Fluke 8600A repair (?)

[AstroNewbie]

Hello,

A newbie here.  I'm trying to fix an old Fluke 8600A bench meter.
It's the version with batteries.  After replacing some of the Nicd
cells, the meter works well on the 200mV and 2V
scale. 

Things go funny on 20V scale and up.  For example, when I tried measuring
a regulated 13.7V and the measurement would bounce around
and is usually above the actual value.  Once in a while, it
would snap to the right value.  The happens seemingly randomly
with both auto ranging and manual ranging.

Going through the instruction manual, it seems some of the transistors
in the input divider section may be leaky or shorted.

The parts list show these transistors as:

XStr, Si NPN Selected  352138
XStr, J-fet, N-channel  F2692
Xstr, J-fet, N-channel  F2691
XStr, J-fet, N-channel  KE4393


I have not been able to find anything on the first 3 parts.  Any
information on these parts or suggestions for replacements will
be greatly appreciated.

I am in search of a new hand held multimeter anyway, so I don't
want to spend too much trying to fix the old Fluke.  If replacement
parts are too expensive or difficult to come by then perhaps I
should just abandon this.

Thanks in advanced for any help.

Cheers.

[wn1fju]

Before jumping to possible false conclusions, check the obvious first, namely the various switches.  Almost all DMMs that old can usually benefit from a shot of cleaner into the switches.  This ultimately may not be the problem, but it's a good place to start.

Try the thing on the other functions, like ACV and Ohms.  If things seem erratic everywhere, I would start with the switches.  Many of these DMMs have a nasty habit of running the inputs through various switches sequentially and everything has to be right in order for it to work properly.

(Note: my 8600A was erratic only on ACV and that actually was a defective JFET being used as a simple current source.  But I would say mechanical failures outnumber electrical ones by a large margin)

[ModemHead]

But I would say mechanical failures outnumber electrical ones by a large margin

A very large margin.  Always check the simple stuff first.

[RJFreeman]

Things go funny on 20V scale and up.  For example, when I tried measuring
a regulated 13.7V and the measurement would bounce around
and is usually above the actual value.  Once in a while, it
would snap to the right value.  The happens seemingly randomly
with both auto ranging and manual ranging.

OK my memory was rather hazy (it has been a while since I serviced a Fluke 8600) so I had to look it up and confirm - according to the manual (and from my hazy memory) the input divider uses Reed relays and I suspect the Transistors and Jfets to which you refer (you did not provide reference numbers, e.g Q11, Q14 etc) are used by the A/D converter so I would avoid poking around in there (unless you have to).

In the 200mV and 2V ranges the Input divider board is bypassed by Relay K1 (located on the main board), and you say that these ranges are OK, so that would suggest the A/D is working fine.

20V and above are switched out to the voltage divider board (by K1) where the higher ranges are selected by Relays K2, K3 and K4 which is where I would be looking;
check for dodgy joints to the voltage divider board, dodgy resistor (R3, R5, R7) or trimpots (R4, R6, R8)  in the divider chain - as any one of these will affect any of the higher ranges. and of course you could have noisy contacts in K1....

[RJFreeman]

Another thought, I wonder if the contacts to the range control prom could be noisy, try easing one end of U9 up in the socket, then push it down and do the same to the other end....

[AstroNewbie]

Thank you.   I will play around with the switches and other mechanical
aspects when I get a chance this weekend.

Incidentally, when I measure the voltage at U4 pin 3, (with no input),
I do not get 0V, but it bounces around within maybe a +/- 0.5V
range.  (I couldn't even find TP11. ) (Please see attached, section b)

Thanks

[RJFreeman]

hmmm,
look a lot of the ADC etc is usually socketed and given the age of the 8600 I would start by simply lifting all of the socketed ICes and pushing them back in again, usually just lever up one end (not so much the pins come out) then push it back in and repeat with the other end of the IC.

For gear of this sort of age (I often have old, 'classic' audio/digital effects in) just reworking the contacts like this often clears many faults.....

Just trying to find the section of instruction you posted in reference, since there is all sorts of switching going on around the front end of U4, flicking between +/- 0.5V may be normal or at least not mean a leaky transistor, I suspect that they mean a permanent offset rather than switching.

[AstroNewbie]

Hi,

The snippet was from section 4-45 on the input divider in the version of the manual here:

http://assets.fluke.com/manuals/8600a___imeng0300.pdf

I'm happy to hear the fault may be of a mechanical nature rather than with unobtainium parts.

Does the fact that once in a while, the reading on the 20V scale is correct, further point to a mechanical problem?

Thanks.

[RJFreeman]

The snippet was from section 4-45 on the input divider in the version of the manual here:

http://assets.fluke.com/manuals/8600a___imeng0300.pdf

Interesting, my Manual is dated November 1974, as is the one you have linked to, although mine is revision 2 while yours is rev 3....
I am guessing that they were probably frantically doing the documentation around that time.

I'm happy to hear the fault may be of a mechanical nature rather than with unobtainium parts.

Emphasis on 'May Be' can't say for certain of course, but if the 200mV and 2V ranges are reading correctly then it looks very promising....

Does the fact that once in a while, the reading on the 20V scale is correct, further point to a mechanical problem?

I would be basing my opinion on the 200mV and 2V ranges being correct (you may want to double check that), if that is the case then the fault is back towards the divider, and you should be able to obtain equivalents for the precision resistors, pots (come to think of it try cleaning/moving the pots, this will change calibration, but a 40 year old meter will need calibration anyway) or reed relays without too much trouble....

[nightmechanic]

Hi,

some questions:
Did you measure the power supplies?
Did you measure the power supplies when the miss-behaving ranges / modes are used?

going back to mechanical / contract issues: did you try re-seating the divider board?
have you seen the attached table in the manual?


can you fill the table for your multimeter?


Ran

[AstroNewbie]

The meter is at my mom's place and I won't be back till the weekend.  So, I
haven't done any of the cleaning/reseating of parts as suggested.

When I was playing with it last weekend, the power supply rails look ok
after I replaced a Nicd cell.  That said, I didn't measure the rails while the
meter was taking measurements.  I will try that.

I don't have any AC source (aside from the wall socket and I wasn't going there).
I ordered some basic parts for testing the resistor and dc current values.  Hopefully
they will arrive this week.

On a high precision 1k resistor, the value looked ok.

Thanks again for all the suggestions.