The tuning of phonograph reproduction

by proper cartridge loading & channel balancing along with modifying an inexpensive PCB and a bit of historical perspective

Robert E. (Robin) Miller III BSEE AES SMPTE BAS Filmaker Technology www.filmaker.com 9/30/2014iv

Setting up a turntable, tone-arm, pickup, & preamplifier must be attended to by users in order to

realize good performance from their phonograph systems for enjoying and restoring 78, 45, and

33 rpm disc recordings. Two crucial issues are often ignored that diminish sound quality: 1) proper capacitive loading of moving magnet/moving iron (MM/MI) cartridges for best resulting

frequency, phase, and transient response; and 2) proper balance for best stereo soundstage, and for monaural reproduction to cancel not-recorded vertical distortions. Both are addressed,

and implemented in an inexpensive ($30 less case & power supply) audiophile phono stage.

Vinyl has made a comeback from the past and from technological obsolescence. Along with shellac 78rpm discs, this

analog medium is a treasure of more than a century of musical

performances, forgotten or deceased artists, and broadcasting

history. Phonography is difficult compared to the ease of use of

the CD and streaming media, and little new hardware is being

manufactured. Vintage used equipment was built to last and be maintained, but replacement styli are a quagmire of quality.

Before exploring the disc and its modulated groove, and the

pickup cartridge and its stylus and preamplifier, an appreciation of

the most difficult technology to design and most expensive to

build: the cartridges stylus a tiny cantilever with a diamond tip. Micro-fabricated of aeronautical aluminum with a needle magnet

of rare earth Samarium Cobalt, the sound of disc reproduction is arguably the sound of its stylus. Pulled backwards over a records landscape like a cowboy caught in a stirrup and dragged by his

horse, following ruts rushing up from behind, bumping over

rocks, but only grazing hoof-print imperfections in vinyl too small

to feel. Intense pressure causes extreme temperature that deforms

temporarily the vinyl at the point of contact, akin to an ice skate

melting a thin film of water so the blade can glide. Advanced tip

"profiles" reduce this pressure, and its eventual wearing of groove

and stylus, by elongating the point of contact up each wall, even

as they narrow it in order to trace better the high frequencies cut

by the flat-across, chisel-shaped cutter of the disc recorder. Yet

the stylus assembly needs replacement only every 500~1,000hr!

The market for used vinyl is thriving. Direct-drive turntables are still made, DJ heirs to Technics SP-broadcast models. L, an audiophile tone-arm for high-compliance tracking at 1g; R, a 14in broadcast arm tracking at 2g.

Subjects of graphs on p2, Stanton #881 (L) & #681 with Stereohedron styli trace the same groove a fraction of a second apart for direct comparison. These "calibrated" (Steinfeld prefers "certified"1) models were chosen by broadcasters and disc mastering engineers for quality control while cutting.

1. Capacitive & Resistive loading of moving magnet (MM)

and moving iron (MI) phonograph pickup cartridges

One determiner of phonograph quality likely has tainted the

reputations and reviews of cartridges for decades improper electrical loading. This impedance completes a manufacturers design of its cartridge. Although critical, they are external, user/

installer-supplied components beyond the manufacturers control.

All MM/MI cartridges rely on user/installer-supplied components to meet their manufacturers designed characteristics. Loaded off -spec results in replay often described as too dull or too bright. Coincidentally it might be perceived to compensate for errors elsewhere in the system [Pickering].

1 Richard Steinfeld Handbook for Stanton and Pickering Phonograph

Cartridges and Styli, self-published (2010) rsteinbook@sonic.net a compendium of more than 100 products based on employee interviews.

p2 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

Oblivious to a loading mismatch, audiophiles might blame a

cartridge for being either dull (C-load too low) or bright (C-load too high), a non-flat response wrecking timbre (tone quality),

the holy grail of sound reproduction. Cartridge manufacturers

intended proper loading be the responsibility of the user/installer.

The chart tells the story of two of the more than 100 products

of musician Norman Pickering and engineer Walter Stanton for

home enthusiasts, disc mastering engineers for quality control,

and the rigors of broadcasting.2 Playing a well-regarded test

record (CBS Laboratories STR- series), their data shows results

tuning their upper-end cartridges by varying the external load capacitance. The 680/681 (certified version) requires 275pF

(pico-Farad) 10%; the lower impedance MM 880/881 also

275pF, but 20%. Unless compensating other system errors,

operating outside these tolerances will have a detrimental effect

on the performance of your cartridge and therefore on your sound.

All MM/MI cartridges employ a C-load for a high frequency

(HF) resonance to extend the HF response. As shown, higher

values of C-load create a peak in the mid-high range, then falling

steeply at HF. Smaller C moves the peak toward ultrasonic VHF

frequencies at some cost to mid~HF. Absent the designed C-load,

the cartridge is not complete, and cannot perform. To say that

proper capacitive loading is important almost goes without saying,

yet we are discussing it precisely because it is so often ignored.

Overall perfect flatness of frequency response curves is the

elusive goal, but failing that, an important perceptual truth is that

bumps in frequency response are more deleterious to the ear than

equal dips. So leveling bumps at the expense of broad, shallow

dips results in the precise 275pF specification for these cartridges.

The total capacitive load, C-load, seen by the cartridge is additive, consisting of the distributed capacitance along the arm

wiring, cabling to the preamp, and the lump of capacitance inside

the preamp at the input jack in parallel with a 47k resistance. A typical preamp has a 150pF capacitor. Cabling typically is 125pF.

It is easier to increase (add) capacitance than to decrease it.

You could, with a pF-level capacitance meter or impedance bridge, measure your cables, lable them, and even measure the

2 Adapted by the author from Ballou Handbook for Sound Engineers:

the New Audio Cyclopedia, published by Howard W. Sams (1987).

preamps input C if it isnt in the manual or you dont open it to see the component. Then add Cspec = Cwiring + Cpreamp to compare

with the manufactures specification. How does your arithmetic compare to typical values for Stanton models: 275 = 125 + 150?

But what if the cables are not as good? Or are lower C/ft?

Or the cart spec is 620pF? Or a different C (or none!) is inside the

preamp? Or what if for playing different types of discs, you

switch between differing cartridges. Or different arms/turntables?

Solutions depend upon your need for switchability and your DIY

skills. On the web are Y-connectors with plug-in capacitances. If

you measure and tag your phono interconnects, select the closest

to total each carts spec. Substitute a new capacitor inside your preamp. OR if you have $10, a bit of room inside, and can drill a

in hole, mount a switch and solder six caps, as illustrated below.

DIY C-load selector for $10 in parts inside preamp for tuning a cartridge, adding green switch at right with six tiny (yellow) ceramic capacitors for selection of three values. Ferrite beads filter radio freq. interference (RFI).

For the likely difference in C-load tuning requirements of two

dissimilar pickups, a second turntable or arm, add a second switch

in conjunction with the C-selector. This requires four in holes: two for switches and two for the second pair of phono connectors.

Shure M97 & M35x specify 200~300pF. Audio Technica

95E recommends 100~200pF. Nagaoka MP-11s reported range is 100~620pF! Specifying a range instead of a precise design

value is a cop-out to curry favor with busy buyers. The C-selector

values can easily be substituted for the range needed. With a level

meter and a respected, unworn test frequency record, the loaded

carts frequency response could be measured. But a trained ear on one online forum posted these impressions of an AT 95E that

required an unusually low Cwiring of 75pF:

Cwiring + Cpreamp = Ctotal Subjective quality of sound

75 47 122 "Impressive, but only initially"

75 22 97 "Cleaner, defined soundstage"

75 100 175 "Harsh [bordering on shrill]"

The choices of 22, 47, & 100pF refer to an unmodified preamp in

the Appendix that after modifications provides 21 C-load choices.

These subjective results conjure the Stanton 881 chart. For the AT loaded with 122pF (cf. the Stanton 881 curve at 500pF)

brightness is only initially impressive, but air at very high frequencies is noticeably attenuated. Approaching the highest

specified 200pF (cf. Stanton 1,000pF curve) the AT accentuates

the peak and lowers its frequency and sounds harsh. But the

lowest Cspec=100pF flattens the resonant peak (cf. Stanton curve at

p3 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

275pF) and may be the real AT 95E design value, although it is

difficult to achieve. Low impedance moving coil (MC) cartridges

are not affected by C-load3 none is needed. However for more

ubiquitous MM/MI cartridges, C-load tuning is important!

Varying Cwiring offers another run at the goal, but a lot of

poppycock surrounds audio merchandise, especially cables.

Audio wiring isn't mysterious, nor need it be expensive. Maybe

free, tangled in your box of spares! Even inexpensive audio

interconnects work within limits. Knowledge is power to ward

off being ripped off. The capacitance of a cable is proportional to

its length: Twice as long, twice the C. Cheapest, very thin audio

cables measuring 80 pico-Farads (pF) per foot of length are OK if

limited to 1 ft, with the preamp inside or adjacent to the

turntable plinth, but even at this short length their poor shielding

may be inadequate against radio frequency interference (RFI) at

low 5 millivolt phonograph levels. These cables are for low

impedance, line level connections, nominally ~300mv (10dBv).

Mid-priced coaxial cable might have 30pF/ft, the best 20

pF/ft. To increase Cwiring, pick a less expensive 40 pF/ft cable, or

make it longer. To reduce Cwiring, pick a more expensive 30 or 20

pF/ft coax, or make it shorter. It's not about splurging for exotic,

highly hyped, ultra-low pF/ft cable, but to optimize economically

the total C-load: 3ft of 40pF/ft cable, 4ft of 30pF/ft cable, or 6ft of

20pF/ft cable each adds 120pF toward the grand total Cspec.

2. Balancing two channels for monaural & stereo disc replay

The Golden Mean (aka Fibonacci ratio) the sweet spot of perception. Its like in reproduction of sound in stereo is balance.

Unlike the Golden Mean's 61.8/38.2 split, audio balance is strictly

50/50. A small imbalance can destroy what audiophiles term the

soundstage (audio engineers speak of localization & spatiality). Playing monaural vinyl or shellac on a stereo system, imbalance

means falling short of mono's inherent cleanliness from distortion.

In stereo using the conventional equilateral speaker-listener

triangle, it only takes a difference in level between channels of

15dB [Theile 2001] to pan an auditory event fully to one side or

the other, seeming to emanate from the loudspeaker on that side

alone. In the middle, it takes little difference to move the image

quickly off-center, skewing the soundstage non-linearly. This is

perceived to leave what is termed the "hole in the middle" of the

entire image, shoving farther out voices and instruments recorded

only slightly left or right. If a stereo recording is reproduced only

slightly off balance, say by only a couple dB of the 15, the entire

image will shift quickly toward that speaker, leaving an even

larger void through the center and toward the other side.

Assuming a well-made recording (a big assumption), every

component in the reproduction system can introduce imbalance caused mainly by component variations and degrading further

over time in the present discussion electronic gain or transducer sensitivity errors, some of which cancel, others adding to a worst

case. The "balance" control was invented to compensate for the

sum, adjusted by ear if not by a sound pressure level (SPL) meter.

Even with the rest of the system precisely balanced, playing stereo

vinyl with a pickup cartridge that just meets its specified 2dB max

imbalance will throw off its imaging if not corrected in the phono

preamplifier before the signals go on to the rest of the system.

3 With MC carts, cable noise immunity is less critical, and capacitance is

a non-issue that can be ignored, but not lack of stylus interchangeability.

Cartridge manufacturers did not intend this output sensitivity

variation to go uncorrected in the preamp, just as tape machine

makers provided separate channel gain controls for variations in

the playback tape-head. Meanwhile amplifier/receiver makers fall

back on their systems overall balance control to save the cost of a few parts, or complaints from users of too many fiddly bits.

Playing vintage monaural vinyl is affected differently, but

critically. The mono record groove is usually modulated laterally,

horizontally only, positive-going away from the turntable spindle,

then towards it negative-going. Cylinders and some discs were

recorded vertically, hill-and-dale, but suffer higher distortion. The vertical drive the groove forcing the stylus up but having only gravity for its descent is mechanically single-ended, and so susceptible is to added even-order harmonics. Laterally the

stylus is driven both left and right by the groove walls, and so is

push-pull, which is not subject to adding even-order distortion. Because most natural sounds consist of either all harmonics OR

odd harmonics only, adding even harmonics disproportionately

fouls natural reproduction of the timbre of those sounds.4

Early stereo in a single groove used lateral motion for the

Left channel and vertical for the Right, which contained more

distortion and sounded inferior to the Left. (Early stereo radio

simulcast one channel on AM, the other on FM.) Stereo caught

on after L & R channels shared distortion equally, perceived as

balanced! With no difference for reference, remaining distortion

in disc reproduction was acceptable. For vinyl, each 45 wall of

the V-shaped groove is assigned a channel, each with modulation

equal parts lateral & vertical, as in the following left illustration.

At right [adapted courtesy Ortofon], the vectors of 45/45 stereo and stylus horizontal/vertical motion are mathematically related. Lateral motion results from adding the signals L+R. Vertically, any difference LR, purposeful in stereo, is distortion in mono.

5

While a stereo cartridge, having vertical compliance, is the

safest way to play mono records, it reproduces unwanted vertical

junk that isn't recorded content, but artifacts to be eliminated: The

spitting of sibilants and other HF above LP cutoff 6 due to 2nd harmonic distortion from "pinch-effect" when a spherical stylus

cannot fit high frequency cornering, especially in inner grooves.

Or losing contact due to too little tracking force for a cantilevers compliance, or too little or too much anti-skating force. These

tracing distortions trump those of angular tracking errors of the

tone arm and cartridge [Yamamoto], as does bumping over clods

of dirt that can produce 30dB+ pops, often triggering amplifier

ringing (oscillation). Note that these artifacts remain in stereo.

4Stopped at both ends, stringed instruments generate all harmonics (even

& odd); also conical-bore saxophones. Open at one end and cylindrical,

clarinets & trumpets produce only odd. None has even harmonics alone. 5In the Mid-Sides (M-S) form of stereo, M=L+R (lateral), S=LR (vert.). Then L=M+S and R=MS. (Reduce each result to 70.7%, or 3dB.)

61,970Hz cutoff with 1mil tracing radius [Goldmark, Columbia 1948],

now 2.8kHz with a 0.7mil spherical stereo stylus, 4.9kHz with a 0.4mil

elliptical, or 9.8kHz with a 0.2mil line contact [calculated by the author].

p4 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

Illustrated above is how 2nd

harmonics are mechanically

generated on playback (they are not in the recorded groove). The

flat chisel-shaped recording tool cuts walls (in red) that are

parallel radially. However, while a 0.7mil (18m) spherical

replay tip with circular cross-section fits and drops down where it

can, at waveform peaks, it is pinched upward, squeezed highest at

the modulations midpoints, where the stylus moves fastest. The 0.3mil (8m) elliptical tip fits better everywhere, rising far less.

Rising & falling happens twice for each recorded cycle of

high level sounds produces 2nd

harmonic distortion artifacts an

octave above the recorded tones. Adding even-order harmonics

disproportionately within the natural spectrum of instruments

containing all harmonics, even & odd, alters their timbre, making

them sound discolored. Worse, adding 2nd harmonic to many instruments that normally exhibit only odd-order harmonics in

their spectrum changes their timbre to sound quite unnatural.

To illustrate the range of stylus dynamics, 10 vectorscope images are rotated 90 to correspond to stylus motion, viewed from the front. L, signal only on the Left channel and groove wall. R, signal only on the Right wall.

L, lateral-only (mono) motion. R, vertical (mono out-of-phase). Referring to the illustrations on p3, the groove moving rightward results from positive-going signals in-phase for both channels. But upward/downward motion is 180 out-of-phase intentional in stereo, but distortion in mono.

L-Highly correlated stereo (mostly mono). R-highly uncorrelated stereo (phasey and at high levels more likely to jump the groove). Also illustrative of how a centered voice L can be instantaneously distorted by pinch effect to look like R with added sibilance spit.

L, well balanced, nicely spatial stereo recording of a grand piano. R, same piano softer, accompanying a solo violin standing at right. This ferocious groove and stylus activity is purely mechanical unlike the nonchalantly inconvenienced electrons and digital bits. Steinfeld ponders why vinyl sounds better than it has any right to!

L, after preamp balance trimming using a reliable test record, this kick drum, likely intended to be panned center, shows an imbalance of 1.2dB, discernibly left of center. R, a mono record played with a stereo cartridge contains unintended vertical distortion that should be nulled using preamp balance trim to approach the Lateral-only (mono) screenshot. 2

nd harmonic

pinch effect is out-of-phase in stereo, disembodying transients & HF spit.7

Furthermore from the vector figure on p3, the cartridges output signals for a rising motion are positive-going on the Left

channel but negative-going on the Right, and v.v. Hence the 2nd

harmonic distortion produced is 180 out of phase! Adding pinch

effect distortion suddenly to loud peaks causes the phony partials

(only) of the voice or instrument to appear instantaneous beyond

the speakers, flashing randomly around the listening room. These

errant spitting sounds of vocal sibilants, buzzing brass, and dirt

pops & percussive transient ricochets are ghostly, disembodied

from the image of the voices and instruments in the soundstage!

7Wild forces act on the stylus within the split second of these captures.

Stanton/Pickering cantilevers are suspended inside mounting tubes by a

donut-shaped rubber elastomer acting as a compliant fulcrum, stabilized

fore-and-aft and from twisting by a wire soldered at the back of the tube.

p5 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

The simplest way to cancel these unwanted signals vertical components is to parallel (short-circuit) the L and R outputs of the

cartridge. However, the sensitivities between channels of the

cartridge might be different by as much as 2dB in most

manufacturers specifications. In this case, the reduction of vertical non-content is only -11.7dB.

8 As a change in level of

10dB, softer or louder, is perceived as halving or doubling of

volume, the reduced distortion is still half the volume it was.

Now add that the cartridge might be slightly rotated in the

headshell or arm, or the disc cut with a slight error in balance, as

mono records after 1958 increasingly were cut on stereo lathes. A

total imbalance of 2dB and merely paralleled without balance

correction reduces vertical artifacts only 9.5dB or 67%. In this

example for pinch distortion reaching 10%+, simply paralleling

the cart outputs for monaural reduces it to 3.3% or more 2nd

harmonic distortion not exactly high fidelity. (NB again, in the example, that distortion in stereo remains up to 10%+!)

So it's best to monauralize the L & R signals after a stage of

pre-amplification, where the gain of either L or R is variable to

trim any balance errors. Then combined to mono after signals are

balanced, vertical artifacts are nulled. Even in the real world held

to at most dB difference before combining, the reduction of

distortion and vertical noise would be 93.6%, attenuated 24dB,

resulting in the worst case of 0.67% 2nd harmonic that fits the

definition of hi-fi no more spitting sibilants, in mono anyway. This 500% improvement over simply paralleling cartridge outputs

is achieved by a gain adjustment in one channel or the other by a

potentiometer, or switching up to 2dB in 1dB increments.

Do NOT simply Y the preamp outputs, causing instability from each channel effectively short-circuiting the other! Instead,

the Appendix shows a $30 yet "audiophile-grade phonograph preamplifier and DIY modifications including balance trim, mono

switch, and capacitance & resistance loads for MM/MI cartridges.

An RCA broadcast turntable with the 12in tone-arm at back aligned for 16in transcriptions, shown, the arm at right aligned for low tracking error playing 7~12in discs. A drawer of cartridges and interchangeable styli (p6) play all 33, 45, or 78.26rpm, vinyl or shellac disc recordings, stereo or monaural.

Despite advances in sound reproduction, fueled by digital

audio technology (see post script), the analog vinyl world has revived, and evolved to where the electronics some downright cheap are no longer the weak link. More than ever, it is the

8An exception is the Decca-London cartridge with lateral & vertical coils

(M-S, see footnote #5), so paralleling channels cancels vertical precisely.

electro-mechanical transducers, such as phonograph reproducers,

that require the most attention and financial resources. The

electronic modifications in the Appendix by your humble audio

engineer and skinflint will be comparable to nearly any even though you made it! No matter what others would have you

believe, the results from your cartridge and disc collection as a

result of this paper will be hard to beat. For a change, audiophiles

can be satisfied and maybe even brag about how little they spent!

And how a few hours of DIY paid off in years of enjoyment.

3. A bit of historical perspective re grooved disc recording

Instantaneous acetates, (lacquer) electrical transcriptions (ET), were the only method broadcasters had to record programs

and commercials until magnetic tape, from 1947. Cut using a

lathe, below, at 33rpm but with 2.5mil (64m) chisel, maximum recording time was 15min per side on a 16in disc. For a 30min or

longer program, two turntables were needed to segue between

discs (even ones starting inside for no abrupt change in quality).

The 16in ET format for broadcasting was borrowed from Vitaphone motion picture sound, with a turntable attached to a

projector synchronized by gearing. In the Golden Age of Radio

1930~62, stations were required to announce whether a program

was "transcribed" (recorded) so as not to mislead the public, or

fearing theyd storm the station expecting to meet celebrities! (Today, what isn't recorded?) Stores selling Victrolas (see p8)

offered recording services for church choirs and student recitals to

make albums of several 12in 78rpm discs with 4 minutes a side.

From left: 15min radio program on 16in ET; 10in 78.26rpm holds 3 min per side; a 45rpm single invented 1949 for juke boxes; from 1948 the vinyl 25+min/side 12in album. The ET is inside start. The Path 78rpm hill-and-dale (vertical) U-shaped groove requires a true spherical tip.

p6 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

Hundreds of different recording characteristics evolved by

trial & error from 1926 while new electrical recordings were still

being played on acoustic wind-ups (p8). To thwart competitors,

record labels were secretive about their formulae until consumer

enthusiasts & broadcasters won standardization in 1954 by RIAA

(Recording Industries Assn of America ) using simple RC filters,

adopting RCAs New Orthophonic characteristic [Moyer 1953].

Mono and from 1958 stereo masters are cut with inverse RIAA curve at top. Consumers play using its mirror image, below, frequency & phase response coming out flat. Disc pre-emphasis renders replay vulnerable to issues with mechanical and electrical phase and transients if HF are not equally deemphasized, including correct C-load as in 1. Not recognized, for replay-only, are the ineffective IEC 20Hz filter for infrasonics [Howard], and the Neumann curve some believe in error is required at 50kHz.

While championing standardization and higher quality for

both its consumer and broadcast businesses, RCA debuted with

much fanfare, then silently dropped Dynagroove to counteract pinch distortion of run-of-the-mill spherical needles. Speculating

they pre-distorted the audio by mixing-in inverted 2nd

harmonics,

it ruined the sound for sophisticated audiofans using elliptic styli.

Nearly all stereo discs are mastered monauralizing low frequencies below ~150Hz. While this practice reduces groove

hopping and vertical turntable rumble, it also truncates binaural

perception of interaural time difference (ITD) cues down to 90Hz

[Bose; 45Hz the author] that effect spatiality in real hearing.

The so-called Neumann Curve is ill-advised. Proponents believe in implementing a 6dB/octave rise at 50kHz to counteract

attenuation used to protect the cutter head. However, the cutting

amplifier used a 2-pole, -12dB/octave filter not only to flatten but

to turn downward any ultrasonic drive, and early on at 35kHz

instead of 50kHz! An octave lower at 25kHz, the effect of a 2-

pole filter is a fraction of a dB. So even if any VHF audio made it

through disc-making and cartridge replay, it along with any bogus

correction would be inaudible except for increased pops from radio frequency interference even when a disc is not playing!

Although less than 3% of the CD and download market, new

vinyl sales rose 6-fold 2007 to 2014, recently growing as much as downloads have fallen [The Economist]. Used vinyl shops are

responding to demand by average collectors, who like the larger

album format for visuals and liner notes. And while seniors wax

nostalgic for warm [and uncompressed!] vinyl sound, in Barnes & Nobles vinyl kiosks they are buying for teenaged grandsons. It is likely that demand for record playing equipment will follow.

Big vinyl collectors have records in the thousands, as did

radio stations, ranging in genre & format from Path hill-and-dale

(vertical) 78s requiring true spherical styli; shellac 78.26rpm SP

(standard-play) acoustically recorded before 1926 and electrically

thereafter; late 1950s vinyl 78s; SP & microgroove ETs 10~16in;

7in 45rpm singles from 1949; and from 1948, 33rpm mono LPs, both 10in and 12in. From 1958, stereo mostly at 33rpm with some audiophile releases at 45rpm. In the digital age, studies find

the longest-lived medium for data is the shellac disc, made of lac

beetles and slate dust, that lasts 200+yr just dont drop it!

Needed for grooves from any era to sound great? L, interchangeable styli organized by disc and cartridge/arm. R, "archiving" preamp with variable & preset LF turnover, HF rolloff, VLF rumble shelf for any groove from Edison cylinders to all Berliner discs, lateral or vertical mono (78s, NAB, Columbia LP (mono), RIAA +0/-dB, C-load selection, precise mono combining with stereo pickups, and variable LF vertical cancellation in stereo by author.

Subject for another paper, recovering the fidelity baked into disc recordings requires alignment of pivoted tone arms to minimize other distortion caused by tracking errors. L, assorted templates for arm pivot-to-(platter) spindle, pivot-to-stylus tip, and "offset" angle of the cartridge to orient the cantilever tangent to the groove, as was the disc cutter. R, tangential-tracking player.

Schematic of a $30 preamp with authors modifications for C-load, balance trim and improved frequency, phase & transient response in the Appendix. Beyond the scope of this paper are the merits of various preamp topologies [Holman, Lipschitz+] or preservation & restoration techniques [Copeland+].

p7 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

Appendix: How to modify an inexpensive phono preamp: C- & R-loading, balance trimming, monaural mode

Here are easy modifications for audiophiling an inexpensive Electronics-Salon A-310 phono stage. This JFET preamplifier has accurate RIAA compensation on a populated and tested printed circuit board (PCB) available for $10 including shipping (Hong Kong)

at http://stores.ebay.com/Electronics-Salon . The purpose for these modifications is to improve reproduction for both monaural and

stereo 33 or 45 vinyl disc recordings, or 78rpm issues with re-EQ, using a stereo moving magnet or moving iron pickup cartridge.

Improvements are: a) add balance correction for system channel sensitivity & gain errors of up to 2dB to reduce vertical artifacts

for mono, and improve soundstage for stereo; b) add VLF stability, reduce noise & distortion; c) add to the selection of capacitive load

values for cartridge tuning; d) & e) improve frequency response to within

p8 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

A310 PCB top as modified with 12 component substitutions, eliminating four (that remain onboard) including 1 of 2 audio carrying capacitors.

A310 PCB bottom, as modified, showing four jumpers: two by solder bridges and two by wire. Use of a precision RailSplitter eliminates DC offset in the output (less than 0.3mV) without need for a output blocking capacitor and requires only an external power adaptor of 12~35V DC.

Wiring to/from the PCB Once modifications a~e) are made on the PCB itself, additional mods are made to its signal connectors. Make these four wires/leads as short as possible, but just long enough to reach the opposite channels terminals. (Add four ground wires to the jacks, or mount them in a metal box connected to one of the PCBs GND terminals.) For RF immunity, thread a wire through a ferrite bead and loop around and through again, then connect from each input jack to its PCB IN. Each L & R OUT is wired through a 1,650 resistor to its output connector, plus a single pole switch between them for a mono laterally modulated groove. (For a hill-and-dale recording, temporarily reverse the R+ and R- clips at the cartridge.) Use the stylus appropriate for

standard groove 78rpm & transcriptions or microgroove LPs & 45s. Connect inputs using the separate tone-arm ground wire, and

remove the carts ground strap unless it is insulated from the arm. PCB documentation at http://www.audiowind.com/pdf/A-310.pdf no longer pertains select the cart loading using the table on p7. Finally, trim the phono systems balance per the steps below

How to balance your phonograph system using an A-310 phono stage as modified: C- and R-load switches A~G may be engaged

in any combination - set per manufacturers specifications for your stereo cartridge (same for L & R), taking into account the tone-arm-to-preamp wiring capacitance as in 1. Gain switches H and I may only be used individually, not both ON together. Adjust the

balance of your phonograph reproduction (stereo cartridge sensitivity difference and preamp gain error) with these steps:

1) Begin by setting both channels H/I gain switches to 1 ON, 2 OFF. While playing a mono record with the mono mode switch in stereo, observe the output levels*** and determine the difference, likely not more than 2dB. If the channels are different by less than dB, skip to step #5. (***Use meters of a recorder; or reverse wires of one cart channel, place switch in mono, and listen for the lowest level.)

2) If the systems L-channel is higher than R by more than dB, skip to step #3. However if lower, reverse each pair of inputs & outputs at the preamp PCB so that its labeled L side is in the systems R-channel, and the boards R side is in system L. Continue with steps #3 & #4, but now address channels and H & I switches of the channel written as {L} instead on the opposite side of the board labeled R, and those written {R} on the side labeled L.

3) Now for the systems {L}-channel level being higher than the {R}-channel by more than dB, flip the PCBs {L}-channel H switch OFF, I switch ON, and again observe the level difference. If less than dB, skip to step #5.

4) If the systems {L}-channel is still higher than {R} by more than dB, flip PCB {R}-channel H switch OFF, I switch ON, and observe the level difference. They should now be within less than dB skip to step #5. If not within dB, either the disc, cartridge, or your level metering is defective, and you can proceed no further until one or more are corrected. (Evaluate

metering by swapping channels and note any disparity in readings, or use one meter.)

5) Your system is balanced within less than dB, or >24dB (>94%) reduction of vertical artifacts. Switch to mono for cleaner sound of monaural recordings, or leave the switch in stereo for a finely balanced soundstage.

(Em)powering: Audiophiles have alternatives for phono stages, usually much more costly than $30 for this PCB and modifications,

plus power supply and case, if any. If not pretty enough (or expensive enough) to call attention to, mount the bare PCB under the

turntable plinth near the tonearm post, the arm wires secured directly to the boards IN terminals. (Compensate for the shorter cable capacitance, lower by ~100pF, with a higher C-load selection.) The preamp can be powered with two 9V batteries, wired through a 2-

pole switch, giving about 100hr of use per set of fresh batteries. Better is a regulated bipolar power supply to the OPA2134s max 18Vdc resulting in double the headroom, eliminating over-modulated 45s or pops from latching up the IC, requiring a power cycle. Best as shown on p7 is Analog Devices TLE2426 precision RailSplitter and almost any 12~35 volt DC wall wart in your spares box. The PCB has power filtering including RF, so the power supply may be mounted several feet away. Thus empowered, enjoy.

p9 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

Modified preamplifier performance & conclusion

The modified single-stage amplifier with RIAA equalization

offers simplicity and low cost. Its roots were first optimized by

Lipshitz in 1979, and with improving opamps, evolved like the

A-310, or the Very Simple Phono Stage (VSPS)i. In some, UHF

boost results from non-inverting op-amps minimum unity gain.

In contrast to the original unmodified A-310 (brown) with its MM gain of 30dB, the improved, flatter frequency response of the modified preamp with 39dB gain (green) agrees well with its design simulations below.

Authors circuit design simulations show ideal frequency response of the A310 as modified in the Appendix (blue) compared to its input, the inverse RIAA response (red) output fed to it by an ideal cartridge & recorded disc.

The phase response of the inverse RIAA filter used in disc mastering is baked into the LP groove (red). One objective of this paper is to realize practically a flat response both for accurate timbre and for phase error to turn out correct (

p10 Tuning phonograph reproduction by cartridge loading & channel balancing Robin Miller 2014

The author

Robin Miller is a pianist & orchestrator, Peabody-winning filmmaker, and audio engineer with more than 50 years in music recording & mixing films, television specials, and historic restorations. With Filmaker Technology he is a Patent-holder (full-sphere 3D audio reproduction) who designs, integrates, and publishes re Ambiophonics and other audio innovations.

Having every technological gadget known, the authors grandsons were mesmerized by a century-old wind-up Victrola IX playing an acoustically recorded shellac 78. What audio media will they enjoy as senior citizens?

In future revisions, this space will link discussion of any problems and

successes of DIYers who attempt this papers modifications. Lets start by adding a PCB source, and giving credit where it is due that came to

light several months after this paper was first published 9/30/2014 i VSPS (Very Simple Phono Stage) is coined by Richard Murdey of

RJM Audio - http://phonoclone.com/diy.html . Among his PCBs & kits,

ordered by emailing for a PayPal invoice, is

his VSPS at http://phonoclone.com/diy-pho5.html . The essence of this

paper proper cart loading and signal balancing is fully applicable to Murdeys VSPS, available as of this writing as a bare PCB for $15 and kit with parts for $40 a bargain. His design appears borrowed by Electronics-Salon for the A-310. Using Murdeys kit, measured results using precision, channel-matched components, will likely be better than

reworking a populated and tested A-310 that uses cheaper parts. And it

avoids unsoldering and substituting components. With his PCB alone, a

Bill of Materials can be imported into a Mouser order, incorporating any

component changes from p7. Cartridge loading & balancing selections

must be implemented externally, as in the VSPS these are not included.

Also the Allen Wright modification (altering standard RIAA response with a 6dB/octave 50kHz Neumann zero) is omitted, as advised by the author on p6 (if a unity-gain-stable IC is used), by simply inserting a

jumper instead of R3 in both channels. Murdey also prefers a polypro

cap in the output rather than a larger one in series with gain resistors that

in the authors opinion is worth it in better VLF performance.