Fluke 89 IV - repair after input overload

[ve7xen]

I got what I thought was a good deal on a 'working' Fluke 89 on eBay. Turns out that many of the input protection devices have been blown (see attached photo). It doesn't look like this would be too hard to repair (there doesn't appear to be any damage to the traces or upstream parts), but I'm not sure what replacement parts would be suitable, or if it's just plain a bad idea. Any advice?

[vtl]

Does it power on? Hopefully the overload didnt reach the processor.

I suppose the most obvious thing to do is start desolding all the blown components and soldering in new ones.  Just look at the component markings to find out what they are and you can decide to either desolder it and leave it (MOVs only short under HV) or replace it (input resistors).

They don't have to be up to the same spec, you can just solder in anything functionally equivalent. You probably would rather verify the meter is working before ordering high quality replacements.

[samgab]

I'd start with the two resistors that are either cracked or visibly discoloured...

[ve7xen]

The meter does power up and seems to work otherwise. The current ranges seem to produce reasonable measurements as well, though I haven't tested it extensively.

So far I've determined:

The 10R resistor has gone open
The 100K resistor measures OK, but I'd replace it anyway
The 1K SMD resistor has also gone open

What I'm not sure about:

The part that has blown in half - I think this was a MOV? Thus the meter should work with it open, yes?
The yellow part beside it - is this also an MOV? It appears to be damaged, but if it is open I think just replacing the resistors should restore operation.
I didn't realize this was a 4-layer board and there is what looks like could be internal damage, but it's hard to tell at this point. A schematic would be helpful...

Thanks for your advice!

[vtl]

Service manual on the fluke website is pretty useless. Only has calibration procedures so no schematics unfortunately. The thing thats blown in half I think is a PTC resistor, I could be wrong. If its open, short it out with a piece of wire and see what happens.

[ve7xen]

Yeah, not much of a service manual. I guess it's like most things these days where when it doesn't work they just replace the entire board (or meter...).

I haven't really made much progress, basically where I started. Volts range stays near zero, though does fluctuate a few digits now and then. Capacitance range reads anywhere from 2-10mF or OL - placing a capacitor on the test leads does *something* as the value settles to ~3mF, but this value makes no sense. Ohms settles usually around 2.5-3K, but it does some weird stuff at startup.

Temporarily replaced the 10R and 1K resistors. Shorting that PTC doesn't seem to change anything. I found what I think are some burned traces from the input socket to the 100K resistor and then from the 100K to the pad beside it - when held up to the light this part of the circuit board is traceable, and seeing no traces and some signs of damage there that was all that was logical, so I shorted those. Also removed the two red components, which I think might be NTC resistors, though not too sure. One of them reads ~6K with a pretty strong tempco, the other reads open (presumably dead, but I'm not sure). I tried to find some information on the markings on these, but they aren't very meaningful and I wasn't successful - no idea what they are or what their nominal values should be. The yellow component is I think a spark gap, and reads open as one would expect.

It's really a pain trying to figure out this input stage without a schematic. The best I found was a Fluke patent from 1993 that looked kinda similar, lol.

Had enough of prodding it for now, but would appreciate any suggestions. As it was sold advertised 'working' I'm going to try and get my money back...

[alm]

No specific hints. Fluke often used fusible resistors designed to blow in case of severe overload. Be sure to replace them with similar resistors, otherwise you might lose this safety feature. The Fluke 89 IV is a very close relative to the Fluke 189, so if you can find any information on the 189, this may be usable. Didn't Dave at some point claim to have a partial schematic of the 189?