Agilent U1272A Meter Problems

[Sxhosts]

I have had this Agilent U1272A meter for about 10 years and recently noticed a problem with both the AC and DC volts ranges. When measuring a voltage the meter shows a reading that is about 10% too high and then slowly settles close to the actual value.

For example, if I set my PSU to 5V DC and measure its output the meter will read about 5.5V and then over the course of about 3 hours settle towards 5V but it never quite makes it. If I short the input, it displays -0.1V and again, settles towards 0V but never gets there.

Current and resistance ranges seem okay, capacitance range is wobbly but reasonably accurate.

I'd like to try fixing the meter, can anyone give me any clues as to where in the signal path I should be looking?

[siealex]

Check the whole PCB for water damage.

[Sxhosts]

Thanks for your suggestion.

I have examined the PCB and there is no sign of water damage.

I have now tested the DC millivolt range and it is stable and accurate.

The only ranges that cause this issue are AC and DC volts, does anyone know what path these signals take that differ to the other ranges?

Educated guesses are most welcome!

[wraper]

Try disabling/enabling LPF.

[Sxhosts]

If you mean disabling the LPF in setup then I've just tried that and it made no difference.

If you mean physically disabling the LPF then I don't know how to do that.

[wraper]

As there is no schematic available, I'd try to figure out the (signal path) difference in rotary switch between V and mV positions and start looking from there.

[Sxhosts]

Thank you.

So I can rule out all the input protection components, PTC, GDT?

Would I be correct in thinking that the AC and DC ranges go to the thin film resistor network to be scaled before the input to the ADC and that other ranges bypass this network?

[Sxhosts]

Meter is now fixed.

If anyone has a similar issue the culprit was C35 found between pins 1 and 2 of the thin film resistor network. Refer to Dave's teardown photo at https://www.flickr.com/photos/eevblog/5739205880/in/album-72157626761928610/

Capacitor is removed and meter now stable and accurate on all ranges.

Best guess for replacement part C35 is ~10pF, 3 kV, 1808 case size.

[voltsandjolts]

Dave also had an issue with that cap, needed cleaned underneath to stop some leakage.
Maybe the meter has some fluid ingress that dried up but messed up this high-impedance area.
youtu.be/CKdeqqo2diI?t=1694

[Sxhosts]

Excellent video! I didn't find it while researching because I was only searching for U1272A but seems like the U1273AX has the same front end.

Although there didn't look to be any contamination, I cleaned the capacitor and the board with some IPA, put the capacitor back in and all seems good.

Higher frequency AC ranges are working now, without the capacitor only up to about 100Hz is accurate.

I measured the capacitor out of circuit and it seems to be about 10pF but my measurement capabilities are limited for small value capacitors.

Thank you so much.

[bdunham7]

I guess that would have been a good location for a PCB slot!