EEVblog® Electronics Community Forum
Electronics => Metrology => Topic started by: GigaJoe on September 17, 2025, 05:09:36 pm
-
looks like 2 version,
approx 1993 , no composite
(https://i.ibb.co/G4Q74xgL/01G.webp)
711 should be where R811, just a single 602 as integrator .....
like ... probably possible to add mod ada4077 + 2 resistors .... question: would it screw up something ( at most) , does it help with noise ( goal )
-
Can you provide some background on what this is, as well as clarify what you're wanting to do?
-
There are 2 different versions. One with and one without a 2nd OP-amp. The OPA177 has slightly better low frequency (e.g. 30 Hz to handle 1 PLC or 3 Hz for 10 PLC) noise than the OPA602. Still the integrator noise is only one noise part. One may not even see the difference.
The divider at the compund integrator is used to slow down the precision OP-amp. So the relatively fast ADA4077 or similar OPA205 may want a different divider than the OPA177. The 1 K / 100 ohm in the schematics are a bit odd choice: rather low resistance and a bit much for the relatively slow OPA177 already. Beside a little less noise the 2nd OP-amp can give a slight DNL improvement: the constant impedance at the integrator is no longer important and less error from not having it. In theory there may be corrections in the software, so that a better integrator could result in a worse result from the corrections for an not ideal integrator.
The ADC in the K2001 is quite noisy for different reasons, with no easy way out. In addition there is a gain of 0.5 for the 20 V range and the input buffer is also not super low noise.
-
thanks,
so " do not rock the boat" ....
if i understand it correct: if i mimic schematic , 1.1K basically loaded 177 , skewed DNL, but as it constant , it may hardcoded in a software ,with part i dont have ...
and if i replace 602 by let say composite , it still negligible due to overall design ...
so in overall fight with noise the way by averaging only, means ... ideally to make this noise as close as possible to uniform , but ... electronics usually produse gaussan noise ..
-
There are several reasons why the ADC is relatively noisy. One point is switching only one reference side and add the other side as on offset. This is an integral design point an nearly impossisbly to change. As a result there is quite some noise gain ( ~ 5.3) for the integrator (and similar also some other noise sources). The OPA602 is relatively low noise for an old JFET op-amp aready. This is not so much white noise, but the 1/f noise as much is about relatively low frequencies like 3 or 30 Hz for 10 or 1 PLC operation. With the somewhat complicated AZ cycle of the K2001 even lower frequencies may matter too.
I have not looked at the noise in detail - maybe if I find time I could enter the numbers (needs still a few thoughts for design detail) to my spreadsheet.
For the linearity the problem would be some hard coded (or after special factory calibration) corrections for the relatively slow OPA602. The simple 1 OP-integrator input acts like like some resistance ( 1/ (C_in*2Pi*GBW)) to the virtual ground. And this may need a correction for the fast run-down step, as it looks like this happens with a different impedance at the input. With the rather large integration capacitor they may still get away without a correction.
For a compound integrator with ADA4077 + OPA602 I would more choose some 10 K and 1 K for the divider. One only needs an effective speed ratio of about 10 (can be as low as 6 with suitable RC to ground a fast amplifier that is good driving capacitive loads). No idea why they choose so much devider with the slow OPA177 in the other version.
-
The point for possibly upgrading the integrator would not so much be noise, but the linearity. One could test for the DNL errors that are expected from the simple integrator without keeping the impedance constant. The expected error would be a saw tooth like contribution to the INL with jumps when the RU steps change. In between the differential gain would be a bit lower than ideal. One may be able to see the issue with 1 PLC conversions.
-
it can be really difficult for me to test linearity even if improved
Guess that idea of change getttin discarded ...
-
The relatively high noise level makes INL testing a bit challanging. One could give it a simply try by looking a slow drifting voltage (e.g. RC discharge). If I found it right the Run-up modulation is 4.5 or 5 kHz. With 20 V range and 1 PLC this would make some 500 mV per RU-step.
So a signal would be something like a slow RC discharge from some 1.5 to take some 15 or 30 minutes and record data at 1 PLC non AZ at the full rate.
After subtracting a fitted exponential and maybe some averaging, one may be able to spot jumps that reat about after 500 mV. I my crude estimate is right I would expect something like 5 to 10 µV.
-
second look at schematics :
added polymer 100uF for zener 8V and 6V
input short for max resolution no filter , 100 counts measured - 0.23 - 0.27 uV P-P ( 8V seems more important then 6 )
10V with filter ave-100, has around 6uV P-P ; but this most likely contributed by V source .... pattern of moving back and forth aperiodic, guess some random spikes change the picture ...
there is no circular moving of measured result as was before , it aperiodic now , most likely caused by source ....
I'm kinda happy for such result ... maybe next round to change something ... who knows ... I need a better source and logging for now ..
-
The digital offset compensation (AZ mode) should be relatively fast. It's purpose is to reduce the very low frequency noise. Ideally there should be not big jumps for the offset, but more thermal like fluctuations. My suspicion is that the offset correction does not work all that ideal in the K2001 (and also quite some other meters).
In the Allan variance curves there is usually some extra noise bump visible already with a shorted input. The somewhat complex AZ cycle with addtional steps for gain correction complicates things and there can be additional jumps in the gain (some are seen with the DMM7510 during warm up). One point that could go wrong is the wrong internal zero point (e.g. at one end of scale, or anominal zero that is a bit off) for the ADC scale, so that gain correction could carry over to the offset.
Tests with an external voltage add noise from the external reference, the meter internal references (ADC zener and long term LM399) and the gain corrections. The transition between the 2 separate references in the K2001 is a tricky part with not many things known. It looks like things are different for 1 PLC and 10 PLC (or longer). The longer than 10 PLC cases are likely just averaged from 10 PLC readings and not that interesting for the understanding.
For the start and for the ADC noise the shorted input case is more suitable to look at ADC noise. I is the easier case to understand first.
A good way to looks at the noise is the allan deviation with relatively long data sets, something like 10-50 x the longest time of interest. So this would be a few hours unter stable thermal conditions and well warmed up.
p.s.
Some of the visble jumps could well be from the LM399 ref. These commonly show popcorn noise jumps of the some 4 µV which would translate to some 6 µV with 10 V voltage.
-
The statistics is about noise. With enough averaging one gets the noise (or at least what the simple statistics thinks) down. Accuracy is a different thing and inludes effects like temperature drift, aging and the adjustment / calibration.
The test with the external 7 V reference tests the amplifier, ADC and the refences. To get only ADC and amplifier one does a noise test with a shorted input.
Especially with the K2001 it can make sense to look at the noise not just with the simple statistics, but also plot the Allan deviation. This because the K2001 is known to not have white noise as the simple statistics assumes. So the simple noise estimate is to be taken with a grain of salt.
-
The Allan deviation curve looks like there is something wrong. The slope it is going down too steep. It should initially be like the square root (or a little less) and will normally only get less steep and at some point turn up from drift.
-
I think al;so that there is something or many thinks wrong.
My guess is that this is measured with input short - but even then i expect to flatten and go up
Other option is that is used wrong function in alan python module
-
The new curve looks much more realistic. The y-scale it quite zoomed in and x and y scales thus quite different.
The maximum in the Allan dev. curve is quite common with Keithley meters, but the exact source is not known. Chances are it is linked to some background averaging in the auto zero mode that also includes gain correction.
-
some draft ....
coefficient a bit need adjustment ... but how it looks like on my non filtered data , again .. 10NPLC , no any filter
spikes cutting really well , not sure what would be next ... averaging , then stats ...
(https://i.ibb.co/Hfh691m7/Figure-1.png)
-
The noise level indeed looks high. Even the 9 µV_pp are not really good. The K2001 is known to be a bit noisy in the 20 V range. Part is from the extra divider (gain -0.5 at the amplifier) that doubles the ADC noise and also from the the buffer, that can have quite some 1/f noise. This 1/f noise could be quite different betwen units.
The extra Lf noise from the AZ mode oddity can also add to the noise.
In theory the protection with MOSFETs can also contribute a bit more than usual thermal EMF, but it should not be that much unless the temperature changes a lot.
Leakage at some switches is another possibity. For a test one could check the input current, if it is higher than normal.
-
If they have a mix of DG211 and DG411 they probably have a reason for this. If possible one would reduce the BOM and use only eine type. The DG411 is lower resistance, while the DG211 is lower leakage, capacitance and likely charge injection. So both have there pros and cons.
The 20 V range is actually not using the CMOS switches very much (only a few in the gain setting).
The positive input side gets switched to ground with relatively low resistance. I would not so much expect leakage there to be an issue. If the CMOS switches fail they tend to fail completely and not so much start leaking.
Before doing anything to the meter, it would be a good idea to first do more tests and get more data.
One point would be the noise for the 2 V and 200 mV range (with a shorted input). Here it would likely be data for 10 PLC with no extra digital filtering / averaging and than look at the Allan variance curves.
Another point is the input leakage (e.g. looks at the voltage dirft with some 10 nF of low leakage capacitor (PS or PP type) at the input. Leaking switches (specially JFETs would often also effect the input current.
Another point is looking at the supply ripple - extra ripple there could cause higher than normal noise.
AFAIK the strong points is having a 20 V high impedance range, good gain stability (low TC) and relatively high BW AC ranges. The 200 mV range may be reasonable low noise. AFAIK the K2001 is a bit noisy by design and not really made for highest precision or low noise.
It is a rather old design, that is likely only kept alive because of some (military, avionics ?) customers.
-
Hello,
in comparison to my K2000 in 10V range the std dev over 100 samples seems much too high.
on a 7V reference I typically have below 1 uV std dev.
So it should be maximum the double for a 7.5 digit instrument in the 20V range. (or even better).
I would also look if there is excessive ripple on the power supply.
with best regards
Andreas
-
From the old DMM noise thread it looks like the K2001 is quite noisy in the 20 V range also for there units. In comparison the 2 V range looks quite good.
https://www.eevblog.com/forum/metrology/dmm-adc-noise-comparison-testing-project/msg726789/#msg726789 (https://www.eevblog.com/forum/metrology/dmm-adc-noise-comparison-testing-project/msg726789/#msg726789)
So the 20 V noise measurements could be real and normal for the meter.
There may be a principle issue with the 20 V range. One possible issue could be the buffer amplifier (U341 = AD548) in combination with an AZ cycle that is not good in suppressing 1/f noise. Another weak point could be settling after switching between 0 , 20 V and 7 V reference.
It may be worth checking what AZ cycle is actually run for a 10 PLC setting in the 20 V range. A first step is looking at the amplifier output with an DSO.
-
An OPA140 (or cheaper OPA141) on an adapter board would be an option for a lower noise buffer. It still depends on the AZ cycle if this would really make a difference, but it could.
-
Continue:
NE5534 ( U802 ) on ADC that basically kinda null detector or something ...
at first did try to replace by ADA4077 , but my gut tellin me it wrong .. mostly not enough speed ... and i was correct ... gettin random number generator. wow - a bit wrong timing and everything skewed ... a good news that the point ...
2nd shot: OPA140 (correction caps intended for NE - ignored) .... and it looks much better then NE5534, it really much better as after about one hour of warmup only last digit jumping without any filter .... my guess to idealize, it need to replace LM311 by something much faster and precise.
I'm thinking ... that NE5534 has X20 amplification, in consideration of offset and noise amplification..... What happens if I bump it to let say X40
Will it increase precision of "null detection" therefore decrease the noise ?
the next thing: OPA602 , im not sure how much the noise i'm able to reduce by replace it. TI official telling OPA145 as a replacement ... so def OPA140 would be ok .... Im' thinking ..... ( composite on small add-on board with ADA4077 + OPA140, same as an original schematic, but diff opamp looks much more attractive)
for reference, schematic https://i.ibb.co/840yf0s1/ADC-K2001.png
damn ... it gettin better and better
-
The slope amplifier has 2 purposes:
1) it amplifies the signal around zero crossing and this way makes the noise and offset drift of the comparator less relevant. The comparator also sees more voltage and thus gets faster. The amplifier noise is usually better then the comparator noise. There are still other noise sources, especially the integrator (OPA602) and the 5.1 K resistor at the slope amplifier input.
2) it sets the bandwidth for the detection of the zero crossing. It is a balance between enough speed and low bandwidth to reduce the noise. For the speed there is GBW / gain and C882*R800, which should be the more relevant part. Still the ADA4077 may well be too slow and add too much delay.
I would not expect much of an improvement from more gain. The LM311 is not that slow and a gain of 20 is already quite a bit when it comes to noise. Precision of the LM311 should not matter that much, with the extra gain in front.
The OP602 (U809) is the fast part of the integrator. That op-amps noise has a limited effect for the final charge (like the NE5534) and thus mainly for short integration time (maybe 1 PLC, but mainly less). A good substitute could be OPA1641 or OPA1677 (with an adapter PCB). Higher speed could help a little with linearity. No need for a low offest (OPA140).
The more relevant OP-amp for longer integration (e.g. 10 PLC) is the slow OP-amp (U813 OPA177) it sets low frequency noise with a noise gain of some 5.3 . The current noise is hurting even more. The ADA4077 (or similar OPA205) could be a substitute here, but may need a change (larger) to R864 to compensate for a higher GBW. One would want R864/R865 >= 6 * GBW(U813) /GBW(U809) and maybe larger to get a similar value for R865/R864 * GBW (U813), even if the rest of the integrator is faster.
-
Kleinstein, thanks a lot for input ...
My current step, try to replace an old ADC with single OPA602. So i removing OPA602 and substitute it by composite OPAMP1 + OPAM2
What would be ideal types for both opamps to substitute it ?
just wonder .. i thought about 5K resistor, but more concern about 100K resistor . so calculation for 100K give me around 1 microvolt - is it right?
still why not like 1K 20K . or less load with 100K on opamp output so better recovering time ... it seems timing are crucial , as with 4077 , only 2 digits after comma was solid ... rest random flipping
-
A single OPA602 would give a poor integrator with rather high low frequency noise. I think the 1 OP-amp version used the expensive OPA627.
A single OP-amp solution could use an OPA140/OPA141/OP1641.
For a 2 OP-amp solution one could use something like ADA4077 or very similar OPA205 for the precision part (replace OPA177) and OPA1677 for the fast part. The divider between the OP-amps has 2 options:
a) aim for fast settling and a ratio like 2:1 to 3:1 (for the amplifiers above).
b) aim for settling slower than the modulation so a ratio more like 200:1 to only correct the slow part, similar to the K2001 2 OP-amp version.
Smaller resistors at the slope amplifier would cause more loading to the integrator. This would definitely be an issue with the 1 Op-amp version and even the 2 amplifier version may not like 1 K, but maybe 3.3 K.
The 100 K resistor is less of an issue as it is relative to the output and thus after the gain of 20.
The noise of the slope amplifier should be mainly an issue with short integration, less at 10 PLC.
-
actually my first post with the image ... single OPA602A :)
okay ...
so options
1 - throw out OPA602A in place OPA140 - see result ...
2 - add on board with ADA4077 ( like it) + hmm .. I have OPA1678 (a lot ... ) technically simply shorted one to ground to test an idea ... I probably go to this way if opt-1 give unsignificant result
original K2001 2 OP-amp version resistors 1K and 100 so 10:1
-
hook OPA140 ... I would like to say it about the same ....
Im still thinking that zero detector still a culprit for a noise generation, triggering a bit different time it produce timing inconsistency.
may happens due to dif. input voltage level.
LM311:
(https://i.ibb.co/LXX52Rbv/lm311.png)
some offset correction added, but i have no idea how well it, due to non symmetrical power, or natural offset correction .....
-
The zero detector is only relevant for the final charge. This is normally only a rather small part of the result and the importance changes with the integration time. The comparison of 1 PLC and 10 PLC mode would tell if the final charge (comparator+slope amplifier) or more the low frequency noise (e.g. integrator) is more relevant.
For the final rundown step there is a consistant ramp for the comparator.
The divider between the OP-amps is there to adjust the speed of the slower OP-amp. So one may have to adjust that divider when the OP-amp(s) are changed. The GBW of the slow OP-amp divided by the resistors should be at most 1/6 of the GBW of the fast OP-amp in the integrator. Otherwise the integrator can oscillate or ring excessive.
Another point is the time needed for settling. There one has the 2 options:
a) fast compared to the short reference pulses: e.g. ADA4077 and divide by 3. This kind of needs a fast 2nd OP-amp.
b) rather slow (like the OPA177 and divide by 11), so that the settling is slow compared to the modulation period.
A speed setting in between may cause INL issues. It still depends on other details how bad it would be.
The OPA140 as single OP-amp integrator should give quite some advantage over the OPA602.
The 2 OP-amp version would not be much lower noise. The low frequency noise (e.g. 3 Hz as main frequency for 10 PLC mode) of the OPA140 is comparable to the ADA4077 and only slightly higher than OPA205 noise. So the OPA140 (or OPA141) would also be a viable option for the precision OP-amp in the integrator.
There is still the resistor Johnson noise as a major noise contribution. So there are limits on how much can be done without a complete redesign of the ADC.
With the more complex AZ loop (to include the 2nd reference) there can be a higher importance of the OP-amp noise, also from rather low frequencies (e.g. down to 0.03 Hz where the hump in the Allan dev. curve is). So there could still be an advantantage from a good BJT OP-amp over the OPA140.
-
Thanks for valuable comments.
I am really curious why even faster opamps like OPA828/827 are not purposed to use for the composed integrator.
-
An even faster OP-amp could be used. The difference would not be in the noise, but linearity.
However the original OPA602 is only 6.5 MHz and the ADC should be OK with this. The relatively large integration capacitor suggests slow modulation and thus no need for super fast settling.
The OPA1655 (55 MHz) could still be an option, though overkill - maybe a thing for a faster ADC with an OPA140 for the precision OP-amp.
-
I have another question if someone can answer:
reference to the same schematic:
https://i.ibb.co/840yf0s1/ADC-K2001.png
left top corner ... reference voltage source that generate suppose to very low noise +V , -V voltage reference.
R805 , R803 - base transistors hook to positive and negative ±15 power ...
A - I don't understand why it there ...
B - it significantly decrease PSRR
so i simulated to test, even ... injecting 1V AC 1Khz in power ±15 in a simulator - that give me 8mV P-P on positive and 16 mV P-P negative (definitely as it relative)
cut off both resistors: +Vref - 3 mcV ... -Vref - 7 mcV - noise .... ands DC decreased to 20 microvolts ( insignificant )
I think i should remove resistors ...
-
I think i should remove resistors ...
I think you better not, at least not without providing an alternative way to avoid an unwanted condition in the reference circuit - an accidental reversal of the output polarity and reverse biasing of BE junctions on Q800 and Q801.
Cheers
Alex
-
R805 also helps with start up of the reference. Without it, there could be a 2nd stable point with zero reference voltage. The offset of the OP-amp U810 could be enough to make it start the right way, but this would not work for a production run. How the power rails come up could also be enough.
One could get away with chaging the negative supply of U810 to GND (still tricky which ground to use).
The negative side should follow and would be less critical one.
The ramp / step of some 6 µV could come from popcorn noise of the K2001 LM399 reference. The LM399 commonly show popcorn noise in the form of 2 possible voltage values about 4 µV appart and jump roughly (still rather random) on the minute to hour rate. Scaled to 10 V thus would give some 6 µV.
-
yeah it seem starting ....
Another problem, I think it related to wandering and basically everything ...
I cover the fan filter by paper for 20 seconds ... reading immediately start rising to 50-70 microvolts, and then back when open ....
fan a bit rattling , and probably a bit uneven airflow....
i have no idea what part is so sensitive to the airflow, consider a large plastic cover inside....
the thing that i thinking about ... most elements are emitting the heat are voltage regulators, (and transformer) and they outside that plastic cover , but may transfer heat by legs, so surrounding components affected ...
the only solution i see , modify VR TO220 to more efficient heatsinks and optimize airflow a bit ...
"one battle after another"
and yeah .. on reference part i remove resistors add diodes, and 10uf ceramic capacitor on positive and negative, it seems doesn't affect the noise.
-
There are quite some parts that can be sensitive to a change of air flow / temperature. I would not so much think of the power supply but more the ADC, the amplifier (gain of -0.5 stage), the amplifier, the input protection.
In my DMM I use a somewhat similar input protection and I can see some thermal effect - though more in the 5 µV range, though with only moderate heat . 10 x more would be quite a lot.
A question is if the air flow effect is measured with an external 10 V or under short conditions. If the test was with an extra 10 V (or similar), it would makes sense to repeat the test with a short, as this removes some of the possible sources. One could also compare to the short in the 2 V or 200 mV range. This would still habe the same protection part, but less weight (due to the gain) to the ADC and a slightly different amplifier part.
A rather sensitive part is an interesting finding. The bump in the noise at the 30 seconds is a range where thermal fluctuations could happen.
-
Thermo Effect:
i did shorted:
on every range LSD's moving up about the same rate on every range, 1000V a bit slower but still moving up when fan closed.
spent a Sunday,
fix a fan , a everything need to disassemble , and fix front panel buttons , it became very unsensitive ...
so. of course someone been there ... flux and resoldering there and there ... never noticed, but LM399H (not AH) , very close to how China reclaimed looks like, no "usual" PILLIPINES on side, and NS logo. (but in 93 it should be linear logo) I pretty certain Keithley using selected with a similar, mumbo-jumbo index like HP .... well .. well .. How can i be certain for ADC zener same , or it something different with skewed tempco and current flow, right now do not match wit minimal tempco .... I think i need to think about ..... ( i have 1 selected Lm399 from HP3478A ... )
here an experiment: it running for 2 hours and shows...
7.056,280,0 OK ...
turn off, the immediately turn on back ... and it shows same numbers? ... no
7.056,228,7
then about for 30 seconds it reach
7.056,260,0
and stay there ... -20 microvolts.
( return to 280 after 20 min )
so ADC has a zener to make +10 and -10 for ADC , then -10 inverting and using to feed LM399 to make +7V and +1.7V , capacitance divider ,not resistors. then a while ago I probe input on ADC for 20V range - each cycle of measurement has a packet of "spikes" +10V +7V (as ref) , measured value and "0" from all this it calculated actual #.##### . it also has LM75 ... Yep ... it funky calculator
and finally i did something weird .. it definitely ADC area for sensitive component , so I toss some paper towel under and around ADC, it kinda band-aid, a bit help ..
-
The HP3478 may well have no selected LM399 ref. It is only a 5.5 digit meter and the LM399 is already overkill in that class. So it could even be the left overs (still OK) when selecting good ones for 3457 / 3456.
If the LM399 ref. was replaced, than likely because it was bad. A recycled LM399 still has a chance to be OK and at least is likely well aged - so old is actually good. Finding a LM399 without the typical popcorn jumps is extremely rare and searching for one is not practical, as rare jumps are even worse. With the extra zener and smoothing the K2001 may even want an LM399 that shows the popcorn jump rather frequent.
AFAIK the ADC ref. is a low noise zener diode. The don't fail very often and I would not really expect the zener to be replaced. One may get a glimpse of the noise with an AC coupled low noise amplifier and measuring the 0.1 ... 10 Hz noise of the reference.
The warm up and reading after 30 seconds (~ 3 PPM) looks not that bad, especially given that the meter runs relatively hot. Also the 7 V reference readings may well be averaged over some time (e.g. 1 min) to make good use of the separate, lower noise reference at the ADC.
For a reading with a shorted input the reference voltage should no contribute to the thermal drift.
With the meter using the AZ mode to suppress an input offset it is odd that the temperature at the ADC should effect the zero reading (= offset). It may happen for transients (if the meter uses averaging also for the zero), but should still not give a longer time effect.
-
i need to set a data reading to analyze behavior ..
-
somehow at morning an apt approach appearing, kind of brain behaving ..
that was with replacing NE5534 , now i got a second similar and significant flaw in ADC design ..
Q812 - L7805 ( TO-92 , +5V regulator) , see link for schematics at the previous post:
I always look at as just a small voltage regulator to feed LM311 positive for adjusting TTL level ... right ? wrong ...
it give a +5 rail for comparators.
it a reference for circuit to skewed comparators level , if output high or low (3 resistors connected to pin-5 )
it a reference to trigger 3 comparators , ( see connection up to R861 and so ) , and follow up all this logic of triggering everything ....
it source of instability :
a bit warm up and everything going up drastically
generate noise that cause a random triggering - C814 , add 1000 uF - you wouldn't believed how behavior changed , it slowly up , and stay very calm just right after power up .... no any funky few hundreds microvolts jiggering ...
somehow need extremely stable +5V ( LT1021 ish ), and i have no idea where that "+5V_OUT" can be in use ....
one comparator - 5ma at positive , so 10ma + something else , I assume 20-25 mA consumption ....
ANY IDEAS WELCOME !!!!
( replace L78L05 , + it on edge of the board , - space tight )
-
According to the plan I have, the 5 V_out is a separate supply. It also supplies the FF (AC74) to sychronize the control signals. This could make the supply a bit noisy / EMI infested.
The voltage is also use to provide trigger levels to comparators, but these should not be critical (drivers for JFET control and comparator for run-up part).
A bad (e.g. too much ESR, dried out) C814 electrolytic capacitor could be an issue for the regular.
It may not be a question of the DC stablity but maybe high frequency ringing.
-
C814 - tantalum , i added 560uF polymer ...
so if i heat L78 up the output +5V decreasing down at 80 microvolts, and on display +7.XX V a bit increasing.
much more sensitive are set of resistors for comparator connected to Pin-5 , and JFET transistors - that really sensitive , especially Q808 , 9, 10
seem my jumping around L78 a bit overexcitedly - incorrect.
( but still better than before .. )
here is current status:
20V , 10NPLC , no any filter ,
vertical scale 10.0 microvolt , 100 - 10.0µV , 150 - 15.0µV
(https://i.ibb.co/TMBvY6qf/output-12.png)
I visualize 20K measurements , it 20K pixels wide PNG:
https://limewire.com/d/vgSLw#nyDosaPeRq
seems it maximum that possible to squeeze from this device, it still x20 better as it was before
mostly, i think, it a processing software, all this waves remind me an oscillating process to find equilibrium ... ; even such artefacts like below:
87474 +7.0564125E+00NVDC, +12115.014430secs, +23065rdng#, 00extchan
87473 +7.0564109E+00NVDC, +12115.564474secs, +23066rdng#, 00extchan
87472 +7.0564106E+00NVDC, +12116.113902secs, +23067rdng#, 00extchan
87471 +7.0564106E+00NVDC, +12116.663287secs, +23068rdng#, 00extchan
87470 +7.0564106E+00NVDC, +12117.212746secs, +23069rdng#, 00extchan
87469 +7.0564099E+00NVDC, +12117.762524secs, +23070rdng#, 00extchan
87451 +7.0564193E+00NVDC, +12127.652682secs, +23088rdng#, 00extchan
87450 +7.0564172E+00NVDC, +12128.202087secs, +23089rdng#, 00extchan
87449 +7.0564172E+00NVDC, +12128.751649secs, +23090rdng#, 00extchan
87448 +7.0564172E+00NVDC, +12129.300998secs, +23091rdng#, 00extchan
87447 +7.0564170E+00NVDC, +12129.850759secs, +23092rdng#, 00extchan
I think it need a cleanup of soldering mess, an do recalibration ...
-
When measuring something like 7 V, there can still be quite some noise from the meters reference. This still just a LM399, that ususally come with popcorn noise (e.g. 2 values some 4 µV apart). The low noise zener at the ADC and digital low pass filtering (the not so well understood software in the background) would still leave some waveiness with some 4 µV amplitude. That could explain much of the variations in the orange curve.
Even though the LM399 reference is likely selected and pre-aged, it may make sense to change the reference to a ADR1399, with the small extra changes needed (more current and 5 ohm 1 µF or similar snubber). After all the LM399 ref. is limiting for a 7 digit meter.
To test just the ADC part, a short at the input is the better test case. To keep things easy to compare I would stick to 10 PLC and no filtering.
Q808 and Q809 should not be that sensitive. Q810 and Q807 should be the 2 that have some effect, but should still compensate to a large part if the temperature is the same. The AZ cycle that includes a measurement of the 7 V ref. should be very good in correcting slow variations.
-
... somehow need extremely stable +5V ( LT1021 ish ), and i have no idea where that "+5V_OUT" can be in use ....
one comparator - 5ma at positive , so 10ma + something else , I assume 20-25 mA consumption ....
ANY IDEAS WELCOME !!!!
( replace L78L05 , + it on edge of the board , - space tight )
idea: disconnect the junction R828*R861 (call it junction "A") from the output of Q812. connect "A" via trimpot to emitter of Q108. adjust trimpot to give 5.xx V at "A". if that works, replace trimpot w/ resistor divider.
-
it an option to use some 5V reference (like, ad586 SOIC + capacitor ) , then reroute signal and left +5 power as is .... maybe ... it kinda a lot of meticulous work , and im not sure if it help ...
Replace LM399 I'm also skeptical , I think actual output averaged and keep in a memory as a reference. otherwise no reason to do 2 separate reference.
Shorted:
20V range , 10NPLC ,no filter: ( seems the same as with 7V )
(https://i.ibb.co/XfzS0tGk/output-13.png)
-
The noise in the shorted case is indeed not much lower. At times there is still quite some low frequency noise. It is know that most Keithey meters show such off extra low frequency noise even in AZ mode. In the Allan deviation curve one can see this as the kind of plateau from some 1 -100 seconds.
The short time noise is OK, but that extra LF noise really hurts. The "Period" of some 100-150 seconds could be something thermal, but could also by on the timescale where LM399 popcorn noise happens or the ADC gain is averaged.
Some other investigations (https://www.eevblog.com/forum/metrology/dmm-adc-noise-comparison-testing-project/msg1366061/#msg1366061 (https://www.eevblog.com/forum/metrology/dmm-adc-noise-comparison-testing-project/msg1366061/#msg1366061)) showed that triggering could make a difference and part of the LF noise could be from the way the AZ loop is doing averaging.
It makes sense to use longer time averaging for the ADC gain / the 7 V ref. measurement step. This gives some digital low pass filtering for the LM399 reference, a way to fight LF noise there. However for the simple ADC and amplifier offset averaging over a longer time is not that good. It reduces the short time noise of the zero reading, but it lower the effective chopping frequency and this way lets more 1/f noise through.
1/f noise can come from many sources (input amplifier / buffer, the OP-amp at the integrator, the inverter part of the reference amplification (U811) and also the from excess noise in resistors). The highest noise gain (look like some 5.x) would be from the precision amplifier in the integrator. On top of that there is the factor 2 from the input amplifier gain and the AZ mode taking the difference of 2 readings adds a factor of sqrt(2) for 2 readings. So referred to the input the integrator LF noise would be amplified some 15 times.
The OPA602 specs are some 0.95 µV_pp for the 0.1-10 Hz noise. The relevant frequency range could however be a bit different, but is a bit unclear with unknown extra averaging. With the noise gain this could be a good part of the observed noise.
-
there is no OPA602, it replaced by TI OPA140
consider my experience with NE5534 replacement, I would definitely say timing of that logic signals are critical part. if there an additional noise reduction, need to look at more precise timing for ZCOMP CBCOMP . I think it logically correct , as essentially ADC measuring time period, and jittering of TTL signal recalculating to the noise level ... I'm still curious C817 cap. ; what exactly they tried to compensate ... there no such cap for U804 comparator , but same compensation resistors ...
for low noise like waives I still pointed to +5 source from L7805, when i added cap it does decreased. And yes thermal fluctuation on top ...
i probably try in-place replacement for this regulator ... something (tiny board soic AD586 + OP07 + transistor )
-
The resistor to pin 5 of the LM311 comparator and the resistors around are likely there to create some hysteresis. They may also speed up things a little.
The extra capacitor C817 would make the hysteresis initially even larger and thus avoid ringing even more. For a slight speed up of the comparator one would also connect pin 6 to the positive supply.
On the downside C817 causes quite some current spike to the 5 V regulator - this could be some issue and why the extra capacitor at teh 5 V can help.
An alternative to C817 could be lowering R806 (e.g. 5 K range). This would give a larger hysteresist all the time and not just for a short time.
With an OPA140 at the integrator the integrator should be less of an issue.
The 5 V regulator has 2 properties that can matter:
1) low frequency noise
2) settling speed and tolerance to fast transients
-
another question:
integrator opamp U809: R826 (3K) connecting output (pin6) to -15V
is it something for offset \ precise measurement , or safety resistor to minimize swing on positive side , or inject -15V noise ??
i'm wonder ..
-
The resistor is likely there to make the integrator output stage run in class A mode. So avoid the output stage cross over and reduce the dependence of the GBW on the current.
If one is worried about the noise injected from the -15 V side, one could replace the resistor with a JFET constant current source ( ~ 2-4 mA).
Normally the -15 V should not be that noisy - if so check where the noise / interference comes from.
-
ah , interesting .... to improve linearity , i guess ...
I added 470 uF in parallel to every tantalum 10uF , 6 in total
0_A - before (green)
0_B - after (blue)
same 20V , 10 NPLC, no filter , shorted
(https://i.ibb.co/GfvtftMQ/output-15.png)
standard deviation:
test_m_NOISE_0_A.txt: 2.716 × 10⁻⁶
test_m_NOISE_0_B.txt: 2.378 × 10⁻⁶
so i feed 20K lines to calculate Power Spectrum Distribution
(https://i.ibb.co/X9tVNzr/output-16.png)
that basically tell me nothing new , very slow drift and interference on the frequency of measurement cycle ...
Noise density ≈ 9.8 × 10⁻⁷ V/√Hz = 0.98 µV/√Hz , basically noise floor - 1 µV/√Hz
and this one:
some bins , missing , so some value as fraction of microvolts doesn't exist, like 0.000,220,1 ; ,220,3 ; ,220,5; not really true 8.5 digit mode ... it also shows .. something , my thought not completely formed ...
(https://i.ibb.co/pBqs4jzY/1d5abe48-5f47-4c46-a534-77738d19e5e7.png)
-
It is a bit surprising that the extra 470 µF capacitors seems to help a little with low frequency noise. The frequencies are too low that it would have an effect as normal RC filter. Maybe this is more a thermal effect with capacitors acting a thermal shield. If the extra LF noise is from thermal effects, it could also be just randomly more stable airflow.
The noise spectrum shows a peak at 0.5 Hz. With a sampling speed of 2 SPS this would be 4 readings period.
The other point noting is that the noise density goes up over the whole range. If the software doing the AZ mode is done right, the kind of chopping action (alternate between zero and the signal) should suppress 1/f noise quite well. I am afraid this is a firmware thing an nothing that is easy to fix. I doubt that Keithley would fix this for the very old meter - if lucky they might for new meters (e.g. DMM6500 , 7510, K2010) that seem to have the same issue.
Ideally they would add alternative AZ modes, without averaging for the zero and maybe also without the ref. step (more gain drift, but less noise).
One might get an idea of the ADC performance if this would be fixed by looking at the 10 PLC non AZ mode, take the difference of consecutive readings and calculate the RMS noise of this (like the 1st point of the Allan deviation). This would correspond to a simple HP like AZ mode.
The missing bins in the histogram are roughly equally spaced and there is no extra hight in the neighboring bin (like expected for DNL issues). So this should reflect the actual internal resolution. After all this is not a 8 digit meter but a 7 digit one (noise wise barely).
-
thermal something ... as i said it quite sensitive to temperature, i tossed paper towel around , average wandering down to 3 microvolts up-down.
so i 'm still wondering about R826 (3K) , it simple to disconnect it and to see the result , the question is: if it affect something, how to find out .
and related: does it mean opamp THD directly affect linearity ?
-
The 3 K might effect linearity, but it is hard to tell how and how much. I looked at the effect of op-amp loading, but with the 2 OP-amp integrator. In my testest I could not see a difference (e.g. to about 0.1 ppm FS level).
It may be a bit more important with only 1 OP-amp in the integrator.
I would suspect something like an U² contribution, from the output current effecting the GBW of the OP-amp. So not the classic THD, but related.
The relation is no direct and it can depend on details of the run-up.
One could see an U² contribution as turn over error (change polarity of external reference, with good care about offsets). Not sure if this would be actually visible or still in the noise. The estimate I made for my system was 0.5 ppm FS - so barely visible with the K2001. With the larger integration capacitor and slower modulation in the K2001 the error may well be smaller than this.
-
resistor R826 - removed.
0 - noise affection, std dev 2.63 µV
guess it wrap up ...
20V range: output: -0.000,220 V ( graph: 1.2 - 2.6 µV )
Std dev: 2.63 µV ; LF noise: 1.11 µV ; HF noise: 2.24 µV
(https://i.ibb.co/LdCVcdxL/output-26.png)
2 Volt range: shorted , output: -0.000,001,9 V ( graph: 1.4 - 2.4 µV )
Std Dev ≈ 0.162 µV ; σ_LF ≈ 0.017 µV ; σ_HF ≈ 0.161 µV
(https://i.ibb.co/RprDMVg5/output-24.png)
0.2V range: output: -0.000,002,08 VDC, ( graph: 0.18 - 0.23 µV )
std dev ≈ 0.080 µV ; σ_LF ≈ 0.015 µV ; σ_HF ≈ 0.080 µV
(https://i.ibb.co/xKRSqbky/output-25.png)
0.16 x 10 = 1.6 µV , vs 2.6µV , so for 20 V almost the bottom
and 200mv 0.08 vs 0.016 ... a few times more noisy ..... i think i need to change something ....
-
seems final .. did update the previous post with result ; some cooldown , then will do a second unit ; need to think about opamp replacement for composite integrator ( 2 opamp replacement)
-
The 200 mV range noise is often dominated from the frontend noise (amplifier and resistors in the protection).
The observed 80 nV RMS for 10 PLC is actually quite good (e.g. the HP3458 is similar).
The 2 V range has a very similar input path an amplifier. This means a good part of the 160 nV RMS noise is from the front end. This is normal and not that surprising. The rest of the noise ( roughly 3/4) should be from the ADC and a littel more noise from the FB resistor in the amplifier.
It is remarkible that the 2 V noise is more than 10 x lower than the 20 V range.
This means some ( especially the LF part) of the 20 V range noise is from the front end. For the 20 V range the path is different and looks like more noisy ( U322 and U341). For U322 it would be current noise that can be a problem - the LT1007 has quite some low frequency current noise that is not good for 25 K impedance. It is however hard to find a better alternative that is also very linear.
-
yeah for 20V , i add some modification as well ... U341 replaced to OPA140 , in later version 2001 it BB OPA124; a lot caps for filters especially ±8V and ±6V , zeners was shunted by polymer caps, Q330 not so much for PSRR actually ; so not ADC only ..., and true, i walk around LT1007 - there is no much to substitute , that significantly do an improvement ...
so i measure my 7V and:
7.05647777 ± 0.19µV ( ~95% confidence)
of course it not an exact # , but before i have ±2.6µV ; improvement over an order in confidence ...
interesting note: RST command: and output starting shaking like crazy, then stabilized approximately after 2000 readings , so yes , constants do recalculating and store in the memory
-
The problem with replacing the LT1007 (U322) is that this amplifier also effects the linearity. In the 2 V and 0.2 V ranges the amplifier is used with gain and for good linearity very high loop gain is wanted. In addition the OP-amp drives quite some load (e.g. some 8 K) and thus can see thermal effects.
Already the LT1007 specs don't guarantee good linearity, but it seems to work.
There are a few options with lower noise (e.g. OPA210, OPA828 ?), but they have less loop gain.
An OPA189 or ADA4522 may work, but add AZ switching spikes and may not like the loading and show thermal effects.
The 20 V range is with some 15 K source resistance and thus sensivive to current noise.
Reducing the loading part should be possible with an extra buffer to provide most (like 90%) of the current .
-
LT1007 ( both ) replaced by OPA189
20V range: output shorted : -0.000,220 V ( graph: 1.6 - 3.0 µV )
Std dev: Std dev: 2.28 µV ; LF noise: 0.91 µV ; HF noise: 1.98 µV
(https://i.ibb.co/5X5CbHpP/output-28.png)
a little bit down ... probably LF may be a bit lower after 24h warmup , current state after 2 hours ....
I see more artefacts like:
+0.0002214
+0.0002221
+0.0002218
+0.0002218
+0.0002216
+0.0002209
+0.0002211
+0.0002214
+0.0002214
+0.0002192
will update this post tomorrow , to see noise after 24H
-
The regular correction jumps (roughly every 55 seconds) are strange. I kind of have expected such steps for the refrence voltage or overall gain, as this is also seen with the DMM7510. However with a short there should be essentially (would be ppms of µV range offset) no effect of the gain. Having such jumps also for the offset could explain quite of the LF noise. The obvious part are the jumps, but it also means the some offset part is not corrected in between and thus amplifier or ADC noise from rather low frequencies (e.g. down to 0.01 Hz) making it's way to the result.
A possible explaination for the rather high noise in the 20 V range compared to the 2 V range could be in the more tricky way to measure the ADC+amplifier gain. The 2 V range can directly get a reference voltage near 2 V (e.g. 7 V / 4 via the PWM divider). For the 20 V range I see no direct path to apply the 7 V ref directly to the input. For the gain part there is only the option to use the same resistors as gain of 1.5 and thus get some 7 V in amplifed x 1.5 to get 10.5 V to the ADC. To get the 0.5 gain one would have to subtract another 7 V measurement with gain of 1. This extra step would make the gain correction loop more complicated and possible longer to make up for added noise. The longer loop may also be used to zero averaging and thus more noise there. Alternating between a 7 V ref. and 10 V ref. could also cause the 0.5 Hz background seen before in the FFT data. The measurement before can have a small effect (e.g. dielectric absorbtion and thermal effects).
One could test this theory by looking at the signal going to the ADC with a 2nd meter or DSO over a few minutes. Ideally apply a small voltage (like 1 V to the input to see the difference between the input and a zero and the ref. voltages).
Another point to maybe look at is to plot the allan deviation for the 2 V range with a short. This may need a long set of data well after warm up. The question is if there is also the plateau up to some 100 seconds like in the 20 V range.
-
i told you it the one big wonky calculator ....
i think it final ( The device itself is the only thing left to replace) :
Noise rms ( basically std dev, in our case , the same #):
20V - 2.15µV ( and still around 3 µV wandering , before was 9µV , 3µV - still annoying )
2v - 0.23µV ( previous result optimistic , i skip last digit by mistake in the script )
0.2V - 0.16µV ( prev .. same bad)
now it seems valid as the numbers it in the same circle
waiting parts for the second one ,and measure my 7V couple days ...
addon:
drop 1h sample:
here is output gibberish: https://pastes.io/1OmTOPvV
interesting that noise calculated as 2.91 μV , so the source voltage probably 1.96 μV ( but my exp. below 1μV )
-
To compare to calculated noise, it could make sense to also record a measurement in non AZ mode (1 PLC and 10 PLC).
Especially the high frequency noise part or 1st point in the Allan dev. curve would be interesting. So no need for long records.
From my spreadsheet for ADC noise the K2001 ADC does not look great, but still quite a bit better than the AZ mode measurements.
The 1 PLC case and a little more averaging may not be that bad in comparison.
The difference to the AZ mode data would likely be mainly from 1/f noise that is poorly suppressed with too much averaging at the wrong place.
-
so you saying, I should do, let say for 6 hours, 4 measurements as combination of 1/10 PLC , AZ ON/OFF ?
-
The tests with AZ on and 10 PLC are already done.
For the non AZ mode, there is no real need for a very long test. Some 20 or 30 minutes would be enough.
-
The allan dev. curve looks wrong again. A log-log slope close to -1 is too steep to be realistic. White noise would be -0.5 and with 1/f noise one usually gets less slope and more of a plateau and often bend up later.
The curve could be the Allan variance which is square of the deviation.
-
yep...
seems correct ; i added some presentation from NIST slideshow
(https://i.ibb.co/TxrV7nFY/test-7-V-Final-long-full-madev.png)
-
The test with the 7 V source is quite complex and contains quite some noise from sources besides the ADC.
There is noise from the external reference, noise of the low noise zener at the ADC, noise from the gain measurement steps, and for longer times noise from the LM399 reference.
For the Allan dev. curve I see different ranges:
The intial blue part may be dominated by ADC noise from not too low frequencies.
The rise in noise up the the peak at some 40 seconds looks like the extra LF noise due to the poor AZ implementation. It may include some 1/f noise of the K2001 zener ref, but I would not expect it to be relevant.
The 2nd blue part looks like the range beyound the zero averaging and thus LF noise suppression from AZ mode finally working.
The lower plateau is than likely limited by reference 1/f noise. This would likely be mainly the DMM internal LM399 and not so much the external ADR1399.
The initial point looks not that bad. It it not great compared to modern DMMs, but OK to call it 7.5 digits. At around 1 second the 1 µV level is reached.
The really anoying part is the peak at some 40 s. I am afriad this is mainly a software issue, that fails to suppress 1/f noise around the ADC.
-
ok .. need more data to get ... about week or so ...
with 1 PLC i missing data , a lot , need change to work with buffer
+7.0565416E+00NVDC, +126011.551984secs, +69410rdng#, 00extchan
+7.0565392E+00NVDC, +126011.910213secs, +69428rdng#, 00extchan
10 PLC AZ OFF:
probably that hump def. AZ
(https://i.ibb.co/b5VLxRWS/test-7-V-Final-AZOFF-full-madev.png)
-
set a script to use memory ... 1 NPLC , AZ -off ; a lot of repeated data:
7.05645130;7.05645130;7.05645370;7.05645130;7.05645130;7.05645370;7.05645610;7.05645610;7.05645370;7.05645610;7.05645130;7.05645130;7.05645130;7.05645130;7.05645370;7.05645370;7.05645610
but it sequence:
+7.0564513E+00NVDC, +1.664722secs, +00050rdng#
+7.0564513E+00NVDC, +1.698283secs, +00051rdng#
+7.0564513E+00NVDC, +1.731648secs, +00052rdng#
+7.0564513E+00NVDC, +1.764260secs, +00053rdng#
+7.0564537E+00NVDC, +1.797670secs, +00054rdng#
+7.0564537E+00NVDC, +1.831163secs, +00055rdng#
controller cant keep up ??? ; doesn't make sense for me to measure like this ...
-
With 1 PLC the quantization error can become a bit more relevant. The steps get 10 x larger, while the noise is likely only about 3 x higher than with 10 PLC. This gives more often same values in a row. This is still not bad, just no full 7 digit resolution with 1 PLC.
If the datatransfer is limiting, one could skip the time stamp. The 1 PLC non AZ mode should be pretty fixed sampling rate. Even if there is a little error on the time scale this would no be so bad.
The distorted shape of the histogram for the non AZ mode comes from the drift part: initially slow and than a bit faster.
For the 10 PLC non AZ case the first point of the Allan dev curve is at nearly the same level (even a bit higher) as for the AZ mode. With normal (simple, like HP) AZ mode one would expect roughly 30% lower noise. This clearly points to an AZ mode that uses averaging - this helps for the short time, but lets the 1/f noise through up to some 40 seconds (the time used for zero averaging). Once beyond the averaging time there is the extra 40% (from usinf a 2nd measurement) noise for the the AZ mode, but also suppression of 1/f noise and thus the downward trend for the AZ mode after that.
-
so ,practically , to display a data on the dmm display , ideally, it would be setting as moving average for around 20-30 sec, or 40-60 samples ( not 100 as at maximum ) that would be a lower noise ; and for maximum precision for very stable source, collect around 1200 samples , and do processing ...
-
AFAIK the K2001 has the digital filter option the is doing a moving average. If really needed the averaging could be done later with saved data. I consider those high resolution meters mainly useful with saving the data. A human is hardly keeping up with processing the display data and is too error prone.
If in doubt one would like to look back at data.
How much averaging is wanted depends. For reading the display something like 1 update per second is usually good.
When using saved data there is no need to do much averaging in the meter, 10 or 20 PLC is aready slow enough to not get excessively large files.
If in doubt the faster data could help analysing issue in hindside.
The issue with the K2001 (and as it looks like quite some other Keithley meters), that causes quite some extra noise is how the AZ mode is done. It looks like they average the zero readings over some 30-50 seconds (could also be 60-100 readings) and only than subtract that average value from the readings. This works nice for a short time (e.g. up to 1 or 2 seconds on the allan deviation curve), but it is a desaster for the longer time scales. They really missed out on implementing the simple AZ mode (subract the last zero reading, or at most the average of the last 2) at least as a mode to choose. For some 20-200 seconds this could have reduced the noise some 4-10 times.
Averaging over quite some time (likely even several minutes) is good the reference / ADC gain measurement.
Getting low noise even with the poor AZ mode would require to really keep the 1/f noise down.
-
AZ triggering, quite clear picture:
(https://i.ibb.co/gq9Nsbh/AZ-sample.png)
--
(https://i.ibb.co/rG1y9XCn/image.png)
cover fan fo 1 minute, then open it , spike 100uV P-P:
no idea why,
(https://i.ibb.co/LzkR90Gw/Clipboard-04-23-2026-02.png)
-
Some updates... 4 days sampling my 7V reference.
Good news: the 7V is performing extremely well.
And the not-so-good (mediocre, to be precise) about 2001:
Here are the results: it ran for approximately 4 days, then I calculated a baseline, let's say an average, accounting for drift and daily day/night fluctuation. I then took the raw sanitized numbers, averaged them over 50 seconds (roughly 100 samples), and created a plot showing the difference from baseline. This makes it clear how the device is performing. Red dots = excluded samples.
https://i.ibb.co/xqtf7QnV/7-V-long-full-madev-fixed-window.png
In defense of my 7V reference — I used the 34411A for sampling, as it has one of the lowest tempcos at 5 LSD (OFF to ON state; the fan likely helps), with only a 3 µV difference between the average at the beginning and the end of the 4 days. I consider that precision, noise, etc., cannot guarantee a drift measurement accurate to 3 µV.
So from the plot I can see drift, and my script calculates it as 0.1901 ppm/24h — but it's slowly trending downward. I think it needs at least a month to stabilize. (an assumption , device sitting on the shelf for many years, due to extremely strange behavior )
As for those sudden drops and jumps — I can't attribute them to LM399 behavior, even if that effect exists, it would still be buried in the noise after averaging. So those jumps are something else. So far, those are the 2 disappointing factors: slow drift upward, and jumps ..
PS: i'm assembled some 5V source, ad586 + OP07+transistor ... still thinking how this can improve ADC , but again ... it microvolts there and there .... and it def. can .....
-
The saw tooth like signal is mainly there at the start. It gets less with time.
This part is very likely the meter internal correction of the gain, from drift of the zener diode at the ADC. The measurements of the 7 V reference are used to recalculate the ADC and amplifier gain and this is than corrented at some time. It looks like this is about every 50 seconds.
Chances are there is done together with the AZ averaging / correction (would fit the position of the noise peak with a shorted input). The amplitude seen is however a bit large to be offest drift.
The first hour of data look a bit off from a meter that is still drifting internally and it not yet warmed up. So maybe recalculate the allan dev. for data without the first 1 or 2 hours. This K2001 seams to have quite some drift before the gain corrections.
One can likely improve on this gain drift, by changing the zener current (e.g. change R854). These low noise zeners usually have a current dependent drift and with the right current one may get close to zero drift there. As is the meter is likely still warming up initially. The signal creeps up, which corresponds to the ADC ref. to drift down. More zener current should reduce the drift and thus the saw tooth signal.
So maybe try 1.2 K or 1.5 K instead of 1.62 K. The difference should be visible especially in the first half hour after turn on.
-
I actually had an idea to unsolder it and find out a current equal to precise TC == 0 ...
another wild option: 2DW232 ... but it has around 6ma at 5.4 as a zener, and extremely sharp curve, a bit left or right and it off . with zener +diode it around 6.2 V , and current to TC=0 was around 20mA in may set, even 6.2V can be not enough to make 10V and all this chain reaction happens ....
so yeah ... Zener current for jumpy correction are really good point ; lets do it :)
Upd1:
3k9 - in parallel , 1K14 resultant , I think it too much ..
saw amplitude became higher https://i.ibb.co/93BGzKBt/Clipboard-04-28-2026-02.webp
factory: 2.37mA 6.4199 V
current: 3.37mA 6.4393 V
i think i need to start from 1.5ma , and monitor behavior and increase the current ...
this is kinda decade resistor and logging together ...
will see overnight the behavior ...
btw , it looks much cleaner as the previous one
https://i.ibb.co/rnz31s3/Clipboard-04-28-2026-01.webp
In passing i hook 460uF to fan , it would be really laughing moment if it affect the noise ..
thinking: No ... this is sucks ... it a fundamental flaw in the design ... the zener voltage converting to 10V that a feed source for LM399 , in this chain changing voltage in zener due to thermal, changing lm399 reference, that suppose to be constant , and calculation get skewed ... In the current design , the easiest way would be to cut negative 10 volt from ADC, and make stable -10V.
.... or precisely adjust zener to minimize thermal runaway.
-
I agree that the design with the 2 references is odd. The logical way would be to generate the current for the LM399 from the LM399 itself. Seams to be a little more effort. On the positive side the LM399 is not that sensitive to small changes in the current. It may mess up the calibration buit not much.
I consider averaging the zero as the biggest flaw with the K2001 (and many other Keithley meters) design. With a simple AZ cycle they could get way better noise (a little higher for 10 PLC but than going down as normal and not the noise plateau up to 50 seconds).
Another odd point is that the Ohms more seem to use some 5 V (may be up to 10 V) voltage range, that would be nice to have also for voltage, but AFAIK is not available.
The point for the the extra zener diode at the ADC is low noise. This should be rather low noise (not much worst than 2DW232), but has quite some drift, especially long term drift. The LM399 is a bit noisy, but low TC and low (at least specs for it) drift.
There is also resitor drift at the ADC and amplifier that can effect the overall gain. So the extra gain measurement in the AZ loop makes some sense.
There is still the flaw, that the 50 sec averaging time are not really enough to filter out the popcorn noise - though would need even longer filtering (like 10 minutes, if the ADC gain is not fluctuating too fast), maybe as running average to limit possible steps.
The size of the saw tooth part depends on both the zener TC and temperature drift.
If the sign of the TC changes the saw tooth should also change polarity.
For the 3.4 mA case the saw tooth still has the rising ramp that should correspond to zener voltage going down and if the temperature is going up, this would correspond to a negative TC for the Zener. The zeners have a negative TC with low current and may cross over to a positive TC at higher current. So the 3.4 mA may not yet be enough. The large saw tooth amplitude could be from more temperature drift, likely not from more TC.
It may still need more (like 5 mA).
-
Hehe ... Kleinstein was right about 5mA , according to pdf for 1N4579 2mA TC ==0 ; good luck with that ... even with 4 mA it drifting up but slower ...
in my understanding zeners basically under the same manuf. process then it bin sorted ... so for 1N4580-83 standard 4mA and maximum 8mA .. yeah
PS: if we look at national semi pdf, that was initially come from .. they claim 6uV noise and 20uV maximum ... consider initial design was 87-ish , so it was a make sense to do 2 reference. Bit 100% it was sorted 399 .. and that window of uncertain noise was overstated
( i have really bad connector on ADC - take off back on - ADC overflow , clean, it back to normal , i shorted analog path across connectors, but seems even for digital it affecting timing or fronts i would say)
-
@GigaJoe: what is the tool you are using for the MADEV/ADEV??
-
https://www.eevblog.com/forum/metrology/statistical-calculation/ (https://www.eevblog.com/forum/metrology/statistical-calculation/)
-
this is day and night with zener .... it actually CDI brand 1N4577 , same was used in Keithley 197 models , i would say soldering looks like original , but it literally thermometer ...
I really say it thermometer, you need to imagine: by design the whole board covered by a plastic antistatic cover , additionally i cover ADC buy few layers of a paper towels around, additionally i findout voltage on fan 8V (not 12 , someone add resistor i think) so airflow quite minimal .... And when i cover fan - bam +50 microvolt up !!
i unsolder it and add some thermal insulation - reaction to fan gone ... it still have positive reaction ...
So a few questions again:
1 - maximum safe current for zener , I assume around 10mA - limit , power dissipation etc ...
2 - assuming I want to replace it , how long it takes to settle to reasonable level , ( as well as consequence questions: , assuming i can select 2 or 3 with very close by parameters, possible to do 2 or 3 in parallel, without any balancing - or this is a bad idea ? )
Update: as usually Im' incorrect ...
-
Up to some 10 mA should be acceptable, but that would be a Zener that is poor (out of spec) to start with.
2 Zener diodes in parallel makes little sense - it would only work with additional balancing. For higher typical current (7.5 mA specs), there are 1N827 and similar, though some of those also get zero TC with less current. E.g. I have 2 with TC zero crossing at some 3 mA.
-
Nah .. sorry, data really was good .. Reason i unsolder, zener and did thermal insulation around. But actually it TC level was extremely bad
so i hook voltmeter and decade resistor,
an actual current close to TC 0 appears in the opposite end, around - 1.4mA
final result inside: +25C to +43C , zener V 6.4XX35 - 6.4XX34 ; i got a negative but almost TC = 0
since Zener 0 it quite interesting to see calculation of autozero:
https://i.ibb.co/NdTcRW78/Clipboard-04-30-2026-02.webp
calculated noise RMS (from data sample of 7V connected):
original current 2.3mA , noise was -- 3.2uV
bumped to 4+mA -- 2.8uV ( that where i got jumpy , as P-P was around 10-12uV )
now 1.3mA -- 3.6 uV
for higher current , 1N827 , it no fact i got better noise, by spec voltage span 6.2 - 6.9, with unknown current for TC=0 , i need it alot to pick ..
-
added 10uF 250V polypropylene slug, to zener ADC, low leak , it over 50 Gohm , a bit help ....
I think it grand finale , here is result:
20V range , no filter , 7V source.
https://i.ibb.co/G4QVkpCd/voltage-deviation.webp
each sample around 2-4 microvolts jump now.
is anything at current source, to improve, for ohms ? ,or mA amplifier to opa189 ... i don't know ...
-
With U322 replaced with an OPA189, the amps ranges already have an OPA189 as first amplifier. As this is a low impedance souce, the amps ranges where already relatively good (may be even slightly better, but no big difference) in the original configuration with a rather low noise amplifier.
For the ohms source I see litte easy improvements. One might gain a little by upgrading U324 (AD707), but the difference would be small and likely not worth the effort and messing with loop stability as newer amplifiers tend to be faster.
The main issue with all the precision measurements is the averaging for the zero and thus rather higher 1/f noise. So that would be a software thing to fix.
How good is the pimped version compared to the original ?
-
ohh .. it day and night
for 20V as it would be the the weakest initially P-P noise can reach up to 150 - 200 microvolts. even if you do 100 averaging it floated to 20 microvolts up and down and quite random for a long run.
additional capacitors , especially for 8V and 6V ramps , additional for 15V , drop to approx. 40uV P-P and around 15uV with heavy averaging .... and still floating up and down randomly ...
the breakthrough happens when i throw out NE5534 ... that was awesome and i understand that 40Y.O. chips likely not keep up ...
and second - TC=0 for ADC zener - that eliminated random wandering
so on rate of the most to least influencing:
NE5535 ( null detector)
TC=0 for zener
opamp that passed 20V ; it was TL061, later version BB OP124 - it was positive replacement for 20V range
8V and 6V ramps zeners capacitors shunt , as they derivates from non stable bootstrap
15V additional caps
integrator OP602 , 2xLT1007 - they kinda OK so just 2 cents ...
ADC capacitors shunt 10uF to 680uF - minimal noise drop, 5V infuence a bit eliminated .. another 2 cents
10uF 250V big polypropylene for ADC zener - another 2 cents ( so totally 6 cents - it matter )
on top: bad oxidizes terminal switch , make noise - blindly shorted ; cold soldering on ADC connectors - randomly jumping like crazy. partially fixed, analog pins soldered some jumpers across connector.
with all this changes i think i leave it for a month running before adjustment.
ah .. forgot a bit - some resistors elimination in ADC Vref circuit - im not sure even if helps .. it may be sensitive to ohm measurement , let say, where a minimal noise level coming in adc input.
a bit update,
https://i.ibb.co/svDKRccX/yurhfgsfdhwtrgd.webp
still fan thingi not resolved , still wonder what is it .. spike are fan influence, it think it something to front end input, as i keep shut fan longer it starting to correct, like offset.
-
another problem: ( call it off ; unbelievable ... this an act of my reference .. many month ok , another battle )
(https://i.ibb.co/Kch0gLPW/7-V-err-big.png)
startup, correction ... then bang ...
a saw such thing before, i thing, at time when i bought it , as it was a while ... i think like this , occasionally
oscillation ? , or a digital part ...
-
I would not expect an issue from the digital part. Digital tends to be no so unstable and usually a big error and not just a little.
It could be some oscillation, it could be something getting to hot an going a thermal hot state. Another real possibility are contact problems or cold solder joints.
-
I had a similar problem with the built-in multimeter in my Agilent 34970A.
A ceramic capacitor turned out to be the culprit. I'm not familiar with Keithley circuitry, but it's possible a similar problem could occur here too.
-
Or a nearby amateur radio operator pushing 1kW into his antenna.. ;D
PS: the above reference is same as in the 34401A..
Btw., how did you identify that crappy capacitor??
-
yeah it my external reference 7V ... no .. idea , honestly ; hook ref. to 34411 2PLC sporadically jumping , take ref. out - works, put back ... good, .. then jumping ... take off put back on seems work ... and as it suspected ....
noise as shorted:
https://i.ibb.co/9mVJvwTb/voltage-deviation.webp
that 2-4 microvolts ; and slow movement , guess autozero
-
Btw., how did you identify that crappy capacitor??
Analysis of the device's circuitry, assumptions about possible causes, several measurements for clarification.
-
So I found my 2DW233 thingy... with the ADA4075-2, where one op-amp feeds the zener and the second is just a buffer.
It's slowly drifting down due to self-warm-up in a styrofoam box... (guess I haven't had much patience for the equilibrium point)
I switched to 10 AVE MOV, as it's usually the default — so 10 PLC and 10 MOV AVE; basically the default profile.
First wave is when the script starts — it always makes such a splash, then the device settles itself, slowly drifting down.
https://i.ibb.co/bjffc9xs/2dwsrc2.png
That's what the designers assumed would come out as the final product, I guess, in the late '80s — far from actual.
PS: We're going into season 2: modding the 2001 with 2 op-amps — chips arrive today!
-
Season 2: second unit, made in 2005 ,
so far all this caps .. LT1007 , so far as the firs one
Adjusted TC for zener, cold-hot +700 microvolts, now -20
it was background discussion about null detector opamp with Kleinstein , he won, predicting insignificance for replacement
(I'm do rigorous now, check changes .. ; how significantly was affection for the firs unit it only sense was ..)
( it BW images
Ne5534:
https://i.ibb.co/JRTDWM6q/5534.webp
TLV888:
https://i.ibb.co/hx39zRJx/888.webp
kinda 2 cents , if so .. a bit less spikes ...
my biggest disappointing are spikes: Jiggling with capacitors to find out noise source come with no result, only integrator left ..
https://i.ibb.co/8npfdbwZ/spk.webp
-
here is thing ...
the last image that "disappointing" if average it , it would be sporadic wave up and down 5 microvolts around 10 microvolts between ...
this version has standard 2-opamp composite integrator with 10:1 1K:0.1K divider ....
Let try 1:1 , 100 Ohm removed:
https://i.ibb.co/Xr4gbn4W/no100.webp
100Ohm back, and 3.3K instead 1K:
https://i.ibb.co/yntnLr92/3K.webp
would be correct to say 3K better, probably try 5K 50:1 , and what i see without 100Ohm , basically integrator instability , or oscillation ?
-
What divider ratio is needed between the 2 OP-amps in the compound integrator depends on the choice of OP-amps. The divider effectively reduces the speed (GBW) of the 1st op-amp. To avoid oscillation or long ringing the GBW ratio (with the divider) should be at least about 1:5. So a 1:4 resistor ratio for 2 equal speed OP-amps (e.g. OPA141). For less ringing one may want a bit more separation, like 1:10.
Details of the OP-amps and capacitance at the input can also effect the tendency for ringing.
What OP-amps are used ?
The wave up and down may come from averaging steps up and down in the raw data.
Are the measurements with a shorted input or with a 10 V reference ? Tests with external voltage are complicated by reference drift and possible 4 µV range size jumps of the LM399 reference.
-
no it not an averaging ...
all plots 20V with shorted input ...
now it:
ADA4077 3K3 -- 100Ohm - OPA1655 : (1K - 100 , even worse )
https://i.ibb.co/S4NKZcpJ/4077-3-K3-100-1655.webp
testing with 50 Ohm now ... it seems better , need 2 hours to settle
preliminarily, if look at trend between 2AZ , 10uV spikes still exist ... but no 30uV ..
https://i.ibb.co/213h113R/4077-3-K3-50-1655-prel.webp
speaking of LM399 jumps, I didn't observe such behavior on the previous unit at all , 5+ microvolt are clearly visible now, I'm
quite confidence, 2 reference design come up to mitigate this behavior as well, for original NS LM399 they defined 20uV max as the noise, in 80x same as this design started, probably including that spice behavior as well ... so i 'm pretty sure it averaged constant in the memory from many samples..
-
With the ADA4077 and OPA1655 combination it should be enough sepration in the GBW to be stable even without divider, or only little divider. The rather fast OPA1655 may however need local decoupling, closer to the chip than the original PCB can provide. A little extra capacitor directly at the adater board may be a good ideal. One could check for ringing with a scope at the output of the ADA4077. If ringing is an issue, an RC series element (like 100 pF+220 ohms) to ground at the integrator input may help.
Quite some of the noise looks like on a long time scale, maybe the ~ 50 seconds lengh of the full AZ loop with averaging. This way offset drift (e.g. from instable temperature) could translate to noise. The 3 K - 50 ohms example looks a lot like a saw tooth expected from a constant offset drift in the background.
Reference steps would only be visible with a significant input voltage, not with the input shorted. The jumps from the reference are scaled proportional for the input voltage divided by the 7 V ref, voltage. So with 18 V in the steps get larger, ~ 10 µV.
The extra zener reference and averaging does digitally filter the LM399 reference noise, but this is not be enough for longer steps.
Attached I have one of my measurements of a 18 V Li battery. The DMM uses a LM399 reference and the steps are from the LM399 reference. The points are the average over 16 x 1 PLC conversions (24.5 SPS). The 3rd low phase in the example is around 30 seconds and would thus not be filtered fully with averaging over some 50 seconds. With a different LM399 I have seen much slower steps of multiple minutes, though that seems not typical. I am afraid that the K2001 even with improvements could be still too noisy to clearly see such effects, though the ref. noise could be part of the measurement noise.
-
i probe by scope. nothing like hi freq oscillation.. but without divider it slowly jumping even harder ..... it really looks like high freq. oscillation in DC circuit . and second evidence are higher divider
let me try with 100 pf ... "integrator input" - are you mean between pins 2-3 of first opamp ?
yes that snapshot was early state , nonstationary , here after 7K second:
https://i.ibb.co/FqC8kTw1/4077-3-K3-50-1655-7-K-sec.webp
-
The damping element would be between pins 2 and 3 of the ADA4077.
The dependence on the divider points to something like osciallation / ringing. One slight issue can be that probing with the scope can be enough to change things.
-
wow .. amazing , Thanks , it works
1uf for both opamp , 100 ohm removed , so 3K3 left , no divider as such ...
and , i forgot about that resistor that connect 2nd opamp output to -15V ...
that the only thing may add the noise.. i don't see wicked spikes ...
preliminary (always more exited as should be):
https://i.ibb.co/q3mfrpYD/4077-3-K3-NAN-1655.webp
?: back to season-1 , would be some benefit to change OPA140 to 1655 for a single opamp version, or not so much ? better add-on board with 2-opamps .. it probably increase linearity in overall , a bit ...
-
For the single OP-amp version the OPA1655 is not that good. It has more low frequency noise. So the OPA140 is already a good choice for a single OP-amp.
The OPA140 is already quite a bit faster than the OPA602 used before.
For the very low frequencies, like from the window up to 50 seconds due to the zero averaging the BJT based OP-amps like ADA4077, OPA207, OPA205 can be slightly lower noise than the OPA140, even when one includes the current noise. However the difference is not really large.
With OPA1655 one could consider removing the resistor to the -15 V. I don't think they would need the biasing for class A. The actual precision is anyway set by the slower OP-amp that sees essentially no load.
-
so both gettin 200.3 fail for diagnostic , but even as i assume, preliminary test as "Front panel comprehensive calibration" pass even without j-fet that responsible for fast ADC. Some how if fail at first time, on open connectors tests, ( passing DC and ohms ) and i missed to write the error , and test completed second time no errors ... i saved constants as it better then before ..
But transistor need back , as the test simply hang on that 200.3, and everything freeze, and only power switch works :) ...
in term of stability ... i would say if average the result it still floating up and down on 2-3 microvolts .... but it ok ..
upd: 2nd pass "Front panel comprehensive calibration" , was error again : 363 - 200 µA gain out of spec. Then i realize improper 20V , was set as 19.998XX more effort an it finish without error ....
lab next step ... probably will do linearisation report for both ... hope wouldnt be a disaster :) as i heavy change ADC ..
so yeah that was with some filters ... and it already was good , comparing to original state ...
https://i.ibb.co/Hfh691m7/Figure-1.png
but it looks so miserable .. comparing to now:
(dont look at spikes ... especially at end , that my activity of on\off some stuff .. it picking everything )
https://i.ibb.co/vCKRZ9VS/voltage-deviation.png
(just a note , 10PLC, no filter )
-
I would like to acknowledge the tremendous amount of help provided by XDEVS.COM in getting my 2001 back to life. I appreciate it very much.
Here a link to full report made by Tin to compare a different equipment that was in his possession.
Just a note , that 2 2001 can not be referred as a regular behavior due to heavy modified.
https://github.com/tin-/linkit/tree/master/dmm9
( There was some quite curious result regarding temperature stability, but I lost the link about it.)
-
For a single OP-amp integrator I would consider the OPA140, ADA4625, OPA827 or OPA828. A BJT based type is not really working because of the current noise. So the OP27 is poor, though maybe not as bad as in the 34401 as the impedance is a little lower (some 11 K vs 15 K). The ADA4898 even more noisy.
The OPA1655 is not good with low frequency noise. 10 PLC with good AZ loop would be down to 3 Hz (with 60 Hz mains).
With the poor AZ mode the Keithly meters are somewhat sensitive to even lower frequencies (maybe 0.02 Hz - though likely less gain) can matter.
The 2 OP-amp version has some advantags when it comes to INL. So a daughterboard with 2 OP-amps, the divider and decoupling could be an option.
There are more tings that can cause an INL error than just the OP-amp at the integrator. For my ADC I had little issues with a OPA1641 as single OP-amp integrator. There was a known DNL issue due to non constant impedance, but that applies to all single OP-amp integrators.
-
that above : OPA140
in speaking about composite like:
[something] + ADA4898-1
in this case current noise not so important , or :
[something] + ADA4898-2 + ADA4898-2
i cant figureout of first precision opamp type
what would your ideal combination for composite ?
and , what would be your opinion on this:
https://zenodo.org/records/18300606
same OPA1655
-
The OPA1655 is good as the fast amplifier in a compound integrator, but it needs support for the low frequencies.
The main downside with the OPA1655 is the rather high power consumption. A powerful output stage naturally comes with higher power consumption.
So for 2 OP-amp combination one could pair it would an OPA140 (JFET) or OPA205 (BJT) or OPA387 (AZ type).
I think the linked article is missing a part for the integrator input voltage. The leave out the part from limited BW and rate of change.
I initially only looked at that half and missed on the Z_out part they are considering. Both are relevant for different time scale: the Z-out part very fast (speed of the fast OP-amp) and the GBW limit with the speed of the slower OP-amp. It depends on the parts which part is larger / more relevant.
They look at the noise for the whole frequency range and leave it to the reader which frequencies really matter. With a slow DMM this are usually the low frequencies, like < 100 Hz. Only with faster conversions (e.g. < 1 PLC) also the high frequenies (up to some 1-5 MHz) have some effect from the run-down / residual charge part. How much and which frequencies depends on the integration time and design. I consider the importance of the higher frequency noise less relevant. The K2001 uses a rather large integration capacitors and the residual charge noise may thus matter at 1 PLC.
edit:
The current noise does matter, unless you add an extra JFET buffer, like in the HP3458.
Sensible precision amplifiers are
OPA140, ADA4620 as JFET types - also high BW for very fast integrator (especially with fast OP-amp > 16 MHz)
ADA4077, OPA205 as BJT based types with low current noise and still good BW
OPA387, OPA388, AD8628 as auto zero amplifiers - likely with extra filter and maybe a 3 OP-amp combination (e.g. extra OPA1641)
The old OPA177 as used in quite some older meters is a bit on the slow side and the current noise is not that great.
Sensible fast FET based OP-amps are:
OPA197 (low power), TLV9361, OPA1641 (10 MHz)
OPA1677 (16 MHz - low cost, rel. low power)
TSB 952 (45 MHz low power)
OPA1655
OPA810 ( 70 MHz)
AD8065
TSH4631 ( AFAIK used in newer KS3446x), older ones used JFET + BJT based.
-
i kinda like OPA387 ,388 but they 5Volt supply , i'm not certain if that would be limit for me ,and space on daughter board ..
so it sound like: OPA140 - ( div 2K: 200 ) - OPA1655 ?
-
An OPA140 + OPA1655 is sensible. The GBW ratio is already close to 1:5 and thus not much divider needed. Maybe 1K:1K.
I would plan with space for a series RC to ground (like 100 ohm 100 pF - need to try best value and if needed at all).
The AZ Op-amp would need a 5 V supply and extra filtering, which usually makes them so slow that one would need a 3rd OP-amp.
-
It depends on circuit but in most cases AZ need +-2V5 instead of single 5V
-
500+125 ohms looks like a lot of load, but the OPA140 only has a low voltage ( like 4 x the offset of the OPA1655 and thus < 20 mV). With the low voltage the 600 ohms are not had at all.
-
Overall, I did try to determine the criteria for selection of op-amp and divider. In the simulator, the best combination — theoretically — would be unity-gain-stable, high bandwidth, etc., and so on... seemed great, but in some cases the result was awkward.
Looking at the output, I decided flatness of the slope would be the best measure, but it always looks flat on the synthetic simulator scope...
So I decided to look at a more pronounced criterion — how flat the output from the first amplifier is. So I simply tried a combination of whatever models I have, or added interesting ones from TI, then adjusted the divider accordingly...
I'm not sure the flat-output criterion is correct, but it seems to give the flattest slope.
here is result:
https://i.postimg.cc/Y9Zc7QB9/intewgr-2.png
and output:
https://i.postimg.cc/ZKrkk97n/intewgr-2-output.png
Spikes always exist — either that, or it's an artifact, or you're getting no spikes using a slow op-amp, but in that case the flat line wobbles, or you get damped oscillation, or total disaster... I'm not sure spikes are bad
-
The fast / output op-amp for the integrator needs to be a low bias type. The bias would add to the intergrator bias and it comes with noise.
So it has to be a FET type, not a BJT based type.
The real world integrator has some 12 K input resistance, not 500 ohm. There is also some capacitance at the integrator input (parasitic and from the switches - especially the JFETs that are off), that does effect to integrator. Ideally one has a series RC element to ground as an extra way to tune dampen the integrator response. It is not there with the K2001, but it helps quite a bit and is there with the 3458 and some other DMMs.
Looking at the output of the frist amplifier (OPA140) is good, also in real life hardware, where the scope is more limited than in the simulation.
A little overshoot is kind of normal for the OPA140 output.
In the simulation one can also look at the voltage at the integrator input. In real life this would be tricky as the voltage is small and the probe may effect things. There should be alternating positive and negative pulses that are some 50-500 ns long (depends on the speed / divider) and maybe some 5 - 50 mV peakvoltage (depends on the GBW * integration capacitor and output resistance of the fast OP-amp). Ideally this peak should not have much ringing and should be settling fast.
Some of the simpler OP-amp models don't include the open loop output resistance and one may have to add it as extra resistor after the OP-amp.
-
using FET type now
"12 K input resistance, not 500 ohm."
I remember you told me ... for sim 200Khz and 500 Ohm - it like a magnifying glass . significantly increase imperfection that would be unnoticeable ..
when such primitive model sufficient enough .. than probably add additional elements and see the reaction ..
so far 2 options:
Best: ( i think , I will continue with this one ... )
https://jumpshare.com/s/fGXgldrT0pgnfXttkvO1
2nd Best:
https://jumpshare.com/s/E2kv0EErdTNGxJQT9znP
Yep - pulses exist 1V P-P , around 5nS , 100% that high due to my 500 Ohm current - it amplify anything ... pulses circuit are first best one.
( good luck to catch with basic hobby scope )
https://cdn.imgtree.co/images/nDtE-bKY.webp
https://cdn.imgtree.co/images/J5BcSpLu.webp
"open loop output resistance" - wait a minute .... for the best comparing to 2ndbest pulses width around x5 less, with the same amplitude . Does it mean resistance are less ? , and it actually better ?
did try OPA1641 but seems model bad, it somehow work, change resistor to +5 ohm more all going sideways ...
I open to any other ideas \ combination .... ( i didn't find much positive for 3 opamp .. maybe due to need x10 effort for tuning ... still need to look at )
thing is ... it seem newer opamp models much better comparing to older existing , like i noticed 2022 -22 much better, comparing to 2010 or early , integrator circuit looks kinda difficult to simulate , using some opamps simply crashing convergence and that it ...
-
With the 500 ohm input resistor one can run into current / slew rate limits. With BJT based amplifiers like the ADA4077 there may be additional input current once the input voltage exceeds some 100 mV / 600 mV. If this is correctly modeled depends.
For the simulation there is still no need to make the drive signal extra strong - there is no noise or scope limitations in the simulation.
Real life can be a bit more tricky.
With 2 or 3 OP-amps in the loop the DC loop gain can be so high that it brings the simulation to the numerical limits.
-
A good starting point is reading the following article. It gives good explanation about fundamentals and deign criteria's used to validate the integrator.
https://assets.zyrosite.com/AwvLzKbyDqfJn2B8/22rpt02-true8digit-deliverable-d4-P1OmTBOuSeMgWcfm.pdf
-
i read this pdf, probably a few days after publication ... still puzzling about C4, Figure 53 ..
-
C4 suppress the AZ glitches of OPA388. See TI app note for it. I posted the link many times in this forum
-
The capacitors C4 and C5 are for filtering the glitches from the OPA388.
Normally such an integrator would also have an RC element, like 100 pF 100 ohms range to ground from the integrator input. It can help with ringing and the very fast part (e.g. charge injection from the switches).
The complex integrator is overkill for the Keithley 2001 with it's limiations from the front end and the rest of the ADC.
As one example the relatively low reference voltage requires an unfavorable resistor ratio ( 20 K reference and 57 K for the input) at the integrator input and this gives quite some noise gain (5.33 overall) for the integrator.
The main amplifier in the front end still has the 1/f noise from the JFET pair, that get limiting due to the odd AZ cylce with zero averaging.
There is the gain of -0.5 and thus overall a factor 10.66 to the integrator LF noise. So the integrator with AZ amplifier might make some sense, but it comes with more white noise (e.g. from the 20 K resistor in series).
The simpler version with BJT based amplifier (ADA4077 or OPA205) would not be that bad in comparison.
A good part of the observed INL can be from the JFET Q810 to switch the input path. FET switches are known to have an U² part to there switch resistance. With good Wien bridge oscillators this is usually compensated by adding half the fet voltage drop to the gate.
Chances are the U² part of the INL is largely from the FET and not the integrator itself.
-
C4 suppress the AZ glitches of OPA388. See TI app note for it. I posted the link many times in this forum
can you please specify TI app note
@Kleinstein :
Q810 - actually a good point ... but with 60K resistor how large voltage \resistance deviation need to affect such wobbles ... (btw why 57.8K ? E96 - 57.6 , Custom made ? and why so precise ? )
lets try sim AZ ..
-
the exact value for the 57.8 K should not be critical. The values effect the ADC gain a little and this is similar to variations in the reference voltage.
I don't know why to used that value - maybe they got a batch relatively cheap or left over from another meter.
Q810 is supposed to be 2N5432 as a rather low R_on switch. Like with most JFETs there can be some variations.
With 12 V input one wold get get 200 µA or 1 mV drop for 5 ohms R_on. 1 mV drop could roughly change R_on by about 100 ppm or 50 mohm. This would be roughly a 2 ppm change to the 57 K parallel with 40 K (from the zero shift). So one could expect up to some 2 ppm effect / INL error at 12 V. It would get better if they have some selected for low R_on / high threshold (both helps to lower the nonlinear effect).
-
thanks, ouch .. - about calculation ....
2N5432 metal case ... hardly remember any metal case on ADC board
here is 20KHz , 15K resistor
Best ( magnified) :
https://cdn.imgtree.co/images/cRLFPj7h.png
2nd Best ( oh mistake , this variation actually no divider, as was implemented at the second K2001 , with divider , any, it much worse , so accidently it the best case ) :
https://cdn.imgtree.co/images/bN5v84kZ.png
i think ADA4077 + OPA827 - way to go , and would be start to replacement ..
same "best" schematic tuned to lowest harmonics at unity gain, then converted to integrator:
https://cdn.imgtree.co/images/hImRIvF6.png
-
Keithelys original design calls for OPA177 and OPA602 . So this are relatively moderate speed parts.
The ADA4077 +OPA827 combination should be a goods choice, likely already overkill.
I would replace the OPA827 with the cheaper, faster and lower power OPA1655 . The offset, noise (especially the LF part) and drift of the faster amplifier would not matter.
Q810 is maked in the manual as 5432 type JFET. So it could be a plastic case variant ( ? PN5432). The type, or the equivalent J105 makes absolute sense as a low R_on switch. There are not that many types to chosse from in this range.
-
ok ... 1655
RC , 100 pF 100 ohms to ground at the integrator input
caps for ±15V
what else i forgot , seems that all ..
-
still thinking 4522 - 1655 ..
https://cdn.imgtree.co/images/rvrONSZl.png
-
AZ OP-amps are a bit tricky at the integrator. They have current spikes at the input that may couse interference with the reference switching.
The ADA4522 has relatively high current noise and there might be better AZ amplifiers (like OPA182).
Especially with an AZ amplifier one would like to have series resistors (e.g. some 22 ohms) in the supplies and separate decoupling caps to avoid interference via the supply. Ideally one would want this even with normal op-amps, but likely not needed for the linearity expected from the K2001.
-
so far things stall , i decided to touch 10nF
bought TDK 1kV , and some TW mfg. 1.2kV , 2kV , C0G of course, any basically over 1kV ..
measure leak: as i expected - fail , 30V low noise V-source what i have, and 10M cur. sense - doesn't work , cap. exist or do not exist - same thingi .. , short by paper - numbers over the roof .. so it measure something ...
but what interesting thing - decided to clean up cap. from residue , took 99.99% isopropyl bottle special for electronics .. and numbers bump around x4 more ... then use my lovely contact cleaner - back to the same - nothing .. non chlorinated brake cleaner ( super strong stuff) - nothing ... 70% isopropyl - wow bump up to 10 times .. medical grade isopropyl - a bit more than nothing .. acetone for nails - let say nothing ... acetone from construction store - x3 more that nothing ... and it has some correlation what it leaves on glass when dry up . 70% leave the most some residue.
-
If the capacitor measures super low leakage, there is no need for even more cleaning.
To measure the leakage one may consider building a kind of integrator with the capacitor to test and a super low bias OP-amp (e.g. LMC662 or similar). Than charge and watch how well it holds the charge. It still need a good switch (e.g. mechnical relay / switch) to isolate the input path.
The same integrator would than also be used to look at the capacitor dielectric absorbtion: discharge (same relay/resistor) but change voltage to integrator output with extra switch in series).
C0G capacitors are not all the same. From may limited tests they are typically about as good as PP or PS and a few are better.
-
You have to measure the leakage in vacuum and at higher temperatures.
At normal atm pressure everything on Earth is covered by at least a monomolecular layer of water.
The contaminants may be dissolved in that layer..
>:D
-
You have to measure the leakage in vacuum and at higher temperatures.
At normal atm pressure everything on Earth is covered by at least a monomolecular layer of water.
The contaminants may be dissolved in that layer..
>:D
If there is a water layer or not depends on the material and rel. humidity. A surface usually has a critical humdity at which a dense water layer forms that can caus extra leakage. Depending on the material this may be some 10% RH for a poor material or some 60% for a good isolation material.
The meter is to be used with normal air and can only rely on the self heating to keep the temperature some 20 K higher than environment and this way have roughly 1/4 the ambinet RH and thus operate near the dry end.
-
flight to orbital booked ..
with failure of measure leakage ... we moving forward to failure of DA ( Disaster Assur.. no-no Dielectric Absorption .. )
here some curious paper i believe everyone read it already:
https://nepp.nasa.gov/files/25847/2013-Taverovsky-paper-NEPPweb-MLCCsVabs-n264.pdf
so 2 Lab PSU make 120V DC .. I charge cap, discharge for 10 sec according to paper , multimeter from short to cap - nothing ... I think military has some shielded bulletproof DMM that no picking any noise, internal charges, ets on >1G input ... so fail again ..
Then i decided to use a scope - who care about resistance, i record a pulse ..
measured spike depend on discharge time - 10 sec discharging completely if something even appears lately it not detectable ... so 120V charge - 2 sec- discharge ... 5 min to smoke whiskey ,and touch by scope ...
otherwise with 10S discharge need amp
1kV TDK - average 80mV
2kV - ave 60mV
... i dont have PP to compare ( i mean MKP 4 10nF , 600-1000 V ) , if above are true it about to close to PP film: 0.01–0.02% DA ( pick from Internet .. )
and looks like such technique more less ok , but to compare apple and orange need more less precise timing , maybe amp .. so automated ..
-
C0G capacitor have very low DA and it challanging to measure. It need a meter / amplifier with very low input bias (e.g. low pA range, ideally less).
The rather slow standard test may not be practical with a small low DA capacitor. It is made for large capacitors.
I did some tests on the DA of the capacitor in the integrator. This test is using a much shorter time scale, more like the ones relevant for the use in the MS ADC. So some 100 µs dicharge and looking at the reapearing voltage over some 100 ms or so. There is no need to test different discharge times: With low DA one can just look at the difference in the voltage from the effective discharge time.
The shorter time scale makes things easier with the small capacitors ( 2nF in my case).
-
yeah need some automation ....
so i repeat test , REF - 2001M ; DUT - one_opamp_ADC 2001 ; don't need magnified glass to see a hump ..
source can't go over 13 ... others are too noisy ,need to make one over 20
34411a as a reference would be better choice for this case .. next time ..
REF DUT
00.000203 00.000200 0.000003
01.145235 01.145230 0.000005
02.024411 02.024364 0.000047
03.085980 03.085850 0.000130
04.020174 04.019970 0.000204
05.097865 05.097712 0.000153
06.135209 06.135155 0.000054
07.051455 07.051428 0.000027
08.134891 08.134891 0
09.156290 09.156280 0.000010
10.129100 10.129055 0.000045
11.121470 11.121403 0.000067
12.089108 12.089071 0.000037
13.064085 13.064056 0.000029
if it just integrator behavior , it probably solvable .. if it something deeply .. well .. )
hook that 2kV 10nf cap ,instead what it was .. , didnt made trends \stats , by naked eye- no changes ; P-P on short 20V range same stat around 15 microvolts ...
meantime im thinking to replace OPA140 on integrator by something that i have .. and see changes .... the question is what it can be , just for test .. ???
-
A 0.2 mV difference between 2 meters is a lot. It really should no be that bad. I don't think just having a 1 OP-amp integrator would cause that much of an error. There could be different effects, like contact problems at the connector.
If one is lacking a good 20 V source, the 2 V range could also be worth testing. It uses the amplifier in a different way and the performance in the 2 V range should also be good, both for low noise and good INL.
-
i think Tin did 2V and it was a differ ..
good idea to repeat it on 2V , do it later today ...
"like contact problems at the connector."
so switch back-front on that unit doesn't exist , basically soldered short .. and i disassemble it yesterday to replace fan , clean all connectors and all around ....
Q810 indeed : 2N5432 stamped 5.68 on top .. ( selected one i think ... it can go bad ... something voltage related changing condition ... i think 2V was a different shape )
yeah ...
-
ok .. that thing corrected ....
use 3456a and K2001M , 75% 3456A + 25% 2001M = V-comb
this # also floated due to , not so stable source microvolts floating , still noise of 2001 , and i so lazy to automate, so old fashion pen\paper .. while mulling, a few microvolts off ...
(frac. of µV as V-comb not rounded as shown )
V_comb DUT corrected resid
10.004765 10.004767+1.7 µV
9.000491 9.000488 −2.8 µV
8.000520 8.000515 −5.4 µV
7.000580 7.000574 −5.9 µV
6.000510 6.000509 −1.1 µV
5.000050 5.000047 −2.4 µV
4.000189 4.000187 −2.4 µV
3.000070 3.000073 +2.8 µV
2.000070 2.000072 +2.6 µV
1.000008 1.000013 +5.1 µV
0
-
going futher ... season 2 episode with editors cuts ...
LM311 - 2 comparators replaced by LM211 , forgotten thing, 211 -has much better offset comparing to 311 ..
then i replaced null detector OPA140 to SA5534 ... like lowest ... and it gettin worse , final TLV888 and x20 amplification instead x100. that give me some noise cut. I still have around 20uV P-P for couples hours run, but for a short window let say 200 measurement it 6 uV P-P instead as it was 10 let say....
As I see it: offset and propagation time are important for comparator ; noise, speed, and propagation time - important for amplifier. , so reason why at fast implementation , video amplifier and fast comparator or super fast for some 8.5D .
I just realize ... skale stretched x2 ; it 20V but an actual voltage are half .. so actual measured noise twice amplified by calculation .... I'think i reach a bottom of this...
last one: will replace LT1097 to OPA180 - want less TC and bit lower noise ... 180 good enought .. (or 188)
-
The offset of the comparator should not be that relevant. It would add a little offset to the ADC, but that is corrected anyway.
Similar the slope amplifier offset would not be that critical. It may make the rundown a little longer, but likely still fast enough.
Changing the LT1097 in the reference amplificaton to OPA180 can make some sense. I would prefer OPA207 or OPA205.
The OPA180 is likely good enough, but lower noise version could be used as well, as the circuit part is relatively low impedance.
-
Here is integrator output , opamp pin6 - Cap. , this is single opamp version, current opamp now OPA189 :
some thingi not understandable for me, so let's go:
Set: range 20V, 10PLC , input short.
(bit enchanted pictures .. )
Basically in sequence it measure something , like input , 7V ref, short, and here is sequence.
Bottom marks is when integrated cap. shorted
P00: https://i.ibb.co/b5pmxHrS/00.png
shorted long time, we measure something differ:
P02: https://i.ibb.co/fzg84Xrt/02.png
or short time, measure the same:
P04: https://i.ibb.co/fY5Q8X20/04.png
or twice:
P06: https://i.ibb.co/93ShkC8M/06.png
Hope someone may comment better ...
not clear why it measure like this:
P08: https://i.ibb.co/kLXwNm5/08.png
or like that:
P09: https://i.ibb.co/m5kK2zNS/09.png
P09 - probably scope picking noise , that 5 spikes on horizontal line, but it correlated to steps ( why this steps ? ) on rising edge, and not exist on falling edge. - about 160kHz , and do not exist in certain time, so maybe it not a noise .. bottom flat line P02
-
The rather complex sequence in the 10 PLC AZ case is interesting. It look like there are 2 different zero readings. Also looks like 10 PLC for the actual input and 5 PLC for the Zeros and 7 V ref.
The pattern in P8 is relatively easy to explain. One can get zero from alternating positive and negative. This what happens most of the time and gives the higher frequency. In the center one has a pattern with 2 steps positive and 2 steps negative which also give zero, just with a lower frequecy. This can happen from the delay of reading the comparator to switching the reference. Depending on the start charge one can transition to that pattern when ater 1 step the charge is close to zero. I see the same with my ADC and I don't consider it bad.
The short ripples in P9 with the rising slope part are the extra switching steps to get a fixed number of switch operations. They are missing for the down part, as the down part is only one modulation step long. So switching is before and after the falling slope. With the opposite sign input voltage one would see the extra steps with the falling part, but not the riging part.
The sharp 500 mV spikes up and down are likely from probing or the ground. At least there should not be such spikes are the integrator. The switching part can cause some issues with digital ground current and this way it can make a difference where the probe ground is connected.
These short peaks are sometimes picked up in the first picture and sometimes not, as a kind of aliasing / undersampling.
-
There is more ....
Ground point i use "1" , and according to P00, capacitor reset to "0" so waves never cross "0" it always above it .
( second CH scope, I didn't look at precise setting, so it basically for display not measure, as an output - TTL level)
Output UB03 "ZCOMP" signal:
P10: https://i.ibb.co/gMDztxC8/P10-UB103.png
Or I measure it wrong. but it triggering only when capacitor shorted and it basically zero voltage.
UB04, signal "CBCOMP":
crossing virtual "0"
P11: https://i.ibb.co/DD4R4YFq/UB104.png
I don't really follow how it actually work. Major effort toward to to know when integrated capacitor shorted. and that , according to schematic are critical.
and "CBCOMP" - roughly let it go ...
-
The k2001 has separate comparators for the run-up (CBcomp) and run-down (Zcomp). This is used to have to run-up always on the positive side and this way get the rundown always start from the positive side. This simplifies the run-down, but need a larger (like 2 x) integration capacitor, as half the integrator range is unused.
-
opa182 finally arrive, i did small add-on board, 182 +1655, a decided to turn it standalone , input are func. gen. 50K resistor.
Due to gen offset integrator output tend to move to rail , depend on offset polarity ...
Now: divider between opamp i intentionally set - NONE.
Even with that, the picture drastically differ comparing to simulator - no any spikes on triangle. 182 - generate spikes only when integrator output move to rail close enough, it starting limit output and 182 generate spikes of correction. otherwise - solid line ( i may not tune well was around 1V/div, as it always touch the rail when static, only change offset and it drifting to opp. dir. then you able to see solid line).
So major question about divider between:
Most likely the overall stability enforced by large 10nF integrating capacitor .... Should i leave it -NONE , or for good sake add, let say: 2K : 0.5K adding stability ? ( technically I can build non inv. composite with G=1, as a separate circuit and play with divider, but my guess it a diff. topology, as the result) , or try very small integrating cap 50pF , and see the result ?
Upd.. screw it .. replacement installed ... I think i reach bottom line for everything ... about the same noise level for 20V.
btw ... replace LT1097 in Vref. to OPA180 MSOP - small thingy , even try to find difference in PDF - no words... so replaced , ON, measured OK , OFF , assembled. set to test - Turn ON - dead ... so one opamp (or both) shot during transition ON or OFF ... *&#$^$&$ !!! - replace it to SOIC8 OPA188 .. still alive...
at the current state all but one (OPA140 ) choppers , device became almost immune to temperature , closing fan produce a small spike of 20uV (vs 150 as it was) around an hour cold-hot to settle around '0' , 15min - close to zero ... Of course we have this and that ... but in overall ... it day and night ...
and as I'm deep in that rabbit hole .... I thinking to replace U324 AD707J ( Current gen. for Ohm) -> OPA188 ; U328 AD707J ( LM399) to probably on safe side OPA205A , or chopper - OPA188 , mostly for TC
and probably that the end ..
-
The large integration capacitor is needed, as the FB modulation is slow and works only in the positive side during run-up.
Changing the AD707 in the ohms source would be tricky, as the same OP-amp is also used for very small currents and these would suffer from a chopper amplifier with high current noise.
The OP-amp at the lm399 ref. should not make very difference. It still ads to the reference noise. So the OPA205 should work, but is overkill.
-
so yeah ... it shows numbers ... then jump then back .. will work on divider .. maybe somethin else ...
ohm source -- damn right ... I forgot it really make low current .. it seems OPA205A ( ihave A!) to ohms and leave 399 as it ( but thing is ... it also reference for the current source .. and of cours they save a few cents using a lower grade )
Upd:
play a bit with divider ... stable around 1K:150 . but still some sense it jumpy .. the add 300pf to 150 resistor, and looks like that sense disappear ..
Then i decided to do this: And i don't know is it right ? it over 1Mhz , does it affect something ?
(https://imgtree.co/direct/_PD89fuU.png)
-
Episode 18 ... replacing 1 opamp integrator with a 2 opamp version ...
The divider above kinda doesn't make sense ... obviously. And whatever the simulator produces - hardly applies to ..
OPA182 - doesn't work well either ... speed and frequency don't do well for a stable result, after a few iterations I switch to OPA205A
And here the divider comes as a major player for achieving a stable result:
Original, as from main schematic, 1K:100
So after some iterations I got 5K:511 (I have 300, 400, 511, 670, 820, and some 1K 10 ppm SMD so some variations can be done)
It works more or less, making around close to 4+ uV RMS noise ... and moving from 5K up or down .. makes the result worse .. even without moving much ... like 5K2 or 4K8 - result became worse.,
So kinda a 10:1 magic ratio .. and I need to stick to it.
Like 6K:511 - total disaster ..
( start with 10PLC, 20V range , shorted input)
https://imgtree.co/direct/qWoSJ0Sv.png
Similarly, 4K:511
Then I move down: 4K, 3K, 2K with the same ratio .. where 4K, 3K give a bit worse result, 2K:200 - gives a very good result of 2.5 uV noise ..
1K:100 - therefore makes close to 4 uV noise ...
I have no clue of the reason for 10:1, even such result of going up or down by resistance ladder, but practically it is there ..
I have to involve Mechanical Mind (MM) for calculations, where it compares patterns, calculates noise per each 12-20 hour run, fragmentarily, like comparing specific time frame selections in a few runs and diagnosing which is better or worse ..
MM found out an interesting pattern that exists in measurements: 7 consecutive readings represent a wave up and down as average, and included it as part of the determination less - better, comparing 1, 3,5,10, 30 min fragments as less "wobbling" result etc...
So currently I do 500:50 which probably would be a bit better than 2K:200 .. ideally 2.1 uV , so P-P would be in 10 microvolt frame approximately , will see ..
it seems a second unit i need to review from that point of view , where no divider at all .... as well try ADA4077, instead opa205 , and maybe OPA187 ( regardless of elevated noise) ..
-
Normally the divider between the 2 OP-amps at a compound integrator should not be that critical. There can be a minimal divider for the circuit to start oscillation. With the rather fast OPA1655 and rather slow OPA205 or OPA182 it still has a good change to work with a direct link. A 1:1 divider should already be plenty. The exact limit depends on OP-amp details and integrator loading, so no exact general limits.
The divider is relevant for the AC compensation and stability, but absolutely needs no stable resitors. It needs changes by 10s or % to get any visible change - especially if not close to the stability edge. So any resistor grade should be good enough.
The 1:10 divider is already well on the slow side for the combination of OP-amps. For best speed it would be more in the range of no divider at all to a 1:1 resistor ratio.
The OPA205 is very similar to the ADA4077, kind of made as an 1:1 alternative. So the change should not make much difference.
Normally the divider ratio should not effect the noise, if at all they may be an effect on the INL of the integrator is ringing too much, or an OP-amp reaching it's slew rate limit. So the point to check is more using the scope at the integrator output or the output of the slower OP-amp. This is where on can judge on ringing and possible oscillation.
The choice of the slow OP-amp can effect the noise a little - so here OPA205, OPA182 or OPA145 would be slightly different, especially with the 1/f noise.
The curve shown shows quite some low frequency noise. Like some thermal oscillations and not using the AZ mode. I would not expect this part to depend on the ADC details. There may be something else wrong - a poor connector contact, supply hum or maybe the OPA1655 oscillating from insuficient decoupling.
-
here is something 205 output ..
I tellin right away, it in edge of sense , board upside-down , and diff to reach, ground non ideal as well
time of exact measure - unknown , just random shots , probe x1 (for x10 - nothing ) , and a single, otherwise it just mess ..
https://limewire.com/d/AOS1i#Tq9qNQSRpq
-
*snip*
Limewire , that is a name I have not seen in about 18 years...
-
The OPA205 output shows much of the expected signal. The square wave with some extra wiggles as expected for the run-up. There are however quite some extra spikes. This makes it messy. I don't think the extra spikes are all from poor probing as they are at odd / random times. So the meter may have an issue with a not so clean supply or the decoupling.
-
So the meter may have an issue with a not so clean supply or the decoupling.
that is my fight for a while additional caps everywhere , on top that ADC board itself has 6x680uF polymer, and 8-10 of on main board there and there. adc itself as addon board , 1st opamp 22ohm +10uF and 2nd 5ohm +10Uf ,and 10uF across ramps for each opamp ..
So.. Sorry, miss shot.
it seems i understand now ... lower resistor in divider decrease overall noise , as jumps between samples. below around 150 ohm it a trend to decrease .. currently 50 ... and upper resistor as ratio control let say authority of change - too low (like 1:1 ratio) i'm getting sudden spikes - as eventually noise , too low (12:1) "woobling" the result slowly floating up in down randomly in 5-10 microvolts interval with a few - to 10ish mins period..
so ... for example if i do no divider ... it works .. but generate random spikes , and quite noise. direct and 50ohm shunt ... a bit lower noise + spikes.
1K:50 (20:1) - despite quite low noise ,and now any spikes, random wandering
so i'm collecting stats for diff ratio as most optimal
If .. for example i like to filter ADC input , like in 34401 200pF + 100 ohm - an output gettin crazy and "0" up and down as there is no tomorrow .. so it doesn't work at all ..
-
there is schematic to replace 2 opamp version ..
https://imgtree.co/direct/2UPvJOk0.png ( https://i.postimg.cc/sxbsptqm/2UPv-JOk0.png)
the final version 205 switched to opa187 , as 205 a bit .. like 4-5 faster as need .. considering nature of 187 so far i dont see residue ..
well. i kinda hate Automatic Control System course in university , especially Feedback and Dynamic systems and Controls, and that fire back quite well in this case.
so divider between opamp became critical as for system in overall , not exact as separate for specific part ...
like example:
you have this behavior , more less ok .. https://imgtree.co/direct/VfL1ehwK.png (https://i.postimg.cc/v841VM6T/04-opa187-50-300.png)
and then simply change resistor in divider to 50 ohm more and getttin this: https://imgtree.co/direct/39Vivr86.png ( https://i.postimg.cc/YSX9vyrq/04-opa187-100-300.png)
or this: https://imgtree.co/direct/IkDrWaat.png ( https://i.postimg.cc/63yQnLTt/04-opa187-100-1000.png )
I have no explanation , as it appears random islands of good bad mediocre divider ratio
so i simply log variation of hella lot of samples, an throw all in AI to find the best , with condition lowest momentum noise but priority of minimal wobbling in 1 3 5 10 15 min range. So noise bad but if system in overall wandering up and down - it even worse....
and speaking of noise .. it appears that noise injected in adc ( surprise, surprise :) ) and that appears as pattern in output voltage in most 0.28Hz with 2nd 3rd harmonics 0.56 0.84 .. and that noise added , generate 7 samples pattern that about 20% of total ... So best averaging for example not 10 .. 20 but X7
A second thing .. decreasing divder ratio decrease noise , but increase wobbling, it starting to flow in a few minutes pattern ...
adding shield to ADC , partially to analog part cut another 30% and some nasty spikes ... so the result:
20V range, input short, 10NPLC, no filter, 4 hour warm-up
https://imgtree.co/direct/COfOS5Ll.png ( https://i.postimg.cc/nV9zLk6M/COf-OS5Ll.png )
I dont think I can squeeze more , maybe a bit for another few month ..
-
Q810 is supposed to be 2N5432 as a rather low R_on switch. Like with most JFETs there can be some variations.
With 12 V input one would get 200 µA or 1 mV drop for 5 ohms R_on. 1 mV drop could roughly change R_on by about 100 ppm or 50 mohm. This would be roughly a 2 ppm change to the 57 K parallel with 40 K (from the zero shift). So one could expect up to some 2 ppm effect / INL error at 12 V. It would get better if they have some selected for low R_on / high threshold (both helps to lower the nonlinear effect).
I am currently repairing a 2001 and tracked the fault to the ADC, fault symptom is complete lockup when accessing the ADC on startup.
I wont go into detail here and will give a full photoshoot in the main 2001 repair thread.
5V was slightly down at the ADC and the OPTOs seemed ok but the data is just slow bursts and does not look right.
Swapping the ADC from another working 2001 I have restored operation and she passed all tests, beauty.
Now I found Q810 with its gate and drain never soldered when I first started on this meter and thought that was strange for a flow soldered board.
No evidence anyone has been in here at all so just soldered them and moved onto caps, cleaning etc.
Once i new it was the ADC I fired up the thermal again which led me straight to Q810, ASIC was warmer than the good board also.
Q810 Drain to source measures about 4.5 ohm hmmm well thats stuffed.
Any recommendations on a replacement for the obsolete Q810 PN5432 N Channel Switch?
I have MMBFJ108 SMD and J109 TO-92's on order however, as is always the case, they are not as good as the original RDSon wise.
Plan to shoehorn a 2N4393 in there temporarily from an old HP meter board to see if it fires up fixes it.
-
The MMBFJ108 would be a good replacement. Fairchild AN-6609 given process 58 for both the PN5432 and J108 (the THT version, that may be hard to get). R_on would still scatter a little with the threshold.
For a start the 4393 (or slightly better 4391 / 4392) should also make the ADC work, just with some extra TC (e.g. 5 ppm/K range) and maybe INL (could be a few ppm). So the exact replacement would only matter for performance, not for having the ADC run at all.
4.5 ohms drain to source could be normal. That is in the valid R_on range.
-
The MMBFJ108 would be a good replacement. Fairchild AN-6609 given process 58 for both the PN5432 and J108 (the THT version, that may be hard to get). R_on would still scatter a little with the threshold.
For a start the 4393 (or slightly better 4391 / 4392) should also make the ADC work, just with some extra TC (e.g. 5 ppm/K range) and maybe INL (could be a few ppm). So the exact replacement would only matter for performance, not for having the ADC run at all.
4.5 ohms drain to source could be normal. That is in the valid R_on range.
Thanks Kleinstein.
The 4393 didn't help and suspect there is more mischief afoot, some of the in circuit measurements on the other JFETS don't look good.
Good point that it could be valid as the gate was floating when I measured DS but I doubt it, was a constant 4.5R both ways across it all the time.
I havent done much with JFETS, never had the need so excuse my ignorance.
Hooked up a test circuit and am unable to get Q810 to turn off completely, measured 0.7V G to DS
I pulled Q809 and it performed well in the test circuit, measured 0.6V G to DS.
I can see the CPLD pulsing part of the circuit every second or so like it is trying to settle it maybe?
The fact the CPLD is warm is not good when the other one that is good does not get warm at all makes me think it may be shot.
[attach=1]
This is the good ADC with the cool as CPLD in the upper left.
[attach=2]
This is the dodgy ADC, CPLD visibly warm on the left
U802 the NE5534 and Q801 run quite warm on both ADC boards equally.
The 5V being pulled down a bit also doesn't bode well for it.
What could have done this but not effect anything else on the analog board, dodgy rails from bad caps maybe?
Have a feeling it wont end well.
-
Additional photos of the ADC's I should have added
[attach=1]
[attach=2]
-
It is normal that the ne5534 gets warm - they just have a rather high supply current.
The ASCI (CPLD) temperatures could be different from different versions (newer may use less power).
The transistor that looks a bit warms is Q801, which is a driver for the reference. It could be normal to get a bit warm with some 3.3+3+0.5 mA and some 5 V.
To see how much of the ADC is working one could look at the integrator output.
-
I figured the NE5534 temp is normal, I have worked with them before.
The thermal cam and second board allow for very quick comparisons.
As to the processor I also theorized that an older one might not be as efficient.
The faulty board has an Altera EPM5128LC with a sticker on it for the version while the newer working board has a custom print on its CPLD.
I am new to complex ADC's and Integrators so I will keep reading, learning and poking around.
Again having the working unit makes for easy comparisons.
Update: Swapping the CPLD between boards revealed it is fine.
It just draws quite a bit more current, so I am now certain it is one of the op amps or FETS that is shot.
Meter started up without the CPLD fitted to the dodgy ADC so guessing values are miles out as such it keeps waiting for stable levels.
-
@Wombat:
what on the device screen ? "buffer overflow" or "----------- " like no data ?
you have a ADC , old type - 1 opamp, 2nd newer - 2 opamps , witch one toast ?
" could look at the integrator output." - opamp otput pin, should be triangular wave approx +3v from ground as median.
@Kleinstein:
back to my topic:
I'm kinda not certain .....
During some testing, I lost a 10 µF cap across ±15 V and got wild system behavior. I'm questioning whether the second opamp (OPA1655) is on the edge—is it reason to keep it and add ceramic caps around to avoid oscillation, or replace it with something less fast?
I hooked an OPA172 as A2 for a test. The combination OPA187 + OPA172 works, no divider ,no 10uf cap across, but the first one (OPA187) isn't fast enough (noted ..) . At the same time, for amp-1 , options like OPA205 and OPA182 are stable to some degree (still jumpy) under very aggressive dividers like 15:1. 680:50 , ( i think slew rate a critical )
So I'm wondering: should I have OPA1655 as A2, as the 34401 kind of does well with AD711, which looks bleak compared to 1655?
something that faster than OPA187 ... and something that slower OPA1655 ... ideally no divider ( it visible increase noise , i guess by amplifying on divider , not like Johnson noise )
ADA4077 + OPA140 - ?
OPA196 + OPA140 - ? ( i may order 191 , if 196 work well ... )
196 + 192 ? ( pretty much need some divider )
PS... 3rd opamp as follower at output to isolate integrator ?
-
@Wombat:
what on the device screen ? "buffer overflow" or "----------- " like no data ?
you have a ADC , old type - 1 opamp, 2nd newer - 2 opamps , witch one toast ?
" could look at the integrator output." - opamp otput pin, should be triangular wave approx +3v from ground as median.
Thanks GigaJoe it just stays on "----------" If i press Auto the auto indicator turns off and then nothing further.
If a random button is held on power up you get a different random upper key function and nothing else.
Funny I didnt even notice as it went back in after the photos, it is the single op amp version that is sick.
Yes I probed the OPA602 output and it sits at about 8.6V with a bit of a pulse coming out.
Its input was just a minor ripple so something is up here.
Pulled C825 and get a constant 16-17mV over a 10M resistor across the voltage range so seems leaky.
With it out I can see a good positive to negative transition here.
I do not have anything good enough on hand for C825 so placed an order including a lot of the OP Amps you guys mentioned.
Want to avoid doing anything to the working ADC so not pulling parts from it.
I have worked my way across the circuit from the left and as far as I can tell all the references and CB logic switching are ok.
The LM339 reference is good and the lower first 2 (a&b) check out however I am sus on it so will replace it.
There is no activity from CBCOMP or ZCOMP
Not much happening on the NE5534 output
The center fet switching circuit has done my head in a bit and do not quite understand it but still sus on the other fets I haven't tested.
I do not want to just replace heaps of stuff as some has been specifically selected.
Is there a circuit for the single op amp version by chance?
-
yeah "--------" no integrator activity but logic should be alive ...
basically to test integrator , you may replace it by any, let say TL081, hope you have one.
then you will see triangle, but device will start random numbers noise in the same time ....
so you can try , selftest , it may pass or throw some error's due to TL081 ..
C825 - you may replace by any ceramic for test , as the target to make it running, later on get a good one.
circuit for single same as dual except op177 in integrator -that all
-
For the broken ADC, a rather fixed 8.6 V at the integrator output would cause no action at the comparator.
In this case the FB signal (R835/R836) should be most of the time positive with some GND pulses.
For the 2 OP-amp integrator there are many options that should work.
A2 should be fast, ideally low open loop Z-out, but no need for precision. In my ADCs I have tested / used so far: TLE2071,OPA172,OPA1641 and they all worked similar. Other candidates would be OPA197 (rel. low power, but this may come with weak output), OPA1677 (SOT23-5 not ideal), OPA1655 (high BW, but high power consumption). The offset does not matter (A1 can compensate > 100 mV) and thus no need for an OPA140 or OPA192. The AD711 in the 34401 looks indeed a bit slow.
A1 should be low drift / low LF noise and ideally not very slow (e.g. >= 1 MHz GBW). The OPA177 in the original K2001 2 OP-amp version is a bit slow, but seems to work good enough. The main candidates I see are OPA140 (or related OPA1641, OPA145), OPA205 (or similar ADA4077), OPA182, OPA387 (with +-2 V supply).
I am still confused that the divider has so much effect on noise - it should not. Much of extra noise looks jumpy, not like normal noise. The divider would effect the stability / ringing of the 2 OP-amp integrator. Strong ringing could cause INL/DNL errors and maybe a little extra noise.
An an extra follower would add delay and complicate things. I would more consider replacing R826 with a current mirror.
-
R826 - gone, i think i specific to original chip . If i introduce it , device make large waves to few millivolts on shorted input ...
so ... no cap across A2 - bad behave, only specific range of divider otherwise - bad behave, work with very slow A1 or bad behave ...
everything lead to edge stable A2 and, i thing, a reason why i have islands of bad\good divider, and sensitivity to A1 ....
but .. ADA4077 + OPA1655 work on second one with more less OK, caps across rails, it oscillate if no caps- so stability again close to borderline. yeah ..
i can not decrease integrating cap, as it hardcoded in the code ... 10nF comparing to 400pF or 1nF it quite a bit different req. to A2 ... I think current output are matter ...
btw, OPA1677 appears in SOIC8 , and new OPA1688 - design for low impedance load ....
speaking of noise ... it not a thermal noise , i think it combination of pick everything inside + digital mgm. on adc board , something else ... and no divider increase wandering ... but i cant tell now as it , I think, side effect of stability ..
here is 187 + 172 , no divider, output over few hours:
https://i.ibb.co/kgkzBtDc/17-OPA187-OPA172-300-000-P2.png
1000 samples:
https://i.ibb.co/xSY5gFcd/17-OPA187-OPA172-300-000-P2-1000.png
( it actually not so bad .. )
... maybe 200pF -100ohm would work now on input , before it throws the chaos
-
.
basically to test integrator , you may replace it by any, let say TL081, hope you have one.
then you will see triangle, but device will start random numbers noise in the same time ....
so you can try , selftest , it may pass or throw some error's due to TL081 ..
C825 - you may replace by any ceramic for test
Nope I don't have a TL081 but I do have a TL071 that I have had since 1985, I know it had been breadboarded when I was younger.
This is why you keep everything as you may need it one day but 40 years? Can't believe it still works. Found CA3030's too wow.
Soldered it just in the holes for easier removal and left the MKS 10nF with worse leakage than the original in there.
That was it and 2001 No.1 is alive :clap:
The OPA602 was shot but why, died of old age?
Thankyou guys.
It is quite ironic that you are discussing the exact section of circuit I have issues with at the exact time I am investigating it.
I have ordered OPA827AIDGKR and OPA1655DR (added to orders quickly while reading you guys) which do you recommend I use long term?
For C825 I plan to use FKP3F021003F00KI00 but also got a few metalized polys in Kemet Wima and TDK.
Surprised it worked with the double leakage MKS I have bodged in there and the original would no doubt work too.
Just the one error 200.3 A/D Fast TS Circuits Incorrect, to be expected
Probably that MKS cap wouldn't help and I still have that dodgy 4393 FET in there for Q810 also
Last 2 digits bouncing around badly but it is measuring.
Now have a lot of parts coming I do not need
Entire repair blog over here https://www.eevblog.com/forum/testgear/restoration-glory-of-keithley-2001-dmm/900/ (https://www.eevblog.com/forum/testgear/restoration-glory-of-keithley-2001-dmm/900/)
-
The fast OP-amp need good decoupling / like the capacitor directly at the OP-amp supply. The OPA1655 wants them with short leads - the ones on the ADC PCB may already be a bit far away. A slower OP, like the OPA172 is more forgiving, but still
That is normal and not a bad sign for the integrator.
The OPA602 could have died from heat cycles or ESD.
For a replacement in the 1 OP-amp version, the OPA827 is much better than the OPA1655. The low frequency noise does matter and here the OPA1655 is not great. A OPA140 / 1641 would be another alternative (still better than the original OPA602).
Capacitor leakage would effect the linearity. Similar the dielectric absorbtion of a MKS capacitor would cause linearity erros, that can well be visible.
The suitable capacitor types are PP, PS and good C0G MLCCs. In my tests (2.2 nF) they were roughly similar - with a TDK MLCC quite a bit better.
The fast ADC error may be one of the JFETs for the fine slope - AFAIK the fast conversions use a different fine slope.
-
last 2 digits random - that a normal behavior of 2001 :), reason why i didn't use it, till starting to mod it...
200.3 A/D Fast TS - it fast integration circuit fet with 800k, generate current instead 4M, it running only on the fastest settings, 10ms integration, and not on 11ms, - absolute useless thingi ... you can try fastest mode, i'm pretty sure it will work , that 200.3 - likely false positive.
currently i'm not sure if OPA1655 are ideal for dual opamp as the second one ... and 100% not for a single. For a single opamp integrator i had OPA140 works, but cant confirm 100% as linearity was toasted due to my mistake ... i did just in place replacement , and no R826 , you can play to bias it, but my result was awful , agree on OPA827 , and again .. R826 probably need removed or adjusted.
for integration cap i use something like this:
https://www.digikey.com/en/products/detail/kemet/CKC21C103JEGACAUTO/11696189 (https://www.digikey.com/en/products/detail/kemet/CKC21C103JEGACAUTO/11696189)
look for 10NF 1000 - 2000V C0G 2220
or you can throw money to:
https://www.digikey.com/en/products/detail/knowles-syfer/404021K00103JQRAF9LM/16615124 (https://www.digikey.com/en/products/detail/knowles-syfer/404021K00103JQRAF9LM/16615124)
-
@Kleinstein
"The low frequency noise does matter and here "
I'm not arguing about it, the thing is ... see on my chart ... the best what i have , ignoring wandering was around 10-12μV P-P or around 3μV RMS, in that situation it 4 nV/√Hz , or 8 seems doesn't matter. Like in theory yes, in practice - see trends .. But frankly ... OPA205A make the best level of noise, and heavy low wandering ever. need a compromise between this this and that ... for this specific device ...
-
There is no need to use extra high voltage capacitors. More moderate 40 V / 100 V types should be good enough. The PP types however naturally often come as some 400 V or higher.
The OPA205 is good in the 2 OP-amp integrator(with a suitable fast OP-amp), but not as a single OP-amp.
There is also still noise from the amplifier (especially with original LT1007 in the 20 V range) and the JFETs from the input buffer.
Overall the K2001 is not low noise. Modification can improve somewhat on the noise, but there remains the poorly working AZ mode, that fails to suppress 1/f noise above some 0.01 Hz.
-
i think 1kV or higher a safe bet , due to low frequency of integration, high capacitor, and primitive implementation ... like better safe than sorry ... I use 2kV personally ..
-
Error 200.3 was caused by that 4393 FET, is passing all tests now
It obviously wasn't switching enough and the original Q810 is ok.
Tacked the original Q810 and C825 back in and plugged it back in without extensions.
It has settled down quite a bit but not as good as 2001 No.2
-
@Wombat
do you have an option to hook something and get a plot ?
-
Depends what you mean, what am I monitoring and over what period. TC?
I have plenty of meters but if we are talking 6.5 digit I havent gotten right into the pants of my 34401A yet but have the lead.
There is a full list of my current stash of gear here
https://www.eevblog.com/forum/buysellwanted/wttfs-(au)-datron-wavetek-4600-autocal-transconductance-amplifier/ (https://www.eevblog.com/forum/buysellwanted/wttfs-(au)-datron-wavetek-4600-autocal-transconductance-amplifier/)
I have a GPIB to USB lead but have done nothing with GPIB as yet.
If you can link to how and which software I would be happy to contribute
-
to get data over gpib and make the trend what actually it measuring
-
Sorry I hijacked your thread guys.
After running shorted tests with the shield back on I have concluded that there is 4 - 5 times more noise than 2001 No.2 although it's not great either.
Can the standard noise level be improved as it seems counter productive for a 7.5 digit meter.
@GigaJoe: Let me know what to do and I will do what I can, this interests me.
As to the cap choice yeah nah to a $20 SMD cap, nice one Joe maybe for that fancy custom one you are building.
From what I gleaned a polypropylene only cap is preferred for an integrator however a metalized version will do but not as well.
Full through hole polys are harder to get but the FK3P fits that requirement even if only 250V.
Whilst a higher voltage cap will generally have lower leakage at some point you hit the wall of diminishing returns.
But I am most likely talking out of my...
What are the chances of designing an adapter board that drops into U809 and modifies a single version into a dual op amp integrator easily.
Or would this have no chance due to parasitic interference etc.
OK it sounds like it is not worth considering as the gains are not worth the issues it could create.
When you say low level calibration is that beyond the ability of most calibration labs?
-
An adapter to U809 for a 2 OP-amp solution should be possible. Parasitic capacitance should not be an issue. The integrator input already has quite some parasitic capacitance and other meters even add some extra capacitance to ground there. Ideally one would want a seprate ground connection for decoupling of the extra OP-amp.
The good point of the 2001 is more the 20 V range and a relatively low TC, not low noise.
There are a few points to improve on the noise - so not all from one poor part, but several weak points: For the 20 V range this includes U341 (AD548), U322 (LT1007) and U809 (OPA602) in the 1 OP-amp integrator version (the 2 OP-amp version should already be a bit lower noise, even with the OPA177). The 2 V range would be less / hardly effected by U341 and U322 - so the 2 V range may be performing better and would be worth extra tests.
Another point could be adjusting the zener current on the ADC module to have lower TC there, for less reference noise.
A big contribution to poor noise in AZ mode is the way how the firmware caclulates the result in AZ mode. It is not suppressing 1/f noise well. In the non AZ mode the performance is not that bad compared to other meters. A fix would need an improved / fixed firmware, which I doubt we will see from Keithley for the old meter.
-
changing opamp it a fraction of work, suppress rails, and especially as they using non-filtered bootstrap to feed differential fet input, in assumption it has infinite PSRR - it hilarious. zener TC - another "feature"
but in the end still ... maximum u get it like this:
https://i.ibb.co/8gMdNx3x/voltage-actual-centered.png 20V shorted ,no filter
with 14 average, it will fluctuate , ±5µV - about ... im trying to squeeze it to ±3µV - and call it a day
probably draw on schematic would be most useful
but you need to read data and collect plots as only the way to see behavior , second you need very stable reference to measure for a few days and see if device not drifted or jumping or simply oscillating and shows Mars weather ..
then keep in mind : total redo need a low level calibration, 2001 has regular artifact calibration, need precise 2 voltage , and 2 resistance to calibrate everything , changing internal thing, need low level calibration as well.
lastly - there is chance you may trash INL , as i accidently did , without precise measurement , hard to detect.
-
The idea of a boostrapped supply workes nice for the CMRR with the input FETs and U341 (AD548). It's not infinite CMRR, but quite some improvement (e.g. 60 dB and a bit more for the JFETs). The PSRR should be good enough with U341 - but needs some care with zeners (some are pretty noisy).
The supplies for U341 and the JFETs are linked, which complicates the circuit quite a bit.
For linearity the cross over distortion of U341 might still be an issue. This might cause some a small step near zero in the 20 V range. It may however bit tricky to see error on the order of 1 µV in the 20 V range with all the noise.
-
" boostrapped supply workes nice "
not sure how nice it works , don't want to argue , but in my practice that 4 caps drop down noise almost 3 times , especially nasty peaks up to 200 microvolts P-P .. U341 as AD548 - never exist such type , one unit TL061 , other unit , not remember OP124 or something (but it was ANALOG) .
https://i.ibb.co/PZm01fD5/caps.webp (https://i.ibb.co/PZm01fD5/caps.webp)
and many tantalum capacitor on the back board in parallel to bypass caps ... that alone drop down noise to around 50 μV , and 10-12μV wobbling due to zener.
how it was: https://www.eevblog.com/forum/metrology/k2001-adc/msg6084777/#msg6084777 (https://www.eevblog.com/forum/metrology/k2001-adc/msg6084777/#msg6084777)
After that was a second wave of opamp replacement , major U341 TL061 -> OPA140 , noise decreased a bit became more even .. , replacement LT1007 -> OPA189 - almost non measurable, but couple μV there and there .. , of course zener adjustment eliminate waves and wandering ... , couple more caps :)
and last wave adc , itself a bit there and there : LT1097 -> OPA188 , LM311 -> LM211 , NE5534 -> TL888 (OPA140) + ampl 5K -> 511 , and another resistor. additional bypass caps, 5V that over divider generate reference voltage floor for comparators 680uF , of course couple for the digital part .. again few μV there and there ..
and final most headache 1 opamp integrator to 2 opamps second unit ADA4077 no divider OPA1655. first still in progress (second device more noisy btw ...)
that the story about ...
BTW, replacement NE5534 -> OPA140 made some magic in noise that i can not reproduce, potentially, i assume, that particular NE5534 wasn't good , it has to be fast as replace to ADA4077 made it RND generator. for purists NE5534 ->SA5534A , i pick TL888 .
-
So the LCSC order turned up with a Ti OPA827AIDGKR for U809 and no I didn't notice it was a VSSOP-8 when I ordered it.
[attach=1]
[attach=2]
Luckily I have a reflow oven as there is no way I could solder that, along with some dual type 8pin DIP Ali boards fits nicely.
Added a machined pin IC socket as I intend to try some of your dual op-amp configurations.
Small signal diode legs soldered as pins for the DIP PCB are just perfect for the socket.
C825 is still the original as Fedex are stuffing around with the Mouser order.
[attach=3]
20V range shorted no filter
[attach=4]
Ref nulled to compare the bounce of around 10-15uV
You can tell No.2 has had a bit of use based on how bright the last digit is.
I knew No.1 had low use so the ADC must have failed very early on.
Now K2001 No.1 (Single OPAMP repaired ADC) Left - stabilizes within 10 minutes or less but always has the offset as shown, calibration, zero issue?
K2001 No.2 (Dual OPAMP ADC Worked) Right - Takes a good 30 minutes to settle down and really goes nuts during the warm up but settles to low values.
Both eventually bounce around by about the same amount but with the offset shown.
I haven't even isolated the SOP-8 tracks on the back of the adapter PCB and it seems ok.
Both Units have the shield fitted as it makes a huge difference and the cases sitting on top (clip leaded to the frame, also makes a difference) for thermals and shielding.
K2001 No.2 seems very sensitive to thermals which explains the weirdness on warmup, maybe zener issues also.
Yes I know eye-balling it isnt very scientific but I need an AR488 interface, wasnt able to work out how to log with a keysight GPIB adapter easily.
Still plan to replace C825 when they come and after testing them all, still think the FKP3 will be best.
So as I have OPA1655's and more Ali boards I want to build up some dual opamp trials as Giga-Joe has.
Maybe we can come up with an adapter board.
I wouldnt mind doing some circuits, only found one of yours Joe.
Last combo seemed to be ADA4077 + OPA1655 (Instead of OPA827) is this still the consensus?
Can we do some suggestion lists for both single and dual opamp combos, I lost track several pages back.
Alternatives to the OPA827 would be good as they are pricey.
Whats the latest results Joe?
-
for dual:
ADA4077 + OPA1655 , kinda work , i think not optimal
for single:
OPA205A + OPA197 ; ideal ... i think OPA205A + OPA1655 but i have difficulties to make it stable so OPA197
here last: so noise around 7-12µV and wobbling , i give up ...
https://i.ibb.co/ns91Nqg3/voltage-actual-centered.png
so average x7 (NOT 10 ) give a quite reasonable output of few µV floating
ADC a fraction of whole thingi , before screw-up ADC , add capacitors for filters ( make sure it not start ringing )
replace U341 ( for 20v ramge)
adjust current to zener TC=0 on adc board.
may replace LT1007
and hook GPIB to get data for analyzing .... then try ADC ...
"GK" -TI usually VSSOP-8
-
Good candidates for a single OP-amp version would be OPA827, OPA828 and OPA140 (and similar cheaper ones) and ADA4620 (similar amplifier from AD). They should be better than the original OPA602, as they are low LF noise and higher BW.
The LF noise matters for the noise and the speed can matter for the INL.
For my ADC I have / had good success with OPA145, OPA141 or OPA1641 (the OPA140 should work as well, just a bit more expensive) for the precision OP-amp. As a BJT option (less 1/f noise) the ADA4077 and OPA205 should be rather similar with a slight edge for the OPA205.
A BJT amplifier may want a fast OP-amp with low R_out to keep the excursions to below some 25 mV, where a BJT input may start to become nonlinear. It may help to have a little capacitance (ggf. with some series R) to ground at the integrator input for the super faster transient part.
In theory a AZ OP-amp (like OPA182) could also work, but may want some extra filtering to reduce the effect of switching spikes.
I tried these with TLE2071, OPA172, OPA1641 for the fast amplifier (all around 10 MHz GBW). The main parameters are GBW and open loop R_out. Other options would be:
OPA197 : 10 MHz low power, but has higher R_out - could be an issue, especially with BJT based amplifier.
OPA1677 : cheap 16 MHz, 2 mA supply, R_out looks good
TSB951 : 50 MHz, rel. low supply (2.2 mA typ)
OPA1655 : 55 MHz, but a bit high supply current ( 3.9 mA typ)
OPA810 : 70 MHz - but 27 V max supply. R_out looks good.
THS4631 : 210 MHz - AFAIK used in some newer KS3446x, but high supply current
For the capacitor I measured the lowest dielectric absorbtion with a TDK C0G MLC capacitor - about 1/5 the DA of PP / PS film and Kemet and Samung C0G I tested.
-
a bit practical comment about .
going to fast second opamp may simply caused instability due to non optimal original board layers,
OPA1655 - definitely on edge and oscillated when rail capacitors not in place. I see it as a bad sign of instability... So higher may do not getting any benefits but headache. 10nF , and 30Khz around - do not required something fast , in my opinion ..
Regardless of opamps .. somehow it extremely sensitive to divider, as thevenin resistance and divider coeff. a 50 ohm difference and everything screwed ... it getting absolute instable, or increase high frequency noise , or decrease HF noise but increasing wobbling a minutes period of instability. so blindly replace opamp that a serous gambling ..
OPA187 680 - 300 (ground) - OPA1655
OPA205 300 - 150(ground) - OPA197
that 2 options more less balance between everything....
probably OPA205 + OPA1641 would be interesting to try .. but i really don't see how it improve . unless it affect scale linearity make it much better .. i dont know ...
THS4631 look very interesting with integrator cap in 100-200pF and half megahertz integration -def. not for this appl.
-
A fast OP-amp should have a decoupling capacitor directly on the adapter PCB and not just rely on the original PCB for decoupling.
I agree that with the 10 nF integration capacitor one would not need high high GBW to keep the input voltage small. The point of a reasonable high GBW is that the 2 OP-amp integrator is a bit slower in settling than the 1 OP-amp integrator. The GBW of the fast OP-amp determines how much divider is needed and this way how fast the integrator settles after switching. The point here is not 30 kHz modulation frequency, but the length of the short pulses in the pattern. The difference could be in the linearity, not in the noise.
All proposed solution here are still faster than the OPA602 + OPA177 in the 2 OP-amp version of the K2001.
I see the OPA205 + OPA197 combination a bit critical, because of the relatively large open loop output impedance (some 350 ohm) of the OPA197. This can lead to voltage spikes that exceed the linear range of a BJT based OP-amp (like OPA205). This might cause some nonlinearity. The simple FB pattern is still rather forgiving and the effect might still be linear.
I would combine the OPA205 more with an OPA1677 or OPA1641.
The sensitivity to the divider should not be so large - that points to maybe instablity or ground issues.
-
A fast OP-amp should have a decoupling capacitor directly on the adapter PCB and not just rely on the original PCB for decoupling.
Is this referring to my Ali adapter board also, was thinking that I would have added some stray inductance so it would benefit from a cap.
Mouser order finally came so checking which of the 10nF caps are best out of the Kemet, WIMA and TDK varieties I got.
What are your preferred methods for leakage testing.
-
The capacitor directly at the adater can also help with a standard SO8-DIP8 adapter. For an OP-amp with > 10 MHz GBW this would definitely make sense, especially if the decoupling on the PCB may also not be that good.
For capacitor leakage testing it would be with a really stable reference voltage (e.g. 10 V) and than measure the current (pA range) with a TIA. It may well take some time ( a few seconds for a good capacitor, maybe minutes for a poor one) for the current due to dielectric absorbtion to decay. An alternative test for super low leakage would be to use the capacitor with a fA range bias OP-amp (e.g. LMC6482 or similar) as integrator. The tricky part there is still switching at the input to start with a given voltage or discharge. This would need a very low leakage switch or relay (e.g. better than most simple reed relays).
Getting good enough leakage is likely the simpler part. The more critical parameters may well be the dielectric absorbtion.
-
I see the OPA205 + OPA197 combination a bit critical
Ouch ... I ddin't noticed that ... gosh ... seems another round ...