFET. Without it, the damping factor would be extremely low, and any impedance variation in the loudspeaker would give a large variation in frequency response.

The Zen preamp circuit has a much easier job in this regard, as the load attached to the Drain of the MOSFET will be resistive over the audio frequency range, allowing for intrinsically flat response. We will have much less reason to employ feedback in this circuit, and so we will not.

Another advantage that the preamp circuit will enjoy over the power amplifier is the relatively small dissipation involved. The Zen single-ended power amplifier idles at more than four times its output rating, typically 50 to 60 watts. The Bride of Zen will not typically be asked to deliver more than a milliwatt, but the MOSFET will have no difficulty dissipating a watt or so. This will allow us to bias the transistor at very much higher voltages and currents than are required to drive the load, enormously enhancing the performance.

A single common-source gain device will invert the phase of the signal it amplifies, and this must be taken into account when using the circuit in an audio system. The Zen amplifier also

Page 49: Loudspeaker Builder

does this, but we reversed the phase on its output terminals to compensate, which is not possible here. Using the Zen preamp usually requires that we adjust the phase of connection between amplifier and loudspeaker for phase integrity.

The importance of proper absolute phase has been discussed elsewhere, including the original Zen amplifier contribution. I will simply say that proper absolute phase is a legitimate factor in the overall quality of sound reproduction. Even if I could not hear phase reversal, aesthetic considerations alone would prompt me to assure that it was correct.

Volume Control

We will place the volume control at the output of the circuit, giving us a higher output impedance than is typical. Depending on the position of the volume control, the output impedance will range from 0 to 1.5 KOhms.

Normally you see the volume control potentiometer at the input to the preamp circuit, and in this way you can insure that the input will not be overloaded. Placing the potentiometer at the output has two advantages, however, the first being that all circuit noise will be attenuated as the volume setting is reduced. Placing the volume control at the output allows the input of the gain stage to look directly at the source. This improves the bandwidth and distortion of the circuit by providing a lower RC figure for the input capacitance of the MOSFET. If you desire to place a potentiometer at the input of the circuit, feel free to do so before the input coupling capacitor in place of R107.

Because the input to the Zen amplifier is a resistor operating into a virtual ground summing junction, it is insensitive to a resistive source impedance. The gain of the amplifier will alter by the addition of the source resistance, but the distortion and bandwidth performance will not change appreciably. A circuit driving the Zen amplifier need not have a particularly low output impedance, as long as it is resistive in character.

Is a maximum 1.5 KOhm source impedance a problem? Not in my experience, as long as it is passive in character, as in the case of this circuit which has no feedback. You could conceivably see some rolloff at the top end, but even 1000 pF loading will only take the bandwidth down to 100 KHz.

The Schematic

Figure 1 shows the schematic of the one channel of the Zen preamp. There is one transistor, Q101, and N channel power MOSFET. Input signal is presented to the gate of the MOSFET and the output is taken from the Drain. Everything else in the circuit is used to bias up the MOSFET appropriately.

Page 50: Loudspeaker Builder

While not necessarily true of all MOSFET devices, the types we will be considering perform better at higher supply voltages and higher currents. To take advantage of this, we will use a relatively high voltage supply at 60 volts and will bias the gain device at 40 milliamps.

We will provide approximately +7.5 DC volts to the gate of the MOSFET in order to turn it on. The MOSFET will require about 3.5 volts Gate to Source voltage to conduct the 40 milliamps, and this will leave 4 volts left over to be placed across the 100 ohm resistor, R108, which gives us 40 milliamps of current through Q101.

The DC input bias is delivered by the components R101, P102, R103, R105, and C102. The positive supply voltage is divided down by R101 and P102, and can be adjusted by P102. R103 and C102 provide noise filtering, and the adjusted quiet DC voltage is connected to the input through R105. To keep the source impedance from altering the DC bias, we capacitively couple the input through C104, which is a 1 uF film capacitor. The RC network of C104 and R105 produces a high pass characteristic at the input which is 1.6 Hz.

Page 51: Loudspeaker Builder

R107 is an optional part which is used to prevent transients when the input selector is switched. R106 is a 100 ohm nominal resistor used to prevent parasitic oscillations in the MOSFET. D101 is a zener diode which serves to protect the Gate of the MOSFET from voltages in excess of the 20 volt Gate-to-Source rating.

Signal input at the gate of the MOSFET causes the current through the transistor to vary, producing output AC voltage across R104. The gain of the circuit is the ratio of the output loading divided by the apparent resistance formed by the 100 ohm R108 and the inverse of the transconductance of the MOSFET. At this current with typical devices, that value is about 110 ohms total. At audio frequencies the output load is the 1000 ohms of R104 in parallel with the combination of output potentiometer P101 (the output level control) and the load. This value is always less than 800 ohms, so the gain of the circuit is somewhat less than 8 without an external load.

The 40 milliamps of DC current flowing through Q101 sets up a drop of 40 volts across R104, leaving about 20 volts DC at the Drain of Q101. Since we do not want to pass this along to the amplifier, we use output coupling capacitor C103, which rolls off the output below 3 Hz without an output load.

The circuit will use a regulated power supply, but since any noise on the supply will appear at the output through R104, it is desirable to clean in up as much as possible. R102 and C101 form a passive filter at the supply line to perform this task.

Referencing Fig. 1, note that the component numbers of one channel begin with 100, and the components of the other channel begin with 200. The power supply will have numbers which begin with 1.

So there we are; one transistor, one diode, 4 capacitors, 8 resistors, and 2 pots.

The Power supply

Page 52: Loudspeaker Builder

Figure 2 shows the schematic for a power supply with adequate capacity for two channels. The transformer voltage required to produce the regulated 60 volts will be about 50 volts AC. I used an Avel-Lindberg D4007, which is 60 volts with secondaries in series, producing about 86 volts DC across power supply C1. A stack of Zener diodes is used to produce a 64 volt reference biased by R1 and filtered by C2. This stack drives follower Q1.

Additional passive RC filtering is provided. This filtering, along with the filtering provided on each channel, is essential, because the regulation provided by the zener diodes and Q1 has good DC stability but still carries quite a bit of AC noise. As supply noise will appear at the output, we want the power supply noise to be greater than -100 dB below 1 volt, or less than 10 microvolts.

The Parts

In a simple circuit, there are no small roles, only small parts. With the exception of R104 and R204, I have used RN55D type metal film resistors. You are welcome to use anything you choose.

The capacitors in the signal path are metallized polyester film types. “Better” capacitors can be had; maybe they sound better, maybe they don’t. As with the resistors, you may do as your please.

Page 53: Loudspeaker Builder

The choice of MOSFETS is slightly more restricted. The devices should be capable of withstanding 50 volts minimum, and 100 volts is preferred. They will need to dissipate approximately 1 watt, and should be equipped with the appropriate heat sink for this.

The IRF510, 520, 610, and 620 are typical parts that are suitable. Higher voltage types are fine, but offer somewhat less transconductance. Higher current types have more capacitance, and suffer from higher distortion at high frequencies due to nonlinearities in this capacitance. The transistor actually used was the International Rectifier IRF610.

Table 1 displays the parts list for two channels and the power supply.

Table 1 displays the parts list for two channels and the power supply.

Resistors

R1

4.75k

R2, 3

47.5

47.5

33.2k

R102, 202

22.1

R103, 203

10k

R104, 204

1k, 3W

R106, 206, 108, 208

100

R105, 205, 107, 207

100k

Page 54: Loudspeaker Builder

Capacitors

C1, 3, 4, 101, 102

1000uF,100V

C2, 102, 202

100uF, 63V

C103, 203

10uF, 100V

C104, 204

1uF, 100V

Potentiometers

P101, 201

5k

P102, 202

10k

Miscellaneous

Q1

TIP29CGE

Q101, 201

IRF610 Mosfet

D1-4

1N4004

Z1-6

Page 55: Loudspeaker Builder

10V Zener

D101, 201

20V Zener

T1

60V AC Avel-Lindberg D4007

Heatsinks

TO-220 transistors

The PC Board

Fig 3 and 4 are the circuit board art and component placement diagram. A simple circuit such as this can easily be constructed on perf board or even with point-to-point wiring, so feel free to do it that way. There is nothing exotic about the layout, and no special requirements to avoid oscillation and such.

Page 56: Loudspeaker Builder
Page 57: Loudspeaker Builder

Construction Notes

MOSFETs are sensitive to static electricity. Their gate resistance is unbelievably high, but will break down somewhere between 20 and 100 volts, damaging or destroying the transistor. For this reason, I have provided a protection diode in the circuit. Once the circuit is assembled, the transistor will be safe, but until it is, the MOSFET should be handled carefully to avoid exposure to static electricity.

If you can locate a dual volume control potentiometer at 5 KOhms, you will want to use it. Higher values are OK, but result in higher source impedance. If you use short cables between the preamp and a Zen, you should be able to get by with up to 25 KOhms, which will give a

Page 58: Loudspeaker Builder

maximum source impedance of 6 K ohms. If you can’t find a suitable dual pot, simply use separate level controls for each channel.

The input select system consists of input connectors and a 2 pole rotary switch for a selector. It is preferable to shield the cables between the inputs and the switch, and between the switch wiper and the circuit input. If you want to implement a tape output, you can do so by tapping off the wiper of the selector .

It is useful to put some distance between the transformer and the audio circuitry in order to avoid noise caused by the stray magnetic fields of the transformer and primary and secondary wiring.

I leave the AC power hookup to you. I recommend using an approved commercially available line filter / input receptacle on the AC line, with a .125 amp slow blow fuse.

We will want to ground the chassis to circuit ground, and for safety it is desirable to also earth ground the system through the third conductor of the AC line cord. This occasionally results in ground loop noise, and if it does, I recommend grounding the system to earth through a 10 ohm 5 watt resistor as an alternative to simply floating the chassis. If there is any question about safety in AC power wiring, it is important to consult a qualified technician.

Objective Performance

Fig 5 shows the distortion characteristic of the circuit from .1 to 10 volts output. The distortion is fairly pure second harmonic, generally regarded as the least offensive variety. The distortion curve is monotonic, that is to say, it increases smoothly as the voltage increases, and its first derivative is positive. At the bottom of the curve we begin to see some noise, which is the increase in the curve below .3 volts.

Fig 6 shows the distortion versus frequency of the circuit at 1 volt output. It is flat.

Fig 7 shows the frequency response of the circuit. The bottom end rolloff reflects the output and input coupling capacitors. To extend performance to lower frequencies, use larger capacitors. At

Page 59: Loudspeaker Builder

the high end, the circuit is down 1.5 dB at 200 KHz, and part of this is the test setup. The high frequency rolloff is a function of the capacitance of the MOSFET.

Subjective Performance / Conclusion

It is in this spot that the author elucidates the sonic astonishment which awaits the amateur upon completion of the project. I will not disappoint you. It sounds marvelous. Really. No Kidding.

Even if you didn’t build the Zen amplifier, you owe it to yourself to do this one. Stay tuned for Part 3, The Son of Zen.

White papers and literature still available from Pass Labs, PO BOX 219, Foresthill, CA 95631. tel (530) 367 3690 fax (530) 367 2193

Copyright 1994 Nelson Pass

Balanced Zen Line Stage

Nelson Pass

Introduction

The popularity of the Zen projects points out the interest in very simple linear circuits. They are intended to fuel that interest. The Zen, Bride of Zen, and Son of Zen have been explorations in how much objective and subjective performance can be achieved with a single gain stage. This extreme simplicity has an aesthetic appeal, which speaks to the purist in audiophiles, and the presumption that simple circuits sound better.

At least one "objectivist" has complained (objected?) that the Zen projects do not measure up compared with more sophisticated and complex amplifiers. This is mostly true, but beside the point. The literature and store shelves are full of multi-stage amplifier circuits using generous amounts of negative feedback.

These are single stage amplifiers.

Getting good performance from a single gain stage is a fine technical challenge, and I would say a proper beginning for those who would go on to design and build more complex circuits. Simple circuits have particular value as DIY projects. They are more understandable, they are more likely to be attempted, and they are more likely to work.

So is this the anticipated Bride of the Son of Zen? I suppose it is. It has an identical topology and is perfectly suited for driving the Son of Zen, but there is much utility to this circuit. It also serves as a nice balancedin to unbalanced-out or unbalanced-in to balanced-out converter.

Like the Bride of Zen and the Son of Zen projects, this circuit performs linear amplification without negative feedback. A lot has been written pro and con about the use of negative

Page 60: Loudspeaker Builder

feedback, and I don't propose to bring the debate into this article. It just so happens that we will be getting quite good performance without it.

Balanced Operation

Let's review why balanced operation is desirable. Audio circuits operate in an environment of electrical noise; crosstalk from other channels, ground loops, magnetic pickup from transformers, power supply ripple and other noises. In a balanced circuit, two opposite phases of the signal are present on two otherwise identical input lines. The input of a balanced circuit has a plus and minus polarity, and the output of the circuit also has a plus and minus. The balanced amplifying circuit will amplify the difference between the two inputs and will display a larger difference signal at the output.

What the circuit doesn't do is as important as what it does; it does not amplify any portion of the signal which is the same at both inputs. Ideally it completely rejects the common input signal, and the quality of this rejection is referred to as the Common Mode Rejection Ratio (CMRR), which tells how much of the common input signal gets through.

What the circuit doesn't do is as important as what it does; it does not amplify any portion of the signal which is the same at both inputs. Ideally it completely rejects the common input signal, and the quality of this rejection is referred to as the Common Mode Rejection Ratio (CMRR), which tells how much of the common input signal gets through.

Figure 1 shows the circuit for one channel. Q1 and Q2 are the active elements of the gain stage. They are mutually coupled through R15, and the current through them is controlled by the differences in their Gate voltages. The Gates are nominally at ground through R13 and R14, and the transistors are biased through resistors R3, R4, R5, and R6 connected to a negative voltage supply. Balanced output signals appear on the Drain pins of Q1 and Q2, loaded by R1 and R2 connected to the positive supply.

At the output Drains of Q1 and Q2 we will see a DC potential of approximately one-half of the positive supply, and the two voltages will vary in opposite phase. The maximum peak-to-peak voltage at each Drain is the value of the positive supply, and twice that amount when considered as a balanced output. As a practical matter these two paraphase outputs are passed to the outside world through capacitors which block the DC voltage.

The circuit amplifies a differential input to a differential output. Given a single input, it will amplify it into a balanced differential output, and is useful to convert unbalanced signal to balanced, either because you wish balanced operation or because you wish to convert a conventional stereo amplifier into mono bridged operation.

Of course the circuit is happy accepting a balanced input, and besides using it as simply a balanced gain stage, you can choose to view either output polarity as a single-ended output of each phase. Thus the circuit also functions as a general balanced and unbalanced converter.

Page 61: Loudspeaker Builder

Being that the noise picked up from the environment is usually common to both input lines, it is rejected at the input of the balanced circuit, and thus is much less of a problem. Actual home audio systems using balanced interconnects typically have about 1/10 the background noise and hum.

Another reason to use balanced preamplifying gain stages is that many high end DAC designs offer balanced outputs in which separate DAC circuits are used for each of the two phases of output. Using separate balanced DAC circuits reduces the random noise by 3 dB, the same as if they were in parallel, and reduces common noise by a larger figure. There is also the potential for reduction of distortion with such an approach, but to realize the full performance of these circuits, the gain stage following must have a balanced input.

The Design

This balanced line stage is the classic "differential pair" topology with two identical gain devices connected to develop voltage gain from a differential voltage input. The gain devices (tube/bipolar/fet) couple to each other through their Cathode/Source/Emitter pins. The voltage input is presented to the Grid/Gate/Base pins, varying the current through the devices, and showing up as balanced output voltages on the Plate/Drain/Collector pins.

By the way, unfortunately the designation of Source pin for a MOSFET is often confused with the use of the word "source" which occurs frequently in reference to audio. For clarity here, I will always capitalize the "S" when referring to the pin.

 

Page 62: Loudspeaker Builder

Referencing Figure 1, we note the function of various components. Resistors R7, R8, R9, and R10 serve to slightly isolate the inputs and outputs from the transistor pins, preventing parasitic and other types of oscillation. The values for these resistors can range from 100 to 475 ohms, and 221 ohms is about the best value.

Resistors R13 and R14 serve to provide a reference to ground for the input pins in case there is no source or in case the source is AC coupled. This keeps the Gate voltages near ground so that the amplifier biases with the proper DC voltages and currents. Similarly, R11 and R12 serve to bleed off the DC voltage which initially appears at the output through the blocking capacitors C1 and C2.

R16, R17 and C3, C4 are used to passively filter the supply voltages. The supplies will be regulated, but the filter will remove stray and residual noise at the output of the regulators. The circuit itself will reject supply noise through balanced operation, but this does not help when using a single phase of the output. Passive supply filtering improves both balanced and non-balanced operation.

Resistors R1 through R6 are all 750 ohm metal film power resistors rated at three watts each. All the other resistors in this project are ¼ watt metal film types. R1 and R2 serve as the loads for the transistors, and R3, R4, R5 and R6 are used to bias the circuit with a current source from the negative power supply. The reason I used two 750 ohm 3 watt resistors in series was to get a 1500 ohm 6 watt resistor with the same part as used for R1 and R2.

Page 63: Loudspeaker Builder

The gain of the circuit is the ratio of the impedances of the output circuit divided by the impedances of the circuit that couples Q1 and Q2. The summed impedance of the output circuit is essentially R1 plus R2, and is 1500 ohms. The impedance of the coupling circuit is 124 ohms summed with the approximately 12 ohm apparent source resistance of each of the MOSFETs, or about 150 ohms. The gain of 10 (20 dB) reflects the 1500 ohms divided by the 150 ohms. You can adjust the gain of the circuit arbitrarily by adjusting the value of R15 without affecting the quiescent DC values of the circuit. As you decrease the value of R15 to 0 ohms, the gain will approach 50 (34 dB). As you increase the value of R15, the gain decreases, with 430 ohms giving 10 dB of gain. Some of the performance curves presented later will reflect both 10 and 20 dB gain settings.

Figure 2 shows the power supply for the circuit. Because the voltages required by this project are higher than delivered by off-the-shelf transformers, I have chosen to use two transformers T101 and T102 with secondary circuits in series, one for the positive and one for the negative supply. The primary circuits are connected in parallel, in this case showing 120 VAC operation. The regulated supply voltages at the outputs of this power supply will be at 60 volts each, and to give

Page 64: Loudspeaker Builder

us an adequate margin from the unregulated supply, I have chosen Avel- Lindberg transformers model D4007, rated at 30 + 30 volts AC on their secondaries and 30 watts each. The AC secondary voltages are rectified through B101 and B102 to produce an unregulated plus and minus 80 volts DC.

The circuit is protected by a 1 amp slow blow fuse, and a high voltage (AC line rated) filter capacitor C107 is placed across the line to reduce noise. Earth ground from the AC cord is attached to the chassis for safety, and is connected to the circuit ground through a power thermistor (bright idea from Frank DeLuca). This gives some resistive isolation for prevention of ground loops but goes to small values of resistance in case of catastrophic connection to the live AC line.

The unregulated 80 volt rails are cleaned up by pass transistors Q101 and Q102 acting as voltage followers from stacks of Zener diode voltage references. The Zener diode stacks are 7 X 9.1 volts, or about 63.7 volts, fed a trickle of current by R101 and R102. Allowing for the approximately 3.7 volt loss from MOSFET Gate to Source pins, the output of Q101 and Q102 are 60 volts DC. Capacitors C103 and C104 across the Zener stacks reduce the noise, as do capacitors C105 and C106 at the output of the supply.

Zener diodes Z115 and Z116 are used to protect the Gates of the MOSFETs from exceeding the 20 volt limits imposed on the Gate-to- Source rating of the MOSFETs. As elsewhere, the Gates of the MOSFETs are isolated through resistors R103 and R104 to prevent parasitic oscillation.

It is possible to set up the power supply to regulate at lower voltages to match transformers with less voltage than offered by the D4007's specified. Alternately, you might choose a single transformer that does not develop as much voltage as two D4007's. Simply replace some of the Zener diodes with shorting wires to produce lesser reference voltages in increments of 9.1 Volts, or replace the 9.1 volt parts with other values. To give you a picture of performance at lower supply

Page 65: Loudspeaker Builder

 

packages, and substitutes should have a 100 volt rating. The IRF610 is rated at 200 volts, 20 watts, and 10 amps. An excellent substitute is the IRF510. Devices of higher current or wattage rating are not preferred, as the capacitance of the device is increased correspondingly, and results in degraded performance at high frequency, although I have been able to get acceptable performance out of 150 watt devices with 10 times the capacitance.

It is not essential to match the MOSFETs used in this project, but it doesn't hurt either. I have tested the circuit with matched and random parts with insignificant performance differences. There are other circuits that take advantage of matching, but because of the separate bias sources for each device and the high value of resistance between the Source pins of the two devices, this circuit is indifferent to matching. The exception is the case where R15 has a small value for the purpose of very high gain. Under this circumstance, matching will improve performance, and I recommend Vgs matching to within .1 V or better. For the procedure on matching devices, see one of the previous MOSFET project articles either in the pages of AE or on the Pass Labs website.

Page 66: Loudspeaker Builder

Potentiometers P1 through P4 are nominal values, and you should feel free to use parts that are similar in value. P1 and P2 determine the input impedance if used. While you might feel that it is desirable to use very high values so as to have a high input impedance, keep in mind that the input capacitance of the MOSFETs puts a natural upper limit on the value, so for best performance I don't recommend exceeding 25 Kohms. Similarly, P3 and P4 will influence the output impedance, and as you use higher values, the output impedance goes up. Lower values of output potentiometers will reduce the gain of the circuit, but not otherwise

PC board

Figures 4 and 5 show the PC board layout for both the channels and the power supply. Figure 4 is laid out to give two balanced channels. Note that the artwork is seen from the copper clad side, not the component view. Note that on Figure 4 the references R1, etc. are identical for each channel. For two channels you will want two R1's and so on. Figures 6 and 7 show the stuffing diagrams for these two boards. This view is from the component side, which is the mirror image of the PC artwork.

For those of you wanting to have a production house make boards for you, the website www.passlabs.com has the zipped Gerber files for these boards available for downloading. Keep in mind that this is a simple enough circuit to be wired up point-to-point, and if you have difficulty obtaining the circuit boards, I encourage you to use this approach. This is in fact what I did with the first version of the circuit, and it worked perfectly.

Page 67: Loudspeaker Builder

 

Heat sinks are provided for each power MOSFET, as they tend to run hot otherwise. On the board layout you will see that they are fairly small, and are designed to lay down on the board with a #4 or #6 screw and nut securing the transistor and heat sink to the board. Note that the metal portion of the TO-220 case is attached to the Drain pin of the MOSFET, and so both the case and the heat sink are electrically live.

Page 68: Loudspeaker Builder

A word to the wise is always appropriate when discussing the use of MOSFETs: they are sensitive to static discharge. Gate-to-Source voltages in excess of 20 volts have the potential to damage or destroy these parts. Take some care to avoid unnecessary handling, and avoid static electricity while handling. A modest amount of care in this regard is generally adequate, and when the MOSFETs are installed in the circuit, they are pretty safe if you are using the Zener input protection diodes.

And why wouldn't you use the input protection diodes? I suppose you could imagine that they reduce the "purity" of the circuit and potentially

connected to the chassis or earth ground through the power thermistor TH101. Do not neglect this connection. It is very important that the chassis be earth grounded and that the circuit ground be attached also to earth ground, either through or without the thermistor. Note that the earth side of the thermistor connects to the chassis through the mounting pads of the power supply board. It will be important to use metal standoffs here and check the connection with an ohmmeter or otherwise hardwire the connection of thermistor to chassis ground.

C107 is an interference filter capacitor across the AC line. It is optional, but if you decide to use such a capacitor, make certain that it is safety rated for the AC line. Also be certain to wire it in carefully, avoiding potential for short circuit to chassis or other components.

The primary windings of the transformers shown in Figure 2 are for 120 volt operation. Blue = 115A, Grey = 0A, Violet = 115B, and Brown = 0B. In this case we see that two sets of four wires each will be attached to the AC power, so that both primary windings of each of the two transformers see 120 volts AC. This is accomplished by tying two each Blue and Violet primary wires together (Blue + Blue + Violet + Violet) at the Hot side of the AC line. The cold side of the AC line is attached to the Grey and Brown primary leads of each transformer in a similar manner.

For 240 volt operation, the Grey and Violet leads are tied together, not attached elsewhere, on each transformer, so that the two primary windings are in series. The Blue leads of the transformers then go to the hot AC line, and the Browns go to the cold. Don't leave the Grey/Violet leads sitting bare if you do this.

Keep in mind that not only the primary AC line side of the power supply has potentially lethal voltages, but the secondary system is high voltage as well. Use extreme caution; there are few enough Do-it-yourselfers as it is. If you don't feel competent to handle this end of it, get some help. Most technicians have low resistance to a smile and/or a six pack of beer.

Firing up the circuit for the first time is no big deal. Preferably you use a Variac ™ to raise the line voltage a little at a time to check for excess current draw or smoking components. If you don't have a Variac ™ you will simply plug it in and stand back.

It is possible to test the power supply board without the main board. Simply load the V+ with 15 Kohms ¼ watt to ground and the V- with 15 Kohms to ground and look for the 80 volt values on the plus and minus rails and the 60 volt values on the plus and minus regulated rails.

Page 69: Loudspeaker Builder

A few voltage points will be helpful, all referenced to ground:

The Drain (case) of Q101 should be +80 volts DC or so. The Drain of Q102 should be approximately -80 volts.

The output of the supply V+ and V- should be plus and minus 60 volts DC. This can be seen on the Source pins of Q101 and Q102, or on the output voltage pads.

On the main board the Drain (case) of Q1 and Q2 should be about +30

 

Figure 8 shows the distortion waveform of the circuit at 1 KHz at 1 volt output. Note the second harmonic character of the distortion waveform. This is generally regarded as the most desirable of the harmonics, that is to say, if you have to have distortion, second harmonic is the least objectionable. Those well versed in circuit design might immediately comment that we expect to see third and other odd ordered distortion components in the distortion, given the symmetric nature of the balanced circuit. Odd ordered harmonics begin to dominate at higher levels of output, but at a couple of volts and below, we see even harmonics due to the subtle mismatching of the semiconductors.

Page 70: Loudspeaker Builder

 

Figure 9 shows the THD (Total Harmonic Distortion) & Noise performance of the circuit for output voltages from 100 mV to about 30 volts. This circuit has an R15 value of 124 ohms. Four curves are shown, reflecting the performance for 30, 40, 50 and 60 volt rails, and as you can see, performance is enhanced by the higher rail voltages and bias currents. Unless otherwise indicated, results reflect the 60 volt rails.

The circuit has a high voltage output, about the same as a 100 watt amplifier at the 1% distortion figure. It is not likely to be used at these numbers, but is a part of getting linear performance down at low output

 

will be about 166 KHz into 1000 pF. Lower values of output impedance can be obtained by loading the output to ground, but at the expense of voltage swing. For example without P3 and P4, loading the output with 1000 ohms will drop the output swing by half, but will double the high frequency roll-off point. As there is plenty of output swing, we can safely throw some away if a higher rolloff frequency is required.

This particular circuit, like the Bride of Zen, will not distort into low impedance loads; the gain simply goes down. As a result, you can safely drive 600 ohm balanced loads, getting the same performance as with a comparable input voltage, but at a lower output level. This is because the distortion figure is a function of variation in current through the MOSFETs, which is independent of the load impedance.

Page 71: Loudspeaker Builder

 

One of the figures of merit for balanced circuits is called the Common Mode Rejection Ratio (CMRR). As previously mentioned, one of the benefits of balanced circuitry is that it amplifies input differences while ignoring or rejecting common signal (noise). The CMRR of this circuit is slightly greater than 80 dB, as illustrated in Figure 13, which shows the performance from 20 Hz to 20 KHz. This is a factor of about 10,000 to 1, so that a 1 volt common input comes out as about .0001 volt when measured differentially at the output.. This figure was achieved with unmatched gain devices, but careful matching does not significantly improve the performance.

This figure was achieved differentially at the output. If you are using only

voltages, I have documented the distortion curves for voltages of 30, 40, 50, and 60 volts on each rail.

At 60 volt regulated rails, each channel draws 80 mA of bias current (40 mA passing through each MOSFET in the gain circuit). This is about 10 watts per channel. You can power two channels off this supply, or you can choose to give each channel its own power supply. The current drawn by the gain stages is roughly proportional to the supply voltages, so that the circuit draws about half the current with 30 volt rails.

Figure 1 shows a perfectly workable version of the gain stage, but I can't resist tarting it up a bit with some protection and gain controls. Figure 2 shows the circuit with these added. Zener diodes Z1, Z2, Z3 and Z4 form input protection networks which prevent input voltages in excess of 9 volts or so. Higher values can be used, with 16 volt Zener diodes being the practical limit. It is not essential that the input protection diodes be used, but without them greater care will be necessary in connecting signal sources.

There are four potentiometers also shown if Figure 3. P1 and P2 can be used to attenuate the input signal. Because the balanced input characteristic is perfectly preserved if P1 and P2 are set at the same value, they perform best as precision ganged controls, but this is not essential otherwise. P1 and P2 can be used to protect the input from higher voltage signal sources, or they can be used as volume controls. P3 and P4 perform attenuation at the output, and can also be used either as a volume control or simple gain adjustment. P3 and P4 attenuate the output noise of the circuit as well.

P5 allows intrinsic adjustment of the circuit gain without having to match the potentiometer values, and as shown with a 500 ohm potentiometer, allows adjustment from about 10 to 20 dB gain. The use of any of these potentiometers is optional and independent of each other. The value of the potentiometer is flexible with the value given as optimal for typical use. Ideally P1 and P2 are the same value, and P3 and P4 are the same value in order to preserve the CMRR figure.

Part List

Components from Figure 3, the main circuit, are numbered from 1 to 99, and represent one channel only. Components from Figure 2, the power supply, are numbered from 100 to 199. You

Page 72: Loudspeaker Builder

can use the power supply board to power both channels, or you can use two power supply boards, one for each channel. You will note on the PC board layouts that two separate sets of input power connections, one for each channel, are on the main PC board to facilitate use of one or two power supplies.

With the exception of the power transformer and the PC board, the components used in this project are available from Digikey (800) 344 4539, and I have included the appropriate Digikey part numbers for parts. Digikey handles the Yageo metal film resistors, I used Dale RN55D types.

In general, a wide variety of substitute parts will be acceptable so long as they have the appropriate wattage and voltage ratings. The MOSFETs used are fairly generic N channel devices in TO-220 degrade the quality.

Substitute capacitors should have a 100+ volt rating. Many do-it-yourselfers have called me up and explained the wonders of various better parts, capacitors, resistors, wires, and diodes, that I could have specified that would improve a project. Some of you have been kind enough to send samples, which I much appreciate. Oddly, nobody sends me better transistors, and considering that gain devices account for most of the distortion, this is what I would really prefer.

As a rule, I like to leave the more exotic parts out of these designs, and I have several reasons. First, I want to make this as simple and inexpensive as possible for most of you. Second, exotic parts are not the particular point that I am trying to emphasize in these projects; rather it is the quality that can be achieved with very simple approaches.

Most important, I have saved these embellishments for you to perform independently.

Part List for Main Board (one channel only)Designation

Part

Digikey #

Q1, Q2

IRF610 N channel MOSFET

IRF610-ND

Z1-4 (optional)

1N4739 9.1 V Zener diode

1N4739ACT-ND

P1-2 (optional)

Page 73: Loudspeaker Builder

10 Kohm Potentiometer

381N103-ND

P3-4 (optional)

5 Kohm Potentiometer

381N502-ND

P5 (optional)

500 ohm Potentiometer

381N501-ND

C1-2

10 uF film capacitor

EF1106-ND

C3-4

1000 uF electrolytic

P6476-ND

R1-6

750 ohms @ 3 watts

P750W-3BK-ND

R7-10

221 ohm ¼ watt metal film

221XBK-ND

R11-14

100 Kohm ¼ watt metal film

100KXBK-ND

R15

124 ohm ¼ watt metal film

Page 74: Loudspeaker Builder

124XBK-ND

R16-17

22.1 ohm ¼ watt metal film

22.1XBK-ND

R18-19

221 ohm ¼ watt metal film

221XBK-ND

Heat sinks (2)

1 watt sinks for TO-220

HS104-1-ND

Part List for Power Supply BoardDesignation

Part

Digikey #

Q101

IRF610 N channel MOSFET

IRF610-ND

Q102

IRF9610 P channel MOSFET

IRF9610-ND

B101-102

Diode Bridge

2KBP02M-ND

Z101-116

1N4739 9.1 V Zener diode

1N4739ACT-ND

Page 75: Loudspeaker Builder

C101-102

1000 uF electrolytic

P6476-ND

C103-106

10 uF electrolytic

P5593-ND

C107

.047 uF Line filter Capacitor

P4637-ND ACline rated)

R101-102

4.75 Kohm ¼ watt metal film

4.75KXBK-ND

R103-104

221 ohm ¼ watt metal film

221XBK-ND

TH101

TH101

KC006L-ND

Heat sinks (2)

1 watt sinks for TO-220

HS104-1-ND

F101

1 amp fuse, slow blow

F319-ND

Fuse Holder

Page 76: Loudspeaker Builder

panel mount 3AG

WK0002-ND

AC Power Inlet

-

Q212-ND

T101-102

D4007 30 + 30 volt

Avel Lindberg

Tel (203) 355 4711

Fax (203) 354 8597

Page 77: Loudspeaker Builder

Construction Notes

All the resistors in the project are ¼ watt 1% metal film types. My personal preference is Dale, but there are many good, possibly better types. If you decide to go crazy and use Vishay or other exotic parts, the circuit might sound better; it is very unlikely to sound worse, and no matter what you do, the result will still probably be much cheaper than buying a readymade preamp.

The main PC board shows jumper wires where the pads are designated for P5. Naturally if you use P5, you will replace the jumpers with wires leading to P5.

Page 78: Loudspeaker Builder

There are two sets of power supply connections, one for each channel, and this allows the use of one supply for both channels, or two separate supplies, one for each channel. In any case, you will be using separate wires to bring the supply voltages into the main board.

introduce distortion, although I have to say that I have not measured any substantial distortion, nor have I been able to identify it subjectively. Bad Zener diodes will definitely create distortion, however, and this effect is more common than you would think.

Speaking of input and output connections, I have not specified connector types for this project. Usually for this sort of project I use both XLR and RCA connectors on both the inputs and outputs. The inputs are female XLR with pin 2 = plus, 3 = minus, and 1 = ground. At the output it is a male XLR connector. In parallel with the XLR connectors are 2 RCA connectors, one for each polarity, with XLR pin 2 connected to the "live" of the positive RCA and pin and XLR pin 3 connected to the "live" of the negative RCA. XLR pin 1 is attached to the grounds of both RCA connectors. These grounds are isolated from chassis ground. The case of the XLR connector, if metal, and shield are attached to chassis ground.

The power transformers used here are toroidal types having dual primary and dual secondary windings. The schematic shows secondary output wires A, B, C, D for each transformer which will attach to the ABCD designations on the power supply board. The colors coming out of the Avel D4007 are: A = Black, B = Red, C = Orange and D = Yellow. It does not matter which transformer goes to which set of connections, as they are identical. You will note that B and C wires are simply tied together (on the PC board) for each transformer and do not connect to anything else.

The wiring for the primary AC side of the transformers is not provided on the PC board, and must be accomplished point-to-point. I strongly suggest that you use standard safety rated AC inlet connectors and fuse holders.

Remember safety first. The Earth ground going back to the AC outlet in the wall is connected to the chassis, and then the circuit ground is

Page 79: Loudspeaker Builder

volts. The Gates of Q1 and Q2 should be at ground. The Source pins of Q1 and Q2 should be about -3.5 volts or so.

 

If you get these voltages, everything should be just fine.

Several points about operation:

If you have elected not to use input protection Zener diodes Z1-4, then you will want to be quite careful when you plug sources into the inputs. Generally this means touching the ground of the interconnect cable to the ground of the female connector or to the chassis before inserting the plug. The input Gates of the MOSFETs can take 20 volt peaks. I have seen them withstand 80 volt transients, but not reliably. If you don't use the Zener diodes, but do use P1 and P2, then you can always turn the potentiometers to 0 (counter clockwise) during connection.

Also note that if you do not use Zener diodes for input protection, you will not need R18 and R19, and they may be replaced by wire. Their function is to ensure stability for the circuit driving the input, since directly looking at a Zener diode can drive some circuits into oscillation. This is also why we use R7 and R8, and R9 and R10.

If you blow out a MOSFET, nine times out of ten it will be excess Gate- Source voltage.

Page 80: Loudspeaker Builder

Speaking of transients, I have not provided for turn-on transient suppression at the output of the preamp circuit. As with the Bride of Zen, to avoid potential damage to a loudspeaker, you should ensure that the power amplifier is turned off when the preamp is powered up, or if you have chosen to use P3 and P4, you can reduce the output to 0. The output transient is balanced, by the way, so that it is much lower in value when evaluated differentially.

A turn-on/off suppression relay would be a nice add-on project. I invite interested parties to send their design to Audio Electronics magazine.

If you elect to use P1-5 to adjust the parameters of the circuit, you can gain quite a bit of flexibility in its use. In general, the best performance occurs with the higher value of P5, giving 10 dB gain instead of 20 dB. If you don't need the gain, use 10 dB.

Adjusting the input signal level using P1 and P2 will allow you to optimize the operating point against the signal source level. The best performance occurs at approximately 1 or 2 volts output of the circuit (assuming P3 and P4 are turned up all the way). If you are running 10 dB of gain, this means an input level of .3 to .6 volts, and if your source has much higher output, you can turn down P1 and P2 so as to set the input level to this region.

Once you have obtained the optimal gain and input level, adjusting the output level with P3 and P4 will give you an effective volume control which attenuates circuit noise as well as level. Of course you can use any of these pots as volume controls as you like. Keep in mind that the common mode rejection figures depend of the matching of the values of P1 and 2 and separately P3 and P4. If you are particularly concerned about this figure (and I am not) you might consider using a precision dual pot for these.

voltages. The distortion climbs smoothly with the output voltage, and clipping is graceful.

Below the 1 to 2 volt region, increasing noise drives the curve upward. This noise figure works out to about a 5 microvolt input noise (-106 dbV) or about 17 nanovolt-per-square-root-Hertz input noise floor. This is comparable to your typical noisier variety of FET op amps

Page 81: Loudspeaker Builder

.

 

Figure 10 shows the performance with R15 at 430 ohms, for a gain of 10 dB. The distortion is lower by a factor of about 10 dB (no surprise), and the output noise is reduced by the same amount, reflecting the same input noise.

 

Figure 11 shows the THD & noise, again for four different rail voltages. The output is at 2 volts and is plotted versus frequency. The figure is the same from 20 Hz to 20 KHz.

Figure 12 shows the frequency response, measuring -1 dB at 200 KHz at the top end. The graph does not show the low frequency rolloff, which would be worst case -3 dB at 4 Hz for a 10 Kohm load with a 5K value for P3 and P4. Without P4 and P5 and with a 10 Kohm load, the rolloff is about 1.5 Hz.

The slew rate of the circuit itself is exceedingly high; it does not have a slew rate as such, but rather a classic RC risetime which is dependant on the capacitance seen at the output interacting with the resistive output impedance. The rolloff for the approximately 1000 ohm output

one polarity of the output signal, you will find the rejection is only about a factor of 10 (-20 dB). Obviously this isn't nearly as good, but in point of fact it still represents a factor of 10 better than an unbalanced circuit, and usually this is plenty. I have read assertions that 60 dB CMRR figures are the minimum acceptable, but no good reason to why this figure is essential. In my book, any reduction of noise picked up is a plus.

In actual practice with real systems, I have noticed that there is usually about a 20 dB difference in background noise between balanced and unbalanced systems, and a circuit with 20 dB rejection will preserve this difference fairly well. As an alternative to having to use the balanced

Page 82: Loudspeaker Builder

output only, you can replace the resistors R3, R4 and R5, R6 with 40 mA active constant current sources. This will restore the 80 dB CMRR figure for unbalanced output.

Conclusion

This is a particularly good sounding circuit, and I think it sounds significantly better than the Bride of Zen, although I would be hard pressed to explain why. It seems more liquid and has greater depth, while BOZ is a bit dry by comparison. It might be distortion cancellation in the balanced circuit, or it might be the greater dynamic range afforded by quieter balanced operation and higher output swing. As always, I encourage you to build it and decide for yourself.

Comments and questions are welcomed. The best way to reach me is by e-mail through the Pass Labs website, www.passlabs.com, or more directly: nelson@ passlabs.com. This method gets a reliable, if short, response.

Snail mail is Pass Labs, PO Box 219, Foresthill CA 95631.

This pretty well finishes the preamp offerings in the Zen series. Next: the penultimate Zen amp.

Ground Loops

Kent English

Introduction

Your brand new component is hooked up, fresh from the box, and the fi rst time you power it up is a sonic disaster; it hums, it buzzes and in general sounds absolutely dreadful. Glaring at your equipment or dealer doesn’t help and twisting knobs only makes the noise worse; what now?

From years of experience we fi nd that the vast majority of excessive noise in audio electronics can be traced directly to poor grounding techniques. While we recommend balanced interconnects on your audio components whenever possible, it must be understood that balanced interconnects address only the problems of induced noise. Ground loops are a very different problem, and not at all related to the issues of induced noise.

A Little theory

In order to successfully battle ground loops you must fi rst understand why they occur. Each component in your audio system has at its heart an internal ground. The key points to understand are that there is no such thing as a perfect ground and that no two ground points within any system are ever exactly equipotential to one another.

Wherever two grounds of different potential exist within a system, there is a likelihood of a ground loop related noise issue. When devices are tied together with interconnect cables, these of necessity tie the signal grounds of the interconnected devices to one another. This communication between the two signal grounds is a necessary and desirable circumstance, the

Page 83: Loudspeaker Builder

problem of “ground looping” arises when this connection occurs in more than one instance. The typical culprit being the safety ground supplied by the power cordage or rails in a rack mount being in direct contact with the signal return grounds.

These situations create a closed loop, where current fl ows from one unit’s ground to the other unit and back to the fi rst unit through the additional ground connection, provided by the power distribution network. Typically the impedance’s of these unwanted circuits are quite low, in the order of very small fractions of an ohm. Don’t expect

Pass Labs: Articles: Ground Loops

part of the input circuit only. Shield ground should not be connected on the source end of the wire, only at the input component end; label them and don’t forget! This will of course mean that at the input component end of cable the shield and ground signal conductors will bond together.

This would be the preferred connection for all unbalanced connections where the manufacturer has taken the care to isolate chassis ground from signal ground, unfortunately this is as yet not a universal practice in consumer audio.

The same sort of logic should apply when fabricating XLR cables. Start with a cable that has three wires in addition to a separate shield; pin one on the connector is ground, pin two is positive input and pin three is inverted input. The case connection on the input component end XLR becomes your only shield connection; labeling here is unnecessary, as they are directionally polarized cables by virtue of construction.

If one component has a safety ground isolated from signal and another does not, chances are very good that ground loops will not become an issue. When ground loop problems arise it is most often a result of two interconnected components, each having safety grounds and signal grounds joined inside the component. In these circumstances one of the grounds will have to be abandoned, or you will have make them all more alike….. your choice.

Ok, lets say you have interconnect cables and components that you fancy, and re-engineering or otherwise damaging the product is out of the question, now what?

Logically one would think, you could eliminate ground loops by disconnecting the power-cord grounds on all your gear. Some people might try to break the ground connection by cutting the grounding pin on the power cord, using a cheater plug, cutting the ground wire inside the equipment, taping over the grounding connector etc. As logic predicts this may effect a cure for the noise.

Do not do this. Removing the ground connection isn’t right! It is against electrical safety regulations and potentially very dangerous. Removing a safety ground connection can defeat the actions of the noise fi lter or spike protector inside the equipment. If the ground connection is cut then a fault in the insulation inside equipment may apply dangerous voltages to the equipment case instead

Page 84: Loudspeaker Builder

Pass Labs: Articles: Ground Loops

located closer still too the mains panel. Many power strips have MOV’s and neon lights; if you are seeking the ultimate in RFI free power, do not put these devices on your entertainment system. Both devices can input a small but measurable amount of noise on the power line. Is this small amount of noise signifi cant, no but it is certainly cumulative.

MOV’s have a place in your home electrical system, to be most benefi cial they should be as close to the mains panel as possible. The best of the MOV devices hard wire directly onto the breaker panel buss bars. MOV’s however wear out, and will occasionally need replacement. They tend to fail catastrophically rather than gradually and failed units are not too diffi cult to spot.

Many small gains in noise reduction can and do have a dynamic impact on what you ultimately hear in a highresolution audio system. In many instances taking advantage of these incremental gains represents little additional expense or effort.

to measure this resistance with your handy multi-meter, it likely does not have the required resolution or sensitivity. Accurate measurement requires the use of a device known as an impedance bridge. Fortunately, the cure for ground related noise rarely requires this level of diagnostic sophistication.

In accordance with the revered teachings of Georg Simon Ohm, these voltages while quite low are capable of generating signifi cant current fl ow. It is then these “looping” currents that create the unwanted noise by impressing their signature on low level signals, usually in the form of common mode noise.

In order to minimize ground loop issues Pass Labs never manufactures equipment with signal ground and chassis ground contiguous. By separating signal ground and safety grounds, connecting units together should never cause ground loop issues; however, not all manufacturers follow this line of thinking.

Now What

Once you have an understanding of what causes ground loops, they should with some persistence and effort become non-issues. To the greatest extent possible you will need to keep safety grounds, signal returns and shielding of low level cables separate.

Unbalanced cables have persisted as the consumer audio norm, in spite of their inherent frailties. In systems with very few components RCA type connections work reasonably well, but as systems (particularly A/V systems) become more complex their successful implementation becomes problematic. If you are running unbalanced cables, always use two conductor shielded wire. Using the more common single conductor inside a shield mandates that you combine signal return and shielding to the same wire; thus breaking the preferred protocol.

Page 85: Loudspeaker Builder

Shields are to keep stray noise out of component inputs; the common or signal return is part of the signal path, two opposed tasks. Because of these separate tasks your cables should be directional, and the shields should be

of interrupting a fuse. Running without a power ground will not automatically electrocute you but will make this much more probable if something goes wrong in your system.

Many a noted authority has suggested repolarizing your equipment by inverting the power cord thus reverse the hot and neutral power connections. Do not do this. As a practical matter this may slightly reduce your powersupply related noise issues, but there is a potential down side as well. Inverting your power cord will put the internal fuse and power switch in the neutral power line (goodbye protection). In case of accident as a result, insurance companies will be laughing at your heirs!

If we cannot separate the signal grounds and safety grounds our only other option is to make them as alike as possible by careful confi guration of mains supply and safety ground. There are a number of power distribution techniques intended to reduce or at least minimize ground loop issues. The most common technique is called a star distribution. In star distribution, a point is chosen as the arbitrary lowest voltage potential ground. From this point radiating in as many directions as necessary power will reach all interconnected components. Then all safety grounds will terminate back to the master safety ground at this common point. These star confi gured ground connections must be made of heavy gauge wire and all arms of the star must be the same length and gauge wire.

When all the ground conductors to a star connections central point are of equal length, then the ends of the star are very close to the same ground potential. Assuming faultless implementation of this grounding; signal wiring between any equipment grounded to the star will be at zero potential, thus avoiding ground loops.

The most cost effective way to do this by connecting all your low level components into a quality power strip rather than numerous wall outlets. The wall outlet chosen to plug the power strip into should be the one closest to the mains panel for that particular branch circuit. Anything you do to reduce the overall electrical resistance of the power supply circuit brings the benefi t of having “ground” closer to earth potential. Lowering power source impedance’s in this manner allows your components internal EMI / RFI fi lters to perform as intended.

Any noise producing devices on the same branch circuit, such as fans or portable fl uorescent lights should be