Fluke 89 - Doesn't show zero volts

[spectramax]

Hello,

I have a Fluke 89 IV meter and everythings works on it except the V(DC) and mV(DC) ranges are off.

The V(DC) range shows randomly between 0.2V and 0.1V. Stabilizes to some value betwene 0.1-0.2V.

The mV(DC) range shows randomly between 200-250mV and stabilizes to some value between 150-200 mV.

Is this a calibration problem? I can measure a 9V battery and it shows 9.5V (It offsets by 0.2V I think) but I don't have another meter to cross check it with.

Would very much appreciate any insight. If it is a calibration issue, I can spend the $65 and get it calibrated but I don't want to risk wasting $65 if they come back and say your meter is broken.

Thank you.

[Kim Christensen]

Open it and see if it's had any moisture ingress. If it has, it probably just needs a cleaning and drying out.
I've never see a Fluke go this far out because of calibration drift.

[spectramax]

Looks absolutely spotless. No burned components or any sign of misuse.

I've never taken Fluke meters apart, my god they are beautiful. Any ideas what could be wrong here?

[Swake]

If you short the inputs, is it still off or only when unconnected?

[floobydust]

It's perfectly normal for a multimeter to not read zero V when the input is not connected/floating. They have a high input impedance and will pick up noise and static electric charge.
Your bench may be noisy - ungrounded or non-conductive/ESD mat missing, noisy LED lighting, static charge on your body etc.
Nearby cellphone/WiFi packets will also cause readings to jump around.

Why are you blaming the multimeter, it's calibration etc.?

Short the input and see if you read a true zero. If you don't then there is a problem.

The supercaps are known to take the piss and leak on the board as well.

[spectramax]

It shows 0.001V DC and 0.7mV DC when shorted. But horror: 110 Ohms when the leads are shorted.

Something is totally screwed up with this meter.

[floobydust]

Also look at Diode-test with open/shorted leads, as well as Ohms with open/shorted leads.

[Manul]

It shows 0.001V DC and 0.7mV DC when shorted. But horror: 110 Ohms when the leads are shorted.

Something is totally screwed up with this meter.

Sounds like there is current injection somewhere. Do you have schematics of this meter?

but I don't have another meter to cross check it with.

That's a problem. Most likely you need a meter to diagnose this meter.

[Shock]

Measure one lead at a time into the corresponding socket on ohms. Note that with socket detection you need to tilt the probe at an angle to find the contact. Compare that to the socket solder joint on the rear. Swap leads and repeat the other way. You should be able to rule out leads and socket this way.

You can soak a cotton bud in IPA to clean them. When the cotton bud is soaked you can make the tip smaller to fit. Dirty sockets require multiple attempts with a clean bud till no dirt is seen. Folding printer paper or bending a thin slice of card then soaking in IPA can be used to clean inside the leads shrouded plugs.

If that model has logging and a bad super cap, it needs to be removed or replaced as well as cleaning the pcb but otherwise cleaning the pcb often resolves these type of problems regardless. But first do check it's got correct Fluke fuses in the correct sockets as they throw measurements out. If it's been abused by a former owner you can test movs and ptc.

[coromonadalix]

and a 189 schematic have leaked on the web and here (eevblog)  if you need some verification of input components ...

my bad  it was 189 II   aka Fluke 287 289 ....   

[mikerj]

What's is that speckled contamination on the PCB?

[floobydust]

I think it's just the soldermask texture, maybe the heat press did it. Some of the PC board looks brown near the big fuse? but it might be the photo.
89, 189 service manuals have no schematic that I could see.
If it's reading 110Ω high then I would suspect the clamp transistors or PTC or worse. Best to use another multimeter to troubleshoot. It looks like it's in very good shape and worth a dig.

[spectramax]

Thanks, that's just the texture of the PCB. It looks way worse in the pics. It's just a satin structure, iPhone camera does some computational photography crap that makes it look crazy.

Yea it sucks that this meter is in excellent condition but can't seem to read right.

Here is what I did:

• I got hold of a Keithley 2450 source meter as a makeshift standard
• Calibrated the 0mV range 30mV, 500mV and 3000mV.
• Calibrated 0, 500 ohm and 5kOhm range using a 510 ohm and 5.1kOhm resistor with 1% tolerance. Yeah it is not ideal but I want to see if that fixed the ohm range

None of these worked . I am afraid the meter is a goner. Any parts I should replace for cheap and have it a go?

[floobydust]

You can't calibrate out a hardware problem, that is a silly notion and makes things worse because now it's trying to read correctly despite a fault 
If you already think the meter is a "goner" call the undertaker.
As has been said, you need another multimeter to test the few fault prone parts. Exactly what the multimeter reads on some functions was also requested but I did not hear back.
I don't think I can help because troubleshooting needs patience and a bit of listening or it just results in a trainwreck.

[Kim Christensen]

You can't calibrate out a hardware problem, that is a silly notion and makes things worse because now it's trying to read correctly despite a fault 

Yup!
I used to repair test equipment that had sometimes been rejected by our calibration lab 1/2 way through the cal process. So, I'd repair it and it would still appear broken because the calibration data was now so far out of whack. Then I'd have to do a half assed cal just to verify that my fix had worked before sending it back for it's official cal. It'd be hell if it was an intermittent or multiple fault situation.  So it's better to repair before you attempt calibrating if you know it's got a fault.

[Shock]

None of these worked . I am afraid the meter is a goner. Any parts I should replace for cheap and have it a go?

Noone said calibrate it. The calibration just doesn't drift like that for no reason. It's either dirty or damaged sockets, input damaged (likely from abuse) or contamination from a bad super cap (which you won't be able to see). All really well known issues here on the forum.

Read my prior post on troubleshooting. It's easy to skip straight through to cleaning the pcb or calibrating but those are the last things, after you have absolutely ruled out the sockets, leads, fuses, super cap, input protection. Yes it's very repairable.

[spectramax]

I wasn't calibrating it out of a problem. Jeez. That would not be ideal.

Here is my thought process: I wanted to see if meter would read 0 ohms if I reset the calibration. If it did, that means the meter had drifted or changed somehow from the original calibration point. If it did not (as I found), then that means there is a deeper problem despite of calibrating it to 0.

There is no need to facepalm and be condescending.

[Shock]

Mine was a delayed response, I didn't see others had just said the same thing.