What happened to this poor Fluke 23III?

[NathanFulton]

Hello! Long time lurker, first time poster here.

I have a hobby of getting broken electronics for cheap/free and fixing them up. I recently brought a Tektronix 2235 and Fluke 77II back to life, which I may do posts about later.
(BTW, if anybody needs to 3d-print a new dial detent spring for Fluke 77II's and related, I posted my design here: https://www.thingiverse.com/thing:4858431)

I put in a low bid on a for-parts Fluke 23III on ebay expecting to be out-bid, but it didn't happen! Hooray!

Upon arrival I popped a battery in it and switched it on- nothing. Checked continuity on the fuses, one of them was broken. "This will be an easy fix" I thought... Bridged the fuse terminals with a screw just to test, still nothing. Alright, maybe not, time to pull it apart.


Removing the PCB, I see the first sign of major trouble. Seriously blasted trace from the dial to a via higher up near the LCD. I've done trace repairs before so I'm not too worried yet. I wonder where that via goes...


Oh no... Bare silicon through the "chip window" my heart sank at this point, we now have at least one exploded IC. Let's pull the plastic cover/LCD mount off and witness the carnage...


Kaboom! Somebody sure got a surprise! Look at the blast mark around the dial!


The fusible resistor appears to have tried to do its job, obviously it did not succeed... There are scorch marks on the common terminal and maybe some carbonization of the PCB itself...


Hairline crack in the other IC indicates that both proprietary ICs have been exploded.


Some of the legs of the first IC have been blown completely off and embedded into the PCB.


At this point I'm thinking this goose is completely cooked. If the two blasted ICs were easy to get, I probably would have attempted a full repair. Unfortunately they appear to be very rare and expensive, so I'm probably just going to part this thing out unless somebody has a better idea.

Any guesses as to what could completely ruin a Fluke like this? I've seen plenty of fun multimeter destruction videos and never did a Fluke fail as badly as this. My theory is somebody tried to measure a capacitor the size of a trash can and forgot to discharge it first. I hope this provides some entertainment value if nothing else!

[retiredcaps]

Checked continuity on the fuses, one of them was broken. "This will be an easy fix" I thought... Bridged the fuse terminals with a screw just to test, still nothing.

In most handheld meters, the fuses are there only for the maA/A function.  It doesn't play a role on whether the meter powers on or not.  You can power on a meter without the fuses even installed.

The input protection, MOVs, PTC, fusible resistor all look visibly bad.

[TheMG]

More than likely someone connected it to a source well in excess of 1000V, with a fair bit of current behind it too, or it was subjected to a very powerful transient.

The blown 300mA fuse probably has nothing at all to do with the catastrophic failure, likely someone blew it trying to measure too much current on that range, and just never bothered to replace the fuse. I see that a lot.

[bdunham7]

Can you wipe the red MOVs off and see if they are damaged or just covered in soot from other parts blowing?  Also, what dial position corresponds to that blown trace?

[NathanFulton]

Here are some more closer shots after I removed some components to get a better look:

Can you wipe the red MOVs off and see if they are damaged or just covered in soot from other parts blowing?  Also, what dial position corresponds to that blown trace?

The top MOV has a crack in it but the other two appear mostly okay. Most of the soot washed off with IPA but some of the darkening is permanent on all three of them.
Hilariously the fusible resistor measures 4k after removing it from the board, so it didn't even open despite the gnarly scars! (The schematic says 3.5k)
The thermistor is cracked and blown out, but still measures about 400 ohms at room temp and up to 1k if I touch it with my soldering iron. Schematic says 1.1k so I guess it still works, lol...

As for the dial, counting clockwise from the square hole, 3, 6, 7, 9, and 11 appear to have major damage on their traces. 6 is the one with trace damage to both layers 1 and 4. Layer 3 got a lot of melting underneath the switch, and it looks like both +/- of the 9V battery got in on the action too. I'm really not sure what dial position this was in when it happened... (edit: I looked carefully at slight scorch marks on the plates of the switch and aligned it with the dial. I think they had it set to DC Volts.)

Also, after removing the 683052 IC from the board, it looks like several of the pins were deliberately cut, not blown-off like I originally thought.

[Ranayna]

Wow, it even blew up the inner traces. Even with the chips intact that would appear to me to be not repairable.
Honestly, if the seller knew what has happed to this poor meter, he shows a lot of cheek even trying to sell this Do you mind telling us how much you paid?

[NathanFulton]

Including taxes and shipping, I paid a little less than $30 (my high bid). I don't think the seller had any idea. Judging from the wide assortment of things they sell I'm guessing they don't have intimate knowledge about electronics test equipment. I did ask if they tried installing a new battery and they said "no" so I knew I was taking a risk. I guess it's wise to be generally more wary of portable industrial test equipment just because of the much higher probability of exposure to the really bad stuff.

In the past I've had extremely good luck with other "for-parts, wont turn on" ebay scores, or even just randomly finding things set out next to the dumpster, so I don't feel too bad about this. It's been a good educational experience and I've already harvested some of the precision resistors for the little calibration reference board I'm building for myself.

As far as repairability goes, yeah I wont even attempt it. I've repaired damaged inner traces before by running jumpers to where the trace used to go, but that was for chemical damaged stuff. Since this board is internally carbonized, and carbon is NOT a good insulator, I wouldn't trust this with anything even if I did a full repair. I'll be selling some of the parts to possibly break even money-wise, and it does make me a little sad to see a dead Fluke, but it's been a fun ride anyway. 

[YetAnotherTechie]

Did the divider network get damaged? drifted?

[james_s]

At $30 I'd say you did really well, cosmetically it looks to be in pretty good shape, just keep watching for a beat to crap but working one and swap the guts into the clean housing.

[NathanFulton]

Did the divider network get damaged? drifted?

Z2 measures exactly 1M ohms, and Z1A exactly 10M. I'm not sure how to properly measure Z1B which has several pins...

[fordem]

At $30 I'd say you did really well, cosmetically it looks to be in pretty good shape, just keep watching for a beat to crap but working one and swap the guts into the clean housing.

When I saw the picture I was tempted to ask how much he wanted for it - I have a 26 III (the true RMS version of the 23) that I believe uses the same case & display - as far as I know mine has a fully functional board but the plastics have almost literally disintegrated - if I'd seen that auction he wouldn't have won it.

[nightfire]

As for the reason why it blew up: Probably some guy made that classic mistake of having the test leads still plugged into Amps, switching to Volts and trying to measure something which can output some very good current, like big battery packs in the industry or similar stuff.

On the other hand: Happy to see that this DMM probably has protected the user quite well so no black hands after this accident...

[NathanFulton]

When I saw the picture I was tempted to ask how much he wanted for it - I have a 26 III (the true RMS version of the 23) that I believe uses the same case & display - as far as I know mine has a fully functional board but the plastics have almost literally disintegrated - if I'd seen that auction he wouldn't have won it.

I'll let you know if/when I put this up on ebay!

On the other hand: Happy to see that this DMM probably has protected the user quite well so no black hands after this accident...

Yep! The damage to the plastic appears dramatic but it's only on the surface, everything appears structurally sound and it doesn't look like anything actually escaped to the outside of the case.

[fordem]

I'll let you know if/when I put this up on ebay!

I honestly don't know how I missed it - I'm on every other day or so looking at old Fluke for repair/parts only stuff - my 26 III was my second Fluke, which has since been replaced by a 179, so getting it fixed is more sentimental than anything else.

My first Fluke was the venerable 8020a, which I bought new in 1979, it's had a few LCDs replaced, but other than that, it's still going strong.  Although that instrument only had a one year warranty, Fluke acknowledged there was a problem with the LCDs when used in high humidity areas and shipped me a replacement display every year for three or four years until they had the problem solved, they then shipped me a kit with the redesigned display and the mounts for it.  It's rare to see a manufacturer stand behind their product that way now.

[rsjsouza]

That was quite interesting to see; thank you for sharing.
One remarkable aspect is how the meter's housing still looks new and intact (even the inside), despite the transient/overvoltage event seemed to have been quite dramatic.

[NathanFulton]

Just thought I'd share a quick update cause maybe this will help somebody else down the line: I tried something stupid last night to see if the LCD still works, and it does! 
I very crudely bridged all (most... I think I see the one I missed in the photo) of the LCD pads together, except for the top-right two that are the common, and sent about 3v 100hz AC through it.