Micro/Pico Amp Power Supply.

[Patrick]

Hello,

I'm needing help with a Micro/Pico amp power supply. I'm thinking I should buy one for $32,000,000 since the tax payers money is free, but the one I'm supposed to fixed is powered by 7 AA batteries and was originally used in the 1970s. The physicists want to use it to calibrate things but I'm not sure its worth fixing or how to just build another with all new passive parts.

Should I just duplicate the circuit ?

Pics attached below. It looks like 90M, 900M, 9,000M, 90,000M and 900,000M ohm resistors arranged by a dial.

Ive been using a Nano/Pico/Femto ammeter but without the shield on its hard to make heads or tails with all the radio induced garbage.

[Twospoons]

Is it the switch that's damaged? Those contacts look worn in the photo.  And the wafer is probably a bit dirty, and therefore leaking current.

Could replace it with a set of reed switches operated by a magnet on the switch rotor.  All nicely sealed.

[davekni]

Is it just resistors in series with a voltage source varied by the potentiometer?  That is what the pictures look like.  Or is there active circuitry also?
The glass resistors look exactly like what I salvaged from a broken Geiger counter back then (1970s).  Discovered that those resistors are not good for high voltage.  Must be inert gas inside.  Arcs across internal element long before any atmospheric-pressure arc would form.

The higher ranges shouldn't be too difficult to measure.  A DMM with 10M input resistance and 10uV or even 100uV last digit resolution would give some reasonable indication of 1nA (10mV across 10M).  I've done that before.  For smaller currents, I've used a film cap, let it charge for a measured number of seconds, then connected a scope probe to measure the charge.  Fast enough to capture peak voltage before scope probe impedance discharges cap.

[Patrick]

There is no actives, just passives. Im trying to think about just what the schematic looks like. It hurts my brain, 10^-6 all the way to 10^-12

[davekni]

I'm thinking I should buy one for $32,000,000 since the tax payers money is free

here is no actives, just passives. Im trying to think about just what the schematic looks like. It hurts my brain, 10^-6 all the way to 10^-12

Since I'm one of those tax payers who will save part of the $32,000,000  here's the schematic:

[Patrick]

If you look at the dial, I think you'll see that each value is shorted except the one being used. I think that is to prevent the unused resistors from acting like radio antennas.

Maybe?

[davekni]

If you look at the dial, I think you'll see that each value is shorted except the one being used. I think that is to prevent the unused resistors from acting like radio antennas.

Yes, you are right.  Something like that is going on as I look at the image again.  I'd guess it is more to short out any leakage current through switch.  Unused resistors are likely shorted to "ground" rather than across resistors.  You could check connectivity with a meter.  Even if the switch is bad, hopefully at least some of the connections would still show up to verify which node resistors are shorted to.

[the_anomaly]

Do they know its not a regulated output?  I may be off on my theory here but the load could affect the output.

I would add an extra battery or two and then use a precision voltage source to feed the resistor divider (https://www.digikey.com/en/products/filter/voltage-reference/693?s=N4IgTCBcDaIE4FMBmDEDsDGCQF0C%2BQA).

[Patrick]

Do they know its not a regulated output?  I may be off on my theory here but the load could affect the output.

They use it to calibrate electrometers, so hopefully they have been taking such things into account.