[SOLVED] Fluke 83 multimeter repair attempt

[giosif]

Hi,

I'm trying to repair a Fluke 83 who is currently displaying ~5Mohm (instead of OL) in resistance mode with open leads.
Going into manual range, the only range that seems to be affected is the 40Mohm one (others seem to be ok, but haven't checked thoroughly).
Also, measuring the input impedance while in volts mode, I get ~5Mohm instead of the expected ~10Mohm.

Any idea where I should check next?

FYI, I've already checked the MOV's and they seem to be ok (both measuring OL for resistance and with no signs of physical damage).

[retiredcaps]

Check spark gap E1.  It should also measure 0L resistance.  RT1 should be 1.5k ohm as well.

[giosif]

Check spark gap E1.  It should also measure 0L resistance.  RT1 should be 1.5k ohm as well.

Checked those as well and they are fine.

[giosif]

May or may not be related, but measuring R11 in circuit gives ~5.3Mohm, while the schematics show it should be 10Mohm.
Anyone with a working Fluke 83 that could measure that resistor and confirm the in circuit reading, please?

[stmdude]

I just took apart my works 83, and the only thing I can confirm is that R11 does not measure correctly in-circuit.
For me it measured about 1Gohm (sometimes) and sometimes around 100-150Mohm.

I used two different bench dmms, as it seemed strange that it would change, but they both do the same (Agilent 34410A)

I took out R11, and measured it separately, and mine is spot-on 10Mohm according to both DMMs now.

[giosif]

I just took apart my works 83, and the only thing I can confirm is that R11 does not measure correctly in-circuit.
For me it measured about 1Gohm (sometimes) and sometimes around 100-150Mohm.

I used two different bench dmms, as it seemed strange that it would change, but they both do the same (Agilent 34410A)

I took out R11, and measured it separately, and mine is spot-on 10Mohm according to both DMMs now.

Wow, many thanks, stmdude!
This is beyond going the extra mile. 

[giosif]

However, R11 is not it: I've just removed and measured it at 10Mohm.
Strange that the in-circuit measurement was different from what you were getting, stmdude. 
Maybe/probably fault resides further down the path of R11...

Or anyone has ideas on what may make the meter input impedance ~5Mohm?

[alsetalokin4017]

Sounds like it could be contamination of the PCB. Has the meter been exposed to moisture somehow, stored in high humidity, been peed on by the cat?  I'd suggest a close inspection of the whole PCB under magnification, and cleaning any suspect areas with isopropyl alcohol, at least 91 percent kind, not the cheap 70 percent rubbing alcohol. You might also try gently drying the thing with warm dry air after cleaning.

I have a 30 year old 83 that still works great. I did have to do the LCD contact strip cleaning thing once. Hardly ever use it now since I have the 87-III.

[ModemHead]

R11 is a pull-up for the micro input that senses whether a plug is in the Amp jack.  Unlikely to be related to your problem.  Measuring in-circuit will pick up the diodes in the micro, so you'll get varying readings based on polarity and how much current the ohm-meter uses.

I totally concur with the IPA cleaning suggestion.

[retiredcaps]

Take the entire thing apart including the range switch to give it a good IPA bath.  One person didn't take apart the range switch and found this ...

https://www.eevblog.com/forum/repair/fluke-787-repair-any-idea/

Another reason to examine the range switch ...

http://mrmodemhead.com/blog/ideal-carbon-footprint/

[retiredcaps]

BTW, since the original 80 series I requires you to take apart the entire meter to replace fuses and battery, the chances of contamination is non zero.

[giosif]

Sounds like it could be contamination of the PCB. Has the meter been exposed to moisture somehow, stored in high humidity, been peed on by the cat?  I'd suggest a close inspection of the whole PCB under magnification, and cleaning any suspect areas with isopropyl alcohol, at least 91 percent kind, not the cheap 70 percent rubbing alcohol. You might also try gently drying the thing with warm dry air after cleaning.

I have a 30 year old 83 that still works great. I did have to do the LCD contact strip cleaning thing once. Hardly ever use it now since I have the 87-III.

I wouldn't know, as I got it as faulty off eBay.
I plan to take a close look at the entire PCB, as you suggest and, possibly, give it an isopropyl alcohol bath as well.
Let's see...

[giosif]

R11 is a pull-up for the micro input that senses whether a plug is in the Amp jack.  Unlikely to be related to your problem.  Measuring in-circuit will pick up the diodes in the micro, so you'll get varying readings based on polarity and how much current the ohm-meter uses.

I totally concur with the IPA cleaning suggestion.

Yep, you are right with the purpose of R11 (still learning how to read schematics).
After taking the resistor out, the meter started beeping, which makes sense now.

[giosif]

Take the entire thing apart including the range switch to give it a good IPA bath.  One person didn't take apart the range switch and found this ...

https://www.eevblog.com/forum/repair/fluke-787-repair-any-idea/

Another reason to examine the range switch ...

http://mrmodemhead.com/blog/ideal-carbon-footprint/


BTW, since the original 80 series I requires you to take apart the entire meter to replace fuses and battery, the chances of contamination is non zero.

Yes, I understand now that IPA bath is the best course of action at this point in time.
And good reads those articles you linked there.

Thank you all for your suggestions!

[giosif]

Aaand we have a winner!!! 

First, a tease: can you spot the sp... problem in the picture below? 

[retiredcaps]

Looks like some black residue on the outside of the range switch?

[Monkeh]

There appears to be a burn mark under the grey box cap, just below the MOVs.

[retiredcaps]

First, a tease: can you spot the sp... problem in the picture below? 

Found it.  Circled in red.

edit: Monkeh beat me by 20 seconds while I was in the graphics editing program.

[retiredcaps]

I suggest desoldering that grey cap and seeing if it was damaged and check the pcb area around there.

[giosif]

Yep. 

Here's a better picture of the same.

As a consequence, I desoldered and checked the cap, the diode and the bridge rectifier and they all checked out ok.
So, in theory, I only need to clear the PCB from the burned part and I should be good to go...
...Unless I tried to desolder the input receptacle assembly and managed to break 4 out of 8 pins. 
I need to source an inexpensive replacement now... (I am aware of the fix procedure provided by ModemHead, but I don't have the tools to do that).

Thank you all for your help!

[giosif]

I suggest desoldering that grey cap and seeing if it was damaged and check the pcb area around there.

I've done that and the damage is not extensive.
I've also scraped off all the bruned burned material, so I should be good now.

[retiredcaps]

...Unless I tried to desolder the input receptacle assembly and managed to break 4 out of 8 pins. 
I need to source an inexpensive replacement now... (I am aware of the fix procedure provided by ModemHead, but I don't have the tools to do that).

Inexpensive and Fluke usually don't go together well in the same sentence.  A new one will likely cost in the $35 USD range not including shipping to your country.

[giosif]

Inexpensive and Fluke usually don't go together well in the same sentence.  A new one will likely cost in the $35 USD range not including shipping to your country.

In that case, I probably want to wait for another faulty Fluke to show up on eBay (I paid $50 total for this one).
If I don't manage to damage the receptacle on that too, that is...

[Gangway]

Try using a paperclip cut into a U shape, flatten the middle part, then cut small groove across bottom of receptical, solder U clip, apply small dob of glue & job done. ( check continuity of all receptical pins)
Fixed my 85III receptical & still working ok.

[halexa]

Hi,

I have an Fluke 83 that is broken that I use for parts. Think it got that cap intact. I'm away during the holiday so  I cant check right now.
But if you are intrested i can check it and if it's ok i can mail it to you.

[alank2]

My dad had an old Fluke 73 that got left out and rained on.  It was performing very badly with wildly swinging readings when just sitting there.  I took it apart and you could see quite a bit of corrosion on the tracks especially near the IC's.  I hit the entire backside of the board with some water soluble flux pen and retouched all the joints and then gave it a water rinse and compressed air blow off.  Got it back together, no wild readings any more.  Calibrated its one DC calibration pot and all ranges are in spec now.

[giosif]

Try using a paperclip cut into a U shape, flatten the middle part, then cut small groove across bottom of receptical, solder U clip, apply small dob of glue & job done. ( check continuity of all receptical pins)
Fixed my 85III receptical & still working ok.

Spent most of the afternoon today doing something similar to what you describe: used copper wire bent in tight flat spiral pattern, with one longer end perpendicular to the plane of the spiral pattern; I then tinned the spiral and soldered it to the broken pieces of the receptacle, after which I soldered the receptacle back to the board.
I now have a working multimeter. 
Not sure how solid the solder joints between the broken pieces of the receptacle and the pieces of copper wire are, but it looks to be working fine.
It will certainly be useful as a low voltage/small tasks multimeter.

[giosif]

Hi,

I have an Fluke 83 that is broken that I use for parts. Think it got that cap intact. I'm away during the holiday so  I cant check right now.
But if you are intrested i can check it and if it's ok i can mail it to you.

Thanks, halexa!
The cap, however, looks to be in good nick: capacitance is spot on, resistance measures OL, and visual inspection shows no obvious damage to the body of the capacitor, except for a small char mark which I scraped off.
But thank you for the offer!

[giosif]

My dad had an old Fluke 73 that got left out and rained on.  It was performing very badly with wildly swinging readings when just sitting there.  I took it apart and you could see quite a bit of corrosion on the tracks especially near the IC's.  I hit the entire backside of the board with some water soluble flux pen and retouched all the joints and then gave it a water rinse and compressed air blow off.  Got it back together, no wild readings any more.  Calibrated its one DC calibration pot and all ranges are in spec now.

Hmm... That is some extreme IP rating testing... ;-)
And an extreme recovery too. 

[retiredcaps]

used copper wire bent in tight flat spiral pattern, with one longer end perpendicular to the plane of the spiral pattern; I then tinned the spiral and soldered it to the broken pieces of the receptacle, after which I soldered the receptacle back to the board.

Pictures?

[giosif]

used copper wire bent in tight flat spiral pattern, with one longer end perpendicular to the plane of the spiral pattern; I then tinned the spiral and soldered it to the broken pieces of the receptacle, after which I soldered the receptacle back to the board.

Pictures?

Damn, I knew I should have taken pictures while I was working on it!
But I couldn't resist the urge to see if it was all going to work (had a bit of time pressure as well) and skipped the pictures.

Anyway, I've recreated the pattern and also taken photos of the final setup (as much as it can be seen).
I've used four such wires, two per input plug (the two other input plugs got desoldered without damage in the first place).
Not terribly proud of the end result, but it works and should be ok for the currenty intended usage.
Might repurpose it if/when I get a replacement input receptacle.