Keithley 2100/Array M3500 –750 VAC range bolloxfied!!

[Chris56000]

Hi!

Can anyone offer me any tips for repairing the 750 VAC  Range on my Keithley 2100/Array M3500 4½/5½/6½ digit multimeter please?

There are no p.o.s.t. errors displayed and all the DC and low AC voltage ranges up to 100 VAC range appear to be operational!

When I connect it to mains on "Auto" the meter constantly changes ranges between 100 VAC and 750 VAC every second, displaying only dashes with no numerical readings whatsoever, and on "Manual 750 VAC" it reads "63.899 VAC" (4½ digit setting).

If you then disconnect the mains from the input it stays stuck at "52.4 VAC!"

I first became aware of this fault trying to check the 480 V secondary in a "Mullard High Speed Valve Tester", the meter displayed "131 VAC" which stuck there, and I knew that can't be right because the green spot on the Mullard tester was present, this Keithley meter showed "422 V" on the positive side of the CRT supply and "–525 V" in the negative side of the CRT supply!

The 480 V CRT h.t. secondary is ordinary 50 Hz a c., not a pulse or oscillator type supply!

Rechecking across a convenient valve heater in the Mullard Valve Tester gave me a reading of "6.2532 VAC" on the '2100!

(If there's a way to get rid of all those daft decimal places I haven't found it yet, one or two decimal places is more than enough for me!)

Before anyone asks why I'm using a Lab Quality 6½–digit meter on an old Valve Tester, the answer is all my other test meters need repairs, and the Black Star 3225MP that (did!!) once work has now got ink leaking all over the LCD so that's a goner as well !

Chris Williams

[Kleinstein]

The Keithley2100 is still relative similar to the older 2000 model.
A failure only in the 750 VAC range is strange, as the input part is the same for the 100 VAC and likely also 10 VAC range.
As a first test one could check also all the smaller AC ranges if they read OK with a small 50 Hz voltage (e.g. with a voltage divider) and short.

Looking at an open input does not make much sense, as the input can pick up mains hum. The zero for a voltage reading is with a short.

Still getting 52 V AC with an open input looks suspicious high, like something oscillating or maybe supply ripple.

The Keithley 2100 is more like a low cost precision DMM, not a meter normally used for mains and similar high voltage. There are compromises between low noise and good protection against high voltage spikes that sometimes happen with mains. The CAT rating of the bench meters is thus usually not that great. Unless you need the accuracy a robust handhelp meter may be the better choice. 

[Chris56000]

Hi!

It certainly could be excessive ripple or oscillation due to electrolytic problems, so I'll have a look at them!

Fortunately the Array M3500A S.M. gives a full parts list/B.O.M. for this meter (alas no schematics tho !), so there's no need to make a special note of what I remove !

I'll have a look at the electrolytics and report back !

Chris Williams

[Kleinstein]

AFAIK the circuit is still very similar to the K2000 - even many parts designators in the front end stayed the same.

[coromonadalix]

here's some reverse schematics done of k2000     was found on Xdev's

a Picotest 3500  should be a brother too,  serv manual w parts  but no schematics  can or could help

[shakalnokturn]

There's a short, not so old (2022) topic where someone reported changing cheap electrolytics to solve an ACV problem on these bench meters.

So you're here to remind me I should do this to my 2100 as preventive maintenance...