Agilent 34401A Bubble on PCB ... can it be saved ? (SOLVED)

[Kleinstein]

The votlages at pins 5 and 6 of U201 are a bit different, but this is likely from the loading from the DMM to measure these voltages. The voltages look reasonable.

Being some 4 % off with the 1 M resistor supports that the test current is too low.  Maybe try if the reading drifts just after turn on / warm up. If the issue is from leakage in the semiconductor, it is suspected to get quite a bit worse (like 2 x the leakage) with warm up. With leakage from a dirty PCB less temperature effect is expected.

[The_Spectrum.A_idiot]

The votlages at pins 5 and 6 of U201 are a bit different, but this is likely from the loading from the DMM to measure these voltages. The voltages look reasonable.

Being some 4 % off with the 1 M resistor supports that the test current is too low.  Maybe try if the reading drifts just after turn on / warm up. If the issue is from leakage in the semiconductor, it is suspected to get quite a bit worse (like 2 x the leakage) with warm up. With leakage from a dirty PCB less temperature effect is expected.

Well I didn't leave the instrument warm up with its covers on for 5h. ( I don't have such time on my hands)
The pcb was cleaned with IPA as best as possible. Next step is to use high pressure contact cleaner to get under every IC and diode ( i can't imagine any other way to get under that clean.)

I am hoping for Dr. Frank to give some insight to what might be at fault. I am almost certain that some componet I selected is not appropriate. And I should had just changed the OP AMP only.

never the less, time will tell.

If further warmup is needed I will do such test with different resistors becuase the 1M was the biggest I had in my arsenal at the moment.

Thank you in advance

[Kleinstein]

The current is rather far off, well more than expected for leakage at Q211 or Q202 or the OP-amp bias current. The transistors in the protection would not effect the low current, even if bad or broken.
A point could be if the OP-amp is genuine or a Ebay/Ali part.

The change in current from doing the extra cleaning step points strongly towards a dirt problem, though it could still also be a thermal effect (e.g. higher temperature in the 2nd test).
No need for a full warm up test. The first half hour should show most of the thermal effect.

Contact cleaner may or may not be suitable, as some include lubricants and can leave a resudue. They may also not be good solvents, but more a thing to remove oxide layers. A spray can can also contaminate other areas from splatter. HIgh pressure is more a thing for cleaning a whole board, not so much for loacal claning. That is more a thing for brush, seringe and tissue.

One can still use IPA or similar around / under the chips. It will flow also under the chip and can be sucked out / wicked out with clean tissue. The dirt is removed by the solvent removed, not by the solvent added.

IPA is also only good for some contaminations and there are different purities / grades. More salt like contaminations are better removed with water or a IPA / water mix. More organic / grease like dirt would want high purity / low water content IPA or even a different solvent. It may well need multiple cleaning steps. Usually water or a water+IPA mix first and at the end clean IPA.

One could try / test cleaning of flux residue on a different PCB. So do experiments separate first, to see what works with the flux that was used in the repair. For the flux the soldering step can change things - heated flux is different from fresh or only partially heated. Not all residue from the flux is actually bad - some contain some parafin / wax component, that is hard to remove, but isolating.

[The_Spectrum.A_idiot]

All parts are from mouser. I would never get ebay parts unless are unoptenium.
But what ever I changed / replaced are original mouser parts.

All replaced with soldering iron and soldering wick.

Flux I have a syringe ( i need to find the brand name I can't think of it on top of my head)

Cleaning, before soldering anything on I cleanred the area and pads after wicking them with IPA and a q-tip ( the thing used (incorrectly) for ear canal cleaning) and then soldering on the parts 1 by 1 ( 1 leg at a time)

Let me know.

Thank you in advance

[The_Spectrum.A_idiot]

Flux Type : SMD291/30CC - Chip Quik

[Dr. Frank]

Used a 9V PSU
Tried every range on VDC and values don't change.

re measured μΑ on ranges after cleaning the flux and doing the range change.
Results :
000. ohm range -999.1 μA
0.1k range - 999.3μA
00.1k range -99.46μA
000.1k range -9.84μA
0.1M range -4.83μΑ
00,1M range - 0.36μΑ **
000,1Μ range -0,36μΑ **

The reference currents are all still n.OK, i.e. about 140..180nA too low, which is clearly visible for the 500nA, 5µA, 10µA and probably also for the 100µA reference currents.That's the reason, why you get ERR 612, 618, which both indicate a failure of the 500nA current, i.e. out of tolerance.

U201 was really defective, as initially none of the expected voltages @ pin5, 6 were ok, instead showing the usual faulty values for too high bias currents, i.e. broken input stages.

Now you have to search, where these about 180nA are flowing to GND.
Hopefully your new U201 is OK, then Q211 would be the next suspect. So let's at first rule out both these failure modes:

Please desolder CR202, and then measure the reference current directly in situ towards AGND, i.e. from pin 3 / collector of Q209 to AGND, which is @ INPUT LO.
you might check for the lower currents of 500nA, 5µA and 10µA only, so you have to select 10M, 1M and 100k ranges (see page 98 of service manual). An external resistor is not needed.

Please compare these three currents with those you have measured in the above table.

Frank

[Dr. Frank]

For evidence, I measured the currents of my 34401A:
100, 1k range: 1.0003 mA (+0.03%)
10k range:       99.898 µA (-0.10%)
100k range:     9.9858 µA (-0.14%)
1M range:        5.0198 µA (+0.4%)
10M range:      501.66 nA (+0.3%)

Therefore, the 3 highest currents are about 0.1 .. 0.2% within nominal, and the 2 lower currents are about 0.5% within nominal, so that's the ballpark what to expect as being OK or nOK reference currents.

To be more precise: The nominal values of the currents should be within about 0.5% consistent, relative to each other, as an additional +/-2% for the LM399 voltage reference (6.95V nominal) has to be taken into consideration.

For my 34970A, I found these currents:
100, 1k range: 0.9877 mA (-1.2%)
10k range:       98.712 µA (-1.3%)
100k range:     9.8962 µA (-1.0%)
1M range:        4.9341 µA (-1.3%)
10M range:      494.3 nA (-1.1%)


Frank

[The_Spectrum.A_idiot]

CR202 Removed;
Measurements still the same :
00,1M range - 0.36μΑ
000,1Μ range - 0,36μΑ

next step ?

[mawyatt]

Out of curiosity we took a quick look at our HP34401A and AG34401A {readings}. These were acquired years ago with unknown history and aren't used much these days, and never properly calibrated since acquired.

1.001484mA     {1.001779mA}
100.3637uA      {100.0724uA}
10.00111uA      {10.04446uA}
4.99031uA        {5.026583uA}
497.396nA        {503.761nA}

BTW the above Dr Frank post obviously should be 501.66nA not uA on the 10M range.

Best,

[Dr. Frank]

what about 1M and 100k ranges?

Next step: Please remove Q203 and measure these 3 currents from pin 2 of Q202 to AGND / INPUT LO.

Frank

[The_Spectrum.A_idiot]

the rest of the ranges measure same as before.

okay.

[The_Spectrum.A_idiot]

what about 1M and 100k ranges?

Next step: Please remove Q203 and measure these 3 currents from pin 2 of Q202 to AGND / INPUT LO.

Frank

Q203 removed and cleaned everything flux coated

results :'(
pin 2 of Q202 to INPUT LO
000. ohm range -997.6 μA
0.1k range - 997.6μA
00.1k range -99.38μA
000.1k range -9.89μA
0.1M range -4.92μΑ
00,1M range - 0.35μΑ **
000,1Μ range -0,34μΑ **

Thank you in advance

[Dr. Frank]

Ok, please re-insert CR202 and Q203. You might then test these currents on the outside (Input Lo / Hi) again.

Have you tested the old components, Q203-Q210, Q211, Q202 and CR202 if and how they had been damaged?

If these parts are all damaged there had been a very high overvoltage event, which could have affected U201-B, but also C210 and U101-E.

So my next guess would be to remove or replace C210, that's a 0805 ceramic capacitor .. I don't know if it's a special type for low leakage currents..
It's a COG / NP0, which usually has better insulation than X7R. You might check this components outside for leakage, when applying 20V in series.

Have you got a genuine, new AD706?

Frank

[The_Spectrum.A_idiot]

Q202 I was in between what component to select

I was between :

https://eu.mouser.com/ProductDetail/512-MMBFJ176 (this one is installed)

and

https://eu.mouser.com/ProductDetail/106-SMP5461

I will reinstall the rest and order the capacitor

If U101 is broken I should assume that this is the end of this journey right ?

Thank you in advance

[The_Spectrum.A_idiot]

Ok, please re-insert CR202 and Q203. You might then test these currents on the outside (Input Lo / Hi) again.

Have you tested the old components, Q203-Q210, Q211, Q202 and CR202 if and how they had been damaged?

If these parts are all damaged there had been a very high overvoltage event, which could have affected U201-B, but also C210 and U101-E.

So my next guess would be to remove or replace C210, that's a 0805 ceramic capacitor .. I don't know if it's a special type for low leakage currents..
It's a COG / NP0, which usually has better insulation than X7R. You might check this components outside for leakage, when applying 20V in series.

Have you got a genuine, new AD706?

Frank

I got this from mouser.
https://eu.mouser.com/ProductDetail/584-AD706JRZ

No ebay or unknown source components.

[The_Spectrum.A_idiot]

Ok, please re-insert CR202 and Q203. You might then test these currents on the outside (Input Lo / Hi) again.

Have you tested the old components, Q203-Q210, Q211, Q202 and CR202 if and how they had been damaged?

If these parts are all damaged there had been a very high overvoltage event, which could have affected U201-B, but also C210 and U101-E.

So my next guess would be to remove or replace C210, that's a 0805 ceramic capacitor .. I don't know if it's a special type for low leakage currents..
It's a COG / NP0, which usually has better insulation than X7R. You might check this components outside for leakage, when applying 20V in series.

Have you got a genuine, new AD706?

Frank

regarding the C210
this is what the service manual said about it :

C210 0160-5954 4 Capacitor-Fxd 220 pF ±5% 50 V 04222 08051A221JAT A

This is the closest I can find

https://eu.mouser.com/ProductDetail/KYOCERA-AVX/08051A221JAT2A?qs=j3sIhLrTlovso9JF71DsEg%3D%3D

[Dr. Frank]

Q202 I was in between what component to select

I was between :

https://eu.mouser.com/ProductDetail/512-MMBFJ176 (this one is installed)

and

https://eu.mouser.com/ProductDetail/106-SMP5461

I will reinstall the rest and order the capacitor

If U101 is broken I should assume that this is the end of this journey right ?

Thank you in advance

That JFET should Q211 should be fine, but I meant U201, if you bought a genuine AD706, or have you picked another type?

Please first remove C210, maybe the OpAmp will not oscillate, and you can test the currents, before ordering.

If U101 is damaged, you might be able to re-calibrate the Ohm ranges, which would then compensate for the additional leakage currents.

I expect that up to 1MOhm range, it will still work ok, 10M and higher might be unstable (over temperature, time), and maybe you will not be able to calibrate these upper ranges properly.
DCV seems to be ok, but you will always get these Test Errors 612, 618, because the 500nA current is outside of tolerance.     

If you can live w/o the higher Ohm ranges, your 34401A might be still quite usable, but it needs a full check of all ranges and modes.

Frank

[Dr. Frank]

Ok, please re-insert CR202 and Q203. You might then test these currents on the outside (Input Lo / Hi) again.

Have you tested the old components, Q203-Q210, Q211, Q202 and CR202 if and how they had been damaged?

If these parts are all damaged there had been a very high overvoltage event, which could have affected U201-B, but also C210 and U101-E.

So my next guess would be to remove or replace C210, that's a 0805 ceramic capacitor .. I don't know if it's a special type for low leakage currents..
It's a COG / NP0, which usually has better insulation than X7R. You might check this components outside for leakage, when applying 20V in series.

Have you got a genuine, new AD706?

Frank

regarding the C210
this is what the service manual said about it :

C210 0160-5954 4 Capacitor-Fxd 220 pF ±5% 50 V 04222 08051A221JAT A

This is the closest I can find

https://eu.mouser.com/ProductDetail/KYOCERA-AVX/08051A221JAT2A?qs=j3sIhLrTlovso9JF71DsEg%3D%3D

That's perfect. But please, check the OpAmp and w/o C210 , before ordering.
Maybe it's the wrong OpAmp (low bias type!) or you test another sample first.

[The_Spectrum.A_idiot]

Ok, please re-insert CR202 and Q203. You might then test these currents on the outside (Input Lo / Hi) again.

Have you tested the old components, Q203-Q210, Q211, Q202 and CR202 if and how they had been damaged?

If these parts are all damaged there had been a very high overvoltage event, which could have affected U201-B, but also C210 and U101-E.

So my next guess would be to remove or replace C210, that's a 0805 ceramic capacitor .. I don't know if it's a special type for low leakage currents..
It's a COG / NP0, which usually has better insulation than X7R. You might check this components outside for leakage, when applying 20V in series.

Have you got a genuine, new AD706?

Frank

regarding the C210
this is what the service manual said about it :

C210 0160-5954 4 Capacitor-Fxd 220 pF ±5% 50 V 04222 08051A221JAT A

This is the closest I can find

https://eu.mouser.com/ProductDetail/KYOCERA-AVX/08051A221JAT2A?qs=j3sIhLrTlovso9JF71DsEg%3D%3D

That's perfect. But please, check the OpAmp and w/o C210 , before ordering.
Maybe it's the wrong OpAmp (low bias type!) or you test another sample first.

https://eu.mouser.com/ProductDetail/584-AD706JRZ

this is the OP-AMP Ι could find.

Wιthout the C210 current when up to 0.44μΑ

Waiting on your response.

[Dr. Frank]

and the other currents, 5µA and 10µA?

Does it now pass the test?

That means, DIP 8 is obsolete, and you have used an SOIC / DIP adapter? Or was it an SO package initially? I don't remember..
This type is correct, anyhow

[The_Spectrum.A_idiot]

and the other currents, 5µA and 10µA?

Does it now pass the test?

That means, DIP 8 is obsolete, and you have used an SOIC / DIP adapter? Or was it an SO package initially? I don't remember..

Let me solder everything back except the cap and see how it goes.

[Dr. Frank]

Great!
I have to leave for basketball in 20min., so maybe I can't answer your questions later.

[The_Spectrum.A_idiot]

still Failed. ( I soldered everything back as they were. )

I managed to find on ebay original OP amps.

Let's hope this is the case.(changing to an original part)

Thank you for your time !

[The_Spectrum.A_idiot]

and the other currents, 5µA and 10µA?

Does it now pass the test?

That means, DIP 8 is obsolete, and you have used an SOIC / DIP adapter? Or was it an SO package initially? I don't remember..
This type is correct, anyhow

Well I managed to solder it on .. regardless if it was a different package ... pinout was the same so I just made due.

[Kleinstein]

A damaged capacitor C210 is possible, e.g. due to PCB stress from the bubble. For a test one could try without or with a different low leakage type, possibly also through hole and a slight larger values like 330 pF.  C0G capacitors should be reasonable low leakage, also for more normal types.

With the flux it is also possible that just re-soldering effects leakage on the PCB. The flux linked should be easy to clean and at least for a first test should even be OK without extra cleaning. Excessive flux that did not got hot may however be a problem and likely needs cleaning.

The AD706 should still be available and the photo shows a SO8 foot-print, not a dip. For the wide / narrow version one often can bridge the gap, though this can make PCB cleaning a bit more tricky.