Fluke 189 strange readings

[giosif]

Hi,

I have a Fluke 189 meter with a problem and I am in need of help with troubleshooting it: when set in DCV mode, with no probes connected, the meter will display a reading between 75 and 80 mV.
And, interestingly enough, if I measure the voltage between the COM and V input jacks with another (high input impedance) multimeter, I actually measure 75-80 mV.
If I do the same using a standard 10/11 MOhm input impedance meter, I measure roughly half of that (i.e. 35-40 mV or so) - and both the measuring meter and the Fluke 189 will display this new value.

Next, if I switch the Fluke 189 to DCmV mode (again, no probes connected), the readings start rising as if charging a capacitor (I don't remember if they eventually settle to any specific value).

Back to DCV mode, I measured the input impedance of the Fluke 189, but the readings kind of confuse me (the 189 meter was set to manual range mode, and I measured resistance between COM and V terminals for the different ranges as below):

Range (V)		5		50		500		1000
Measurement_1 (MOhm)		11.60		10.65		10.55		10.54
Measurement_2 (MOhm)		9.57		8.85		8.78		8.77

[Edit] And, of course, I forgot to mention what Measurement_1 and Measurement_2 represent. 
Measurement_1 is resistance read when connecting the positive lead of the measuring meter to the V input jack of the Fluke 189, and, correspondingly, with the negative lead connected to the COM input jack of the Fluke 189.
Measurement_2 is resistance read when swapping the leads (i.e. positive lead going to COM of the Fluke 189 and negative lead going to V).
[/Edit]

I also checked the main components in the input protection section - PTC, fusible resistor, MOV's - and the (in circuit) measurements are similar to those on a Fluke 89 IV board which I know is good.

Does anyone have any suggestions where to look and what to check next, please?

Thanks!

[mzacharias]

Does it measure voltage and resistance correctly?

For the milli-volts function short the inputs and see if the readings persist.

It is normal to pickup low millivolt readings just from ambient noise ESPECIALLY if test leads are connected and even more so on AC function.

[giosif]

Does it measure voltage and resistance correctly?

Yes for voltages.
I will try resistance and report back.

For the milli-volts function short the inputs and see if the readings persist.

I will try this but, in V mode, shorting the inputs gets the readings to 0V.
I expect the same will happen in mV, but will confirm later.

It is normal to pickup low millivolt readings just from ambient noise ESPECIALLY if test leads are connected and even more so on AC function.

I am not entirely sure about this, especially for DCV, as I have a working Fluke 89 IV and the DCV readings on that, even with probes attached, will vary only on the last digit (i.e. 100's of uV), the rest being all 0's.
At the same time, the readings I'm getting on the Fluke 189 are even without any probes attached.

[mzacharias]

Test continuity from the probe tips to the solder connections underneath the input jacks. The 189 types have chronic issues with the input jack assemblies.

Also, if you have not done so, try different probes, especially try the probes you used with the other meters.

[giosif]

Some updates (and information that was requested) on this:
- After posting my previous response here, while investigating the problem, I noticed that if I exhale on the meter PCB (on the side with the input protection circuitry and resistor network), the readings start changing (e.g. in V mode, I could the readings to rise all the way up to over 1V).
- I checked mV functionality and, initially, the readings were about 55 mV with the leads open and -0.23 mV with the leads shorted.
  However, I say "initially" because I then exhaled on the meter PCB and the readings started to change.
  This happened even in the case where the probes were shorted.
- In terms of resistance measurements, the initial readings were about 60 MOhm with the leads open and 0 Ohm with the probes shorted.
  Measuring a precision 10.00 kOhm showed a reading of 9.97 kOhm.
  However, like with the other tests above, even with the probes shorted, the readings changed to a non-0 value when exhaling on the PCB.
- On the input jacks, in fact, the meter has been without the input jacks from the very start of this thread.
  Also, the clips, leads, probes I'm using when doing these tests/measurements, I am quite confident they are all good.

[Samogon]

I have 87IV acting like that.

[giosif]

I have 87IV acting like that.

Ok, so this issue may not be just a one-off.
Have you done any investigations?
Any additional observations you might have made?

[Samogon]

I believe it is ADC chip, meter is on the waiting for repair shelf, will update when i start working on it.

[giosif]

I believe it is ADC chip, meter is on the waiting for repair shelf, will update when i start working on it.

Yes, please, do let us know of anything else you may find when working on it.

Thanks!

[goaty]

Maby the PCB needs cleaning with IPA and ultrasonic to get rid of unwanted leakage/creep currents ?

[giosif]

Maby the PCB needs cleaning with IPA and ultrasonic to get rid of unwanted leakage/creep currents ?

That is something that crossed my mind, but it's one of the last things I plan to try, because I don't have an ultrasonic bath and cleaning manually with IPA is probably best done after removing the ceramic resistor network (and I want to avoid unsoldering / re-soldering that, unless necessary).

[retiredcaps]

That is something that crossed my mind, but it's one of the last things I plan to try, because I don't have an ultrasonic bath and cleaning manually with IPA is probably best done after removing the ceramic resistor network (and I want to avoid unsoldering / re-soldering that, unless necessary).

Dunking the pcb in IPA and scrubbing with a clean toothbrush is usually one of the first things I do (if I think there is leakage).

I have dunked Fluke pcbs with the ceramic resistor network intact without problems.

[giosif]

Dunking the pcb in IPA and scrubbing with a clean toothbrush is usually one of the first things I do (if I think there is leakage).

I have dunked Fluke pcbs with the ceramic resistor network intact without problems.

Thanks for the reassurance!
My main concern with the ceramic resistor network is not the IPA, but the handler of the cleaning task, namely me.
It wouldn't be the first time when, in my attempt to repair something, I damage something that wasn't initially broken. 

[giosif]

Update: after yesterday's post, I decided to "grow a pair" and clean the PCB with IPA and a toothbrush.
And, lo and behold, we have an improvement in the situation:
- in VDC mode, the meter now registers very close to 0V, with no leads connected - only the last digit (i.e. 100's of uV) fluctuates a bit, which is fine.
- in resistance mode, with no probes attached, we now measure OL, which is normal vs. the initial 60 MOhm.
  Also, checking the same 10.00 kOhm resistor from before, we now measure 10.03 kOhm (which is a change from the original 9.97 kOhm, but not necessarily an improvement).

That said, in mV mode, if there is an improvement, it's so small I can't observe it.
So, we're not 100% done here, but I think the above is strong enough evidence that a thorough cleaning of the PCB is needed, before any other type of investigation.

I now need to find a place that would let me use their ultrasonic cleaner at a cost, if one such place even exists (and the cost is reasonable).

[retiredcaps]

So, we're not 100% done here, but I think the above is strong enough evidence that a thorough cleaning of the PCB is needed, before any other type of investigation.

One a couple of meters, I had to do 2 or 3 IPA pcb baths before all the leakage was gone.

If you haven't done so, let it thoroughly dry overnight or use a PC fan to blow on it.

[retiredcaps]

Modemhead's blog on Fluke 189 cleaning.

http://mrmodemhead.com/blog/gallery/fluke-189-clean-up/

General cleaning

http://mrmodemhead.com/blog/dirty-multimeter-clean-up/

[retiredcaps]

BTW, I never asked you if this your meter that suddenly developed the problem or you got this for repair off ebay or something?

Was there ever any AA cell electrolyte leakage in the battery compartment?

[retiredcaps]

Modemhead's dunking method in IPA with clean brush.  From

https://www.eevblog.com/forum/testgear/old-fluke-multimeters/msg633201/#msg633201

[retiredcaps]

More 189 cleaning ...

https://www.eevblog.com/forum/repair/fluke-189-teardown-repair/

[giosif]

Many thanks for all the links, retiredcaps!
It is a pleasure reading Modemhead's blog entries.

I will try multiple IPA baths.

About your question, I got the meter off ebay and, yes, when I got it, there was a lot of battery electrolyte leakage in the battery compartment.
In fact, given the state of the battery compartment, I was surprised to see the PCB didn't look to have been contaminated, apart from a few pieces of dried electrolyte that had broken off from the said compartment and batery contacts (the ones that eventually make contact with the board).

[retiredcaps]

About your question, I got the meter off ebay and, yes, when I got it, there was a lot of battery electrolyte leakage in the battery compartment.

There may be electrolyte that wicked up/into some components that you cannot see or hard to detect.

Ultrasonic cleaning may help if another IPA bath doesn't help.

[giosif]

There may be electrolyte that wicked up/into some components that you cannot see or hard to detect.

Ultrasonic cleaning may help if another IPA bath doesn't help.

Yeah.
Right now, I let the board soak in IPA and scrubbed it with a toothbrush a couple of times.
I will let the board soak in IPA some more over the night, give it another scrub with the toothbrush, then one last dip in IPA and, finally, let it dry (for 24 hours or even more).
If this doesn't make a difference, ultrasonic cleaning is the next step, provided I can find a cheap way to do that (I don't have an ultrasonic bath).

I will keep everyone here posted.

[retiredcaps]

I will let the board soak in IPA some more over the night,

Hmm, I have never let it soak more than 10 or 15 minutes.  I'm not sure about letting it soak overnight.

[giosif]

Hmm, I have never let it soak more than 10 or 15 minutes.  I'm not sure about letting it soak overnight.

Oops. Too late now.
The board is drying right now and I'll look to test it sometime tomorrow.
If there is a change, I hope it is a positive one...

[giosif]

An update on this, as promised: after letting the board dry and testing it, I am happy to report the mV readings look good now.
So, it must have been contamination from the leaked batteries.

Thank you very much for your help!

The above said, we do have one casualty from the IPA bath process: the piezo buzzer, which is much lower volume than expected (forgot to cover the hole before dunking the PCB in IPA  ).
Does anyone know if the buzzer will eventually recover, or I'd better look for a replacement?

Also, if anyone happens to have some spair selector switch contacts, please PM me (I need 4, ideally 8 pcs).