34970A DMM Module Repair

[CrazyTiger]

I am in the midst of repairing a 34970A with multiple issues.
The unit started with  selftest errors as below:

604 Self-test: A/D slope convergence
609 Self-test: DC gain x1
610 Self-test: DC gain x10
611 Self-test: DC gain x100
612 Self-test: Ohms 500 nA source
613 Self-test: Ohms 5 uA source
614 Self-test: DC 300V zero
615 Self-test: Ohms 10 uA source
616 Self-test: DC current sense
617 Self-test: Ohms 100 uA source
619 Self-test: Ohms 1 mA source
621 Self-test: AC rms full scale

Replacing all AD711 AD706 took away all the error, though the actual culprit was one AD711 AD706 (U400), might as well change all I suppose.

Upon changing self-test passed. Measurements were super out of tolerance. Zero adjustment did not go thru stating no signal, That directly pointed me to possibility of defect on the input relays.Had to wait an extra week after the relay that Element14 delivered came broken in half. .

Replacing relays, Zero Adjustment passed. Yet I am noticing in all functionality measurements are slowly drifting up, another weird error i have encountered is if I am to perform -10V adjustment, the meter gives me error 731 Cal: A/D turnover correction out of range, in all positive ranges adjustment is going thru.

Next up will be measuring the supply voltages to see if they are stable or not, maybe its drifting causing this issue.
Any other idea is much appreciated.

[Kleinstein]

The 604 error and thus an ADC problem could explain all other errors. Without the ADC correctly working, the other self tests may easily fail as well.

Normally there should not be very much turn over corrent needed for the ADC. For the turn over error I see mainly 3 components:
1) A problem with some offset at the input, so the turn over error is at the zero point. This error enters twice and adds up. The test from positive to negative side needs a careful offset adjustment.
   Excessive input current could be a cause for additional unexpected offset.
2) A problem with the amplifier, especially the output side cross over error of the OP27. This error would be more like a step at low voltage. So one would  alread see it with some 0.1 or 1 V in the 10 V range.
3) An INL problem if the ADC, especially a square law contribution.

A point to check would be the 5 V supply to the 74HC4053 switch chip. Too low a supply here could cause an INL problem, especially a square law contribution and thus turn over error.  As the AD711 at the ADC seems to have been bad, there is a slight chance that the HC4053 could have been damaged as well (e.g. from some odd transients in the supply turn on/turn off sequence).
Aof cause check the general supplies. AFAIR there are quite a few extra zeners to reduce the supply to some chips that may also fail (e.g. after a latch up).

[Dr. Frank]

I repaired my 34970A last year, with partly same errors:
https://www.eevblog.com/forum/repair/hp-34401a-error-612-613-615-617-618-619-621/msg3571485/#msg3571485

U201, an AD706 was defect (as often) for Ohm mode.
Frank

[CrazyTiger]

I made a mistake on my initial post, the replaced part was AD706 @ U400. U201 have been replaced as well.

Regarding the 5V line, the voltage to the IC is measuring 5.012V, and stable. Will scope the voltage lines for noise later.