Wanted ADC hybrid for 3456A (U500)

[rolkinas]

Hello,
I am looking for ADC hybrid (U500).

Marking on ADC chip: 1QE7 0057
Wanted replacement.
Preferably from EU.

[MadTux]

What makes you believe that it's fried?
High voltage, applied by stupid user, hardly makes it that far into an instrument.
Have you measured the resistors, the resistors inside of U500 are in the schematics in the manual

Since it's only a couple of precision resistors in there, you can probably replace them by low tempco resistors like Vishay or Alpha Electronics from ebay and recalibrate it afterwards.

[rolkinas]

I swapped ADC between two 3456A units. Faulty ADC gives error 4 on both boards. DC volt accuracy on faulty converter is not totaly bad, but last digit on 6.5d mode stays zero and kicks in only in 100cycle (slow) integration mode. AC voltage and ohm modes are totaly non functional. I checked internal resistances of converter according to manual and they are ok but it is very rough test. ADC from other 3456A works perfectly on both units.
It looks like this unit was assembled from faulty parts by previous owner. Too many failures in one device to be honest (:

[rolkinas]

Anyone?

[Ordinaryman1971]

Is it the same hybrid as in 3478A? I have one of those if you need one.

[Kleinstein]

The HP3478 ADC hybrid is a different thing, more than just some resistors.

The main part of U500 are 4 precision resistors at around 100 K. The other resistors are less critical by something like a factor of 10 or more. So a replacement could start with an LT5400-100K.

The other point would be measuring the resistors as far as possible - chances are, that only a few are actually broken. With some luck it is not one of the critical ones (the one at the input and the 3 values near 100 K at the top of the drawing).

[rolkinas]

Thank you for ideas.
I measured all resistor once more and found that particular one 100k between pins 16 and 17 to be drifted to 101,918k .
This stopped unit from displaying 6-th digit and malfunction of resistance measurement. Second resistor between pins 15 and 16 measures 100,117k. I placed temporary 6M in parallel to pins 16 and 17 and unit came to life.
I think that replacement of those two 100k resistors with new quad network should be sufficient while those are in reference side. Other 100k are on ADC side.

[Kleinstein]

A large resistor value between pins 16 and 17 could be compensated by removing JP501 / JP502 if still present.

[rolkinas]

Tried removing jumpers. One was missing initially. It didin't solved problem. drift of pin16-17 resistor seems to be too large in terms of precision. Almost 2k off. Assuming that jp501 was cut sometime ago does that mean that meter started difting earlier...?

[Kleinstein]

I expect JP501 and JP502 to be there to compensate the initial tolerances. The effect should be something on the order of 0.1% and 0.2%. This could be about the accuracy needed for the ratio to get a good DNL and the slow slope work well. So chances are more that JP501 was cut during production / initial adjustment.

The current mismatch is about 1.8 % and thus too large to be compensated this way. However the best ratio may not be with exactly same resistors - some other resistors (also part of U500) also have an effect.

[rolkinas]

 After some experiments I ended with 5,6M resistor soldered in parallel to pins 16-17. This let to calibrate reference without problems. Actually jumpers returned to initial position during calibration. Will leave it powered on for several days to track drift. So far so good, no difference observed compared to other 3456A and 34401A. Not a perfect fix, but better than none
At least since replacement hybrid becomes available

[coromonadalix]

Glad it worked     searching fleabay and others may provide you with one someday ...