RadioShack FET VOM repair has me stumped

[RichardM]

Hi All

I have an old Micronta 22-208 analog FET voltmeter that works perfectly except for resistances on the highest 1M range which measure high by about a factor of 2. I cannot for the life of me work out why.

I have checked the 9.9M resistor R19 out of circuit and it works fine. The S1 switch “fingers” also seem to make contact with the right pads on the dial pcb.

I thought I would try to measure resistance between the COM terminal and the battery positive terminal whilst rotating the S1 dial between the resistance ranges. I measure the correct values of about 10ohms, 100ohms, 10kohms, 100kohm on the lowest ranges but only 2.17Mohms on the 1M range?

Any help appreciated. Scratching my head at the moment.

Richard

[xavier60]

Can R19 be measured in circuit? Maybe a conductive track has formed between the contact pads of the switch.

[MathWizard]

It's a cool circuit, I started to make the circuit, but IDK if LTSpice has any dual JFET's like that. But later I want to use regular JFET's, and alter whatever, to match the input/output range of the JFET's as a diff-amp.

But yeah I've never tried to make a pure meter for V or A or Ohms. I would also like to try making a simple deflection meter, with some coil, even just wrapped around a compass.

One of these days...

[floobydust]

Measuring DCV, is there symmetry in readings between switch S3 in Polarity(+) or (-) ?
i.e. DCV inject almost a volt on say the 1V range, compare what it reads with Polarity(+) verses leads swapped and Polarity(-).
Just to see if one JFET side has some excess leakage current on the gate circuit cause a loading error.
If that is the situation, it would explain R19 appearing low value and Rx1MEG reading low as well.

[RichardM]

Thanks for the replies.

In circuit, R19 measures 9.9M right around the dial EXCEPT when on the 1M range where it measures about 4.8M ?

The DC voltages are symmetric when changing the polarity switch which is a relief as the jfet may be fine?

If I measure between each of the pads for the ranges and the inner arc shaped pad on the switch (one with the 1.5V) I get around 3.68M ohms. If this is in parallel with the 9.8M resister then that would give me the 2.17M I am reading between COM and the battery positive terminal. The 3.68M in parallel with the other resistors won’t make much of a difference. Just some random thoughts from probing around

Richard

[RichardM]

Switching the polarity switch from +DC to -DC the measurement across R19 changes from 2.17M to 3.3M ?

[xavier60]

Has the 1.5V cell ever leaked. The electrolyte can wick along stranded wires and cause contamination at the far end.

[RichardM]

It is a new 1.5v cell

Also the R19 measurements do not change when the battery is removed?

[xavier60]

A BTW for anyone suspecting electrolyte contamination, it usually causes small voltages to be measured in affected circuitry even though all external power sources have been removed.

[floobydust]

If you look at the schematic, the entire metering section, JFET, 9V battery is floating and isolated from the rest of the multimeter (divider strings, shunts, rotary switch etc.)
Only Points G, A, E between them.

It seems to be around 2.7MEG worth of leakage yet R25 is 5MEG so what's with that. But changing the Polarity switch and getting 3.3MEG for a different leakage resistance of near 5MEG now seems baffling as well.
OP what type of multimeter are you using, to know its Ohms voltage and test current etc. if it is throwing things off.
Point "G" should be high resistance unless it's pushing enough to bias the JFET junction or (reverse) turn on zener D1. Check R25 is 5MEG.

Ensure there is no lithium grease used on the rotary switch, that is a no-no. Lithium is a salt and can be conductive lube. I would clean the PC board with Q-tips and IPA, get rid of old flux, soot, carbon etc.
I think it's hard to tell if a component's value is low or it's contamination.

[schmitt trigger]

I also had this same FET-VOM.
I loved it, but for a weird reason, the plastic pieces started to rot away. Disintegrating
into plastic dust.
First to go was the back case, but I could somehow mend it. That is, until the selector switch simply fell apart.

[RichardM]

I have an assortment of meters:) but have been using a Fluke 87V, HP3468a and HP3478a. I also have 2 HP-410C I could try.

I have cleaned the rotary switch contacts well with IPA and a light sand with a fibreglass pen but will look at cleaning the meter more widely. Maybe the polarity switch is the culprit? I will report any changes.

Cheers

[RichardM]

Ok, so it is not the polarity switch. I desoldered it and hard wired the polarity and I see the same behaviour on the 1M range.

[floobydust]

How is R25, 5MEG?
With an eagle eye look for any crud, residue etc. anywhere inside the multimeter, as if a battery took a piss etc. Either you have contamination somewhere, or something is being glossed over. Look at the little JFET board too, it must be spotless. The wires along "G" as well.

A high Z multimeter i.e. HP3478A up to 3V range is 10¹⁰ ohms, you can measure the voltage drop across R25 as I keep thinking the input resistance is lower than it should be. 34401a has a high Z mode as well. Otherwise, the switch row is somehow leaky.
On Rx1MEG range, I would connect another multimeter and measure current at the test leads.
1.5V in 10MEG is only 0.15uA it's very little current in the first place.

[RichardM]

I have been focussing on the little FET board. I think this may be where there is some contamination or something else. I have pulled the dual FET out for now.i have scrubbed it to death. I notice that touching and spraying alchohol around D and R25 while measuring across R19 send the resistance crazy.

R25 tests fine out of circuit. If I measure the resistance between the cathode of D1 and poitG on the other side of R25 I measure 2.7M one way and 4.8M the other way. The 2.7M is when I have the COM on the cathode. This is with the FET removed.

I will try to take some measurements ,nets as you suggest.

[RichardM]

If I lift the selector pcb away from the three finger contacts and then measure between K and the V-Ohms-A terminal I measure open as I rotate the selector dial until I select DCV and I get continuity. This is fine. Problem is if I reconnect the selector pcb I get Isee continuity ins Amps mode, 10M in ACV and about 1.8M in Ohms. What a nightmare.

[schmitt trigger]

Could it be that D1 is leaky? Can you read its part number from the diode's body?

[floobydust]

When a repair goes off the rails like this, it means one's technique is bad and there's an opportunity to learn and grow.
My usual mistakes are making assumptions and overlooking the obvious. Back up and redo things maybe.
We haven't narrowed it (leakage resistance) down between being a component or a mechanical problem.

Again, I would do a visual with a magnifier and look for anything. Maybe that phenolic is simply turning to carbon.
The rotary switch wiper could have leakage, pull it off and test it between poles.

Another technique, when you can read the leakage resistance is to apply a little hot air to heat up parts or regions to narrow it down. Not to melt things but just lift temps up to say 50-80C and see what affects the leakage resistance. It will jump dramatically with temperature.

I keep guessing it's the input clamp- D1, C1 or the JFET.

[RichardM]

Thanks

Yes, I am the first to admit my troubleshooting can often head down a rabbit hole but I have learned a lot over recent years and managed to repair a lot of old test gear successfully. I generally follow the generally accepted approach to fault finding. In this instance I just couldn’t understand or explain the fault. As it turned out it was a mechanical issue with the front selector dial. Depending on how you twisted the dial and applied a little pressure you could get the 1M range working fine. The meter has had a tough life and when I pulled off the front selector dial and it was apparent it had broken in he past and a bodge was done with glue and a new central screw and bolt. The dial was always a bit how do you do. Anyway, thanks for the suggestions. I have parted it out. It didn’t owe me much.

Richard