Restoring a Fluke 8050A

[smbaker]

I'm not sure why I went down this path, just  asking to invest time and money into an obsolete bench DMM, but here I am...

I've already replaced the 14 year old NiCd batteries (which fortunately did not appear to have leaked). The DMM appears to be 10% - 30% out of calibration on nearly every range I try it on. For example a verified 2.25VDC source reads 2.85VDC. DCV and ACV are the only functions I've checked so far. I didn't bother with resistance or current yet, but I would not be surprised to find similar problems. This drift from calibration seems a bit excessive. I'm thinking perhaps I should start by replacing frequent trouble areas, such as electrolytics / tantalums. Does that seem wise? or should I just start by calibrating it first? Any suggestions?

Scott

[lowimpedance]

Has the unit sat unused for a long time ? Perhaps the switches have some oxide build up leading to noisy readings, are they consistently off spec or more random ?
Another possibility is surface contamination so a good clean around the ADC etc might be in order. I would also check the resistance to be sure the issue is across all functions.
After cleaning and exercising the switches there is no difference then move onto the power supply caps, cant hurt to replace although I have not needed to do that on my few units yet !
I would be trying to get repeatable and stable readings before heading for the CAL adjustments when other avenues are exhausted.

Despite being long obsolete its still a good performer with some extra useful features when working.

[med6753]

Check the power supply for correct voltages and ripple. Bench Flukes often had Philips crapacitors which are known leakers.

[bdunham7]

Don't calibrate a broken meter!

The way the voltage reference system works on this is a little convoluted to understand and the hand-drawn schematics have a few rough spots in my version, but I would guess your problem is in the vicinity of VR1/R11/U18/U19/U20/U8A and the resistor divider near VR1.  Definitely check as many ranges as you can at two or three points within each range, just to make sure the fault is consistent and linear.  I had a small pile of these and similar meters. I ended up fixing two or three of them then decided the rest just weren't worth the effort.  You will learn something in any case. 

You should test the power supplies and so on, but I wouldn't go replacing anything yet unless you observe a fault.  Depending on the results you get from testing all the ranges, you might look at the jumper selection process for when there isn't enough room on the calibration pots.  You also can check the voltage and current of VR1--this will require desoldering and lifting a leg, but it will tell you a lot about what is going on.

[smbaker]

Thanks for all the advice, I will dig into it this weekend. Any suggestions on the capacitors? I usually have on hand some decent quality nichicons.

This was an eBay find, and it probably has been powered off a long time. I'll pay some scrutiny to the switches. I would say the error seems consistent though.