Fluke 8600A ranging issues [SOLVED]

[GregDunn]

Starting a new thread since this has nothing to do with the installation of mains power supplies in my formerly battery powered 8600A meters.

Unit #2 (the one which had some oxidation on the main PCB) is pretty solid, but two small problems remain:
1. It won't settle at 00.00 DCV with inputs shorted; off by as much as 200mV.
2. it has a slight nonlinearity in DCV from about 2V up to 200V; up to 0.5V deviation from my other meters.  It reads high at lower voltages, low at higher ones.
Ω and ACV seem pretty close.  I suspect the control FETs Q15-Q21 as suggested in the manual, but can't find a good cross reference part. 

Unit #1 (clean inside) zeroes nicely with a short on DCV, but has some serious problems with the input ranging, as follows:
1. AC, DC, and Ω seem to read as though the divider is set to a fixed attenuation but the DP moves when I select different ranges.  Autorange attempts to function; K1 is definitely activating.
2. The ADC tests out very accurate if I inject a signal at TP11.
More testing is needed on the input divider, but I removed the divider, Ω and ACV daughter boards and individually tested the relays; they are functional (I can hear them activate) but the state of the contacts is unknown.  It's pretty difficult to check them in circuit but I can measure their resistance out of circuit.  It might be the relay common voltage, since that would also cause them to fail to activate; that's much easier to check since it has a common source.  One thing I did test was the PROMs, and to my great relief swapping them between the two meters changed nothing; I believe they're both OK now that I've tested the behavior more thoroughly.

So I probably need to find substitutes for the control FETs, and in the meanwhile isolate the ranging issue on meter #1.  A busy evening ahead!

[Ordinaryman1971]

Clean the range switches and function switches, they are notorious for going bad.
I've had similar problem with one of mine....
Another tip is to give it a bit of time. Spray it with some solvent of your liking, exercise the switches, but don't expect it to work right away.... let it soak and work them again.
Happy switching
Those multimeters are incredibly stable and accurate so your readings should be right on the money regardless of the age of the meter.... unless you've messed with the calibration adjustments..... did you.
And I love the 1200V range... show me one of today's meters that has this range.
One more reason to have one of those handy.

[GregDunn]

I think perhaps I mentioned it in the prior thread, but yes - first thing I do with any older Fluke once I have the display/ADC working is apply deoxit to the switches.  It's saved many a meter from the landfill.

[Ordinaryman1971]

Of course...

[Kleinstein]

The linearity error could be from leakage in the FET switches or "diodes" from the protection. Leakage going up faster than linear it not that rare.
The ranges not switching could be a stuck relay contact.


I think not having ranges higher than 1000 V has to do with regulations - different rules can apply if hight. So nobody would bother to specify a range for 1200 V, it's usually just 1000 V max, or considerably higher with special, expensive gear.

[GregDunn]

The linearity error could be from leakage in the FET switches or "diodes" from the protection. Leakage going up faster than linear it not that rare.
The ranges not switching could be a stuck relay contact.

The manual does indeed suggest the FETs as a possible cause of offset/linearity issues.  I'm just trying to figure out what devices are appropriate since the cross-reference doesn't give a useful part #.  If nothing comes up, I guess I'll just try to make a "best fit" based on voltage ratings. 

I worked with #2 last night and although the relays seem to be getting the right signals, it's unknown whether they are passing the unknown voltage/current properly through the contacts in-circuit.  It has to be the input divider plus possibly the contacts on K4, if I'm reading the theory of operation right.  DC current actually works and is accurate over a pretty wide range, and AC current is a bit low on the bottom 2 ranges but fine above that.  K4 is in circuit for ACA but not DCA, and for all ACV ranges.  K1 is the common factor for DCV, but Ω uses different relays for each range so I'm not sure on that one.  There's also U16, a CMOS mux handling some of the range selection, that I should have a look at too.

Tonight I'm going to put them side by side and try to work out the problem; ranging works OK on #1 so I should be able to find the cause on #2 by comparing voltages.

[Kleinstein]

For the FETs it looks like 2N4117 should do the job. The rather low values for C30, C24 suggest that at least Q14 is low input capacitance and thus a high R_on FET like the 2N4117.  They seem to only need some 25 V voltage rating, so BF245A might also be OK.

[GregDunn]

I will have a look at those for my next parts order!

[GregDunn]

OK, finally had time to examine unit #1 more closely.  Swapping the input divider AND the U9 ROM with the working unit made no difference.  This meant it had to be either the range switches or the U8 DVM chip.  The range switches are working, as changing range manually causes the DP to move to the expected location.  So I put the scope on pins 28-30 of U8 and... nothing on pin 30, ever.  This explains why the unit is trying to autorange but failing to select the correct divider values; pins 28-30 select one of 8 outputs from the ROM, which turns on the appropriate divider relays.  Moving the scope to the working unit #2 shows correct logic levels on pins 28-30 as the range is selected manually, and they cycle in autorange.

I thought it was really unlikely that only one of the range selector signals was bad, especially since closer examination showed a signal with no noise or activity at all; a bad output should either float or be susceptible to some glitching when other signals were active.  On a hunch,  I pulled the U9 ROM and U8 from their sockets and measured pins 28, 29 and 30 to ground.  28 and 29 measured about 10MΩ; pin 30 was about 1Ω.     Grabbed the magnifying glass and followed the trace around the board.  It went through the board a couple of times, which slowed me down as I flipped the board over to find it again.  Then, over near the power supply, was a tiny whisker of solder bridging the signal line and the adjacent ground trace.  It was so delicate that just heating it caused it to retract into the blob where it came from.  Signal trace now measured 10MΩ!

Replaced U8 and U9, and powered it up.      Ohms and DCV looked pretty good immediately.  I will now replace the AC converter (have to pull it to get to U9) and clean up some of the flux still lingering on the board , then run a full sanity check on the meter.  This makes me happy, because of the two, this one is in better cosmetic condition and zeroes with a shorted input; which gives me some confidence that the input buffer FETs are not leaky.

[med6753]

Good deal!  Me and it's 8600A kin are very happy. 

[GregDunn]

The bench is happy, too - 3 in-spec 8600A and 3 in-spec 8800A.     That's one more of each than I was aiming for when I started the latest round of DMM acquisitions.  The 4th 8600A needs new control FET(s) and a new PROM, which I'm not in a particular hurry to take care of until some other gear gets repaired.  The 4th 8800A is the one which has a defective U11 custom chip, but is otherwise OK, I think.  That one will probably require a parts donor, whereas I can probably fix the last 8600A without one.

So I'll wrap up this thread and move on to the next gear, which will be either the HP3470 or the HP3490 depending on how much space I can now clear off.