Current Mode Electronics for the Qweak Experiment

Des RamsayUniversity of Manitoba/TRIUMF

Qweak Collaboration MeetingTRIUMF, VancouverOctober 14-15 2005

Synthetic Quartz Scintillator Bars

Shielding Wall

ToroidalMagnet Liquid Hydrogen

Target

ElectronBeam

Layout of the Qweak Experiment

Front-end electronics in extra shieldingbehind the wall

800 MHz

50 p.e.per event

x1000

50,000 eper event

6.4 A 6.4 V

VME digitalsignal integrator

1 M I-V

to DAQ

in shielding outside hall

Nature of the Current Mode Signals

QIBin 22 shot noise:

I-V Preamplifier specs

• Gain: Vout/Iin = 1 M with option of up to 10 M. Set by switches on board.

• Output: 0 to +10V. Adjustable 2V offset. Drives 130 m RG-213

• Input: 10 A range. (e.g. +1 A to -9 A with +1 V offset.) Tolerates 5m of RG-62 on input. (Noise set primarily by length of input cable.)

• Bandwidth: f3db = 30 kHz. (settles to <10-4 in 50 s)

• Density: two amplifiers per module (one per detector bar).

• Uses 5 V DC Supply. Ground fully isolated by internal DC-DC converter.

• BNC connectors. Center conductor negative on input.

• Small size for ease of shielding.

MK1 Preamplifier at TRIUMF

Offset adjust

Chan 1

Chan 2

OutIN

Chan 1 gain

Chan 2 gain

+5 V DC

10521

0.512.55

M

M

MK2 Preamplifier

• Reduced power supply noise• Switchable gains

Noise Measurements on Preamps

• Noise measured using x200, 50 kHz 5-pole amp, scope 500 kSPS• Noise referred to preamp output.• Used various lengths of RG-58C/U on input – measured capacitance.

Compare to shot noise (for 1 M) :70,000 V beam on10,000 V LED test 300 V best-case battery test

Noise in Vrms

Cin (pf) Measured Pspice

open 50 24.4 93 70 51.2 179 100 83.0 275 120 112 726 190 182

Gain (M) Noise (Vrms)Channel 1 1 110 2 205 5 470 10 825

Channel 2 5 475

MK1 -- 1 M gain MK2 -- 225 pf

• four 1 ms integrals• target bubbling not seen on few ms time scale• Tsettle should be short to minimize dead time

one spin state – (1/250) second

1 ms

t

next spin state

50 s settling time(not to scale)

Qweak “1 kHz” Integration Scheme

Existing Gzero Ion Source Signals

• signals derived from 20 MHz crystal clock.• Qweak integrator should use this clock as well• Integration triggered by MPS (is present form OK?)

• Integration time software selectable 1/300 s to 1/30 s To be set as a fixed number of samples.

• Integration triggered by external NIM signal (e.g. MPS) Triggers a selected time after leading edge and runs for pre-set number of samples.

• Clocked by 20 MHz NIM signal from ion source. Divided by pre-scale to get sample rate.

• Full differential high-impedance input provides common mode noise rejection.

• 50 kHz, 5-pole anti-aliasing filter.

• 18 bit ADC, 500 ksps.

• Four 32-bit sums per integral available to DAQ on VME. No dead time between the four periods.

• Buffered output permits reading previous integral during integration.

• Eight integrators per single width VME.

TRIUMF Current Mode VME Integrator

VME registers

• Sample period multiplier, k. Sample period = k*(1/sys_clock)• Number of samples per block, I• Number of sample blocks, m.  m = 1, 2, 4.  • Gate to Trigger Delay, n. Delay = n*(1/sys_clock). • Gate source, EXT/INT. • System Clock source, EXT/INT.

•Block integration time = k*l*(1/sys_clock).Total integration time = k*l*m*(1/sys_clock).(In the FPGA it may be simpler to do 4 sample blocks and a total at all times.)

Data Readout

• Sequence number. Shows if a measurement has been lost or skipped. • Block 1-4 sums. 32 bit. • Total sum. 32 bit. • Total number of samples actually taken -- Should equal k*l*m. • Valid data flag or new data ready flag. • Data buffer empty flag, if the data is buffered.

VME section details

Outstanding Questions

PREAMPS:• What radiation dose will they take? (Dave Mack is testing one)

• What gain range should we use on the production amps? (MK2 has 1,2,5,10 and 0.5, 1, 2.5, 5 for evaluation.)

• Where, exactly, should we mount them?

INTEGRATORS:• Are we using 1/250 s spin states, and is 50 s settling time reasonable?

• Is ion source clock 20 MHz and can we get it as a NIM signal.?

• Is MPS good as a trigger and can we get it as a NIM signal?

• approximate signal with samples• Q = (average I)(T)• band limit signal to small fraction of sampling frequency to eliminate the wiggles and kinks.• we impose an analog cutoff at 1/10 the sampling frequency

Integral From Samples (rectangular rule)

1 2 3 n…

T = nt

t

I

t

Averaging of Digitization Noise

• The 18 bit ADCs have ~0.5 LSB rms noise per sample.• This is reduced by averaging ~500 samples per integration.• This will only work if raw signal spreads over enough channels.

• Assuming equivalent noise bandwidth 47 kHz (f3db= 30 kHz) and 18 bit ADC at mid range:

condition Q rms noise before channels channels (e) integration () (FWHM)

beam ON 50,000 69 mV 1420 3339LED test 1,000 9.8 mV 201 472battery test 1 0.31 mV 6.3 15

So this is OK even for very quiet signals.

rms noise on a 1 ms integral

Condition noise (ppm)beam-ON shot noise 1120shot noise during LED tests 160shot noise during battery tests 5preamplifier noise 2 digital integrator noise 1-2

Comparison of Different Noise Sources

chan 1chan 2Input side

Output side

megohms 5 2.5 1 0.5 1 2 5 10

• Channel 1 is shown set for 1 M and channel 2 for 0.5 M

• To change gain move the set switch back towards the input side and move the one you want towards the output side. Note the gains increase away from the center.

• The offset is set for 1.0 volts when shipped, but can be changed with the offset pot

• To open the preamp for adjustment, remove the hex nuts from the OUTPUT side and remove the black screws from the INPUT side.

Changing gains and offset on the TRIUMF MK2 preamp

Qweak Collaboration Spokespersons

Bowman, J. David - Los Alamos National LaboratoryCarlini, Roger (Principal Investigator) - Thomas Jefferson National Accelerator FacilityFinn, J. Michael - College of William and MaryKowalski, Stanley - Massachusetts Institute of TechnologyPage, Shelley - University of Manitoba

Qweak Collaboration Members

Armstrong, David - College of William and MaryAverett, Todd - College of William and MaryBirchall, James - University of ManitobaBotto, Tancredi - Massachusetts Institute of TechnologyBruell, Antje - Thomas Jefferson National Accelerator FacilityChattopadhyay, Swapan - Thomas Jefferson National Accelerator FacilityDavis, Charles - TRIUMFDoornbos, J. - TRIUMFDow, Karen - Massachusetts Institute of TechnologyDunne, James - Mississippi State UniversityEnt, Rolf - Thomas Jefferson National Accelerator FacilityErler, Jens - University of MexicoFalk, Willie - University of ManitobaFarkhondeh, Manouchehr - Massachusetts Institute of TechnologyForest, Tony - Louisiana Tech UniversityFranklin, Wilbur - Massachusetts Institute of TechnologyGaskell, David - Thomas Jefferson National Accelerator FacilityGrimm, Klaus - College of William and MaryHagner, Caren - Virginia Polytechnic Inst. & State Univ.Hersman, F. W. - University of New HampshireHoltrop, Maurik - University of New HampshireJohnston, Kathleen - Louisiana Tech UniversityJones, Richard - University of ConnecticutJoo, Kyungseon - University of Connecticut

Keppel, Cynthia - Hampton UniversityKhol, Michael - Massachusetts Institute of TechnologyKorkmaz, Elie - University of Northern British Columbia Lee, Lawrence - TRIUMFLiang, Yongguang - Ohio UniversityLung, Allison - Thomas Jefferson National Accelerator FacilityMack, David - Thomas Jefferson National Accelerator FacilityMajewski, Stanislaw - Thomas Jefferson National AcceleratorMammei, Juliette - Virginia Polytechnic Inst. & State Univ.Mammei, Russell - Virginia Polytechnic Inst. & State Univ.Mitchell, Gregory - Los Alamos National LaboratoryMkrtchyan, Hamlet - Yerevan Physics InstituteMorgan, Norman - Virginia Polytechnic Inst. & State Univ.Opper, Allena - Ohio UniversityPenttila, Seppo - Los Alamos National LaboratoryPitt, Mark - Virginia Polytechnic Inst. & State Univ.Poelker, B. (Matt) - Thomas Jefferson National Accelerator FacilityPorcelli, Tracy - University of Northern British ColumbiaRamsay, William - University of ManitobaRamsey-Musolf, Michael - California Institute of TechnologyRoche, Julie - Thomas Jefferson National Accelerator FacilitySimicevic, Neven - Louisiana Tech UniversitySmith, Gregory - Thomas Jefferson National Accelerator FacilitySmith, Timothy - Dartmouth CollegeSuleiman, Riad - Massachusetts Institute of TechnologyTaylor, Simon - Massachusetts Institute of TechnologyTsentalovich, Evgeni - Massachusetts Institute of Technologyvan Oers, W.T.H. - University of ManitobaWells, Steven - Louisiana Tech UniversityWilburn, W.S. - Los Alamos National LaboratoryWood, Stephen Thomas - Jefferson National Accelerator FacilityZhu, Hongguo - University of New HampshireZorn, Carl - Thomas Jefferson National Accelerator FacilityZwart, Townsend - Massachusetts Institute of Technology

Shot Noise

QIBin 22

)( 23

dbfT2

1

one-sided shot noise, [A2]

equivalent noise bandwidth [Hz]

or

chargequantum [C]

current [A]

Example, 1 ms integration with beam on, assuming 800 MHz: • Q = 50,000 e• I = 6.4 A (800 MHz x 50,000 e)• B = 500 Hz• in = 7.2 nA rms (7.2 mV with a 1 M preamp)

Note that in 1 ms, N = 8 x 105 counts. = 1120 ppm, same as 7.2 nA/6.4 A N/1

Qweak Front End Electronics

• four 1/120 second integrals• multiples of 60 Hz cancel in sum• individual integrals show if 60 Hz (or odd harmonics) was present

one spin state – (1/30) second

(1/120)

t

next spin state

200 s settling time(not to scale)

Old “120 Hz” Qweak Integration Scheme

Block Diagram of Current Mode Electronics

Prototype Dual Preamplifier

time

current

3.6 pA(0.6 ppm p-p)( ppm)

6 A

helicity- + - + - + -

Size of Qweak Signal

• figure shows regular spin flip; in practice use + - - + or - + + -

• for 50 kHz noise bandwidth, rms shot noise is 70 nA

• on a scope the noise band would be 100,000 x the signal !

3.0zA

charge

counts

Q0

- helicity+ helicity

charge

ADCerror

+s

-s

• ADC reads S channels low below Q0 and jumps to S channels high above Q0

• This causes the measured asymmetry to depart from the real asymmetry, A0, by an amount , where is in channels.

• The DNL won’t introduce an asymmetry when none is there, it only changes an existing one.

)(0 sAA

Differential Nonlinearity (DNL) Example

Summary of Electronics Needs

DSP

eprom

external trigger

serial port (diagnostic)

channel 1

channel 2

TRIUMF E761 VME Module

Number of samples

Offset

Sample rate

Calibration

-16-bit ADC- DSP summing

- calibration mode

- channel offset and gain

- takes N samples when triggered

- adjustable delay from trigger

- delivers gate during integration

- features can be programmed

TRIUMF Parity 2 module