Keithley 2000 Measurement Anomalies

[reinardvanloo]

Hi All,

A bit dumbfounded by the following. Using my Keithley 2000 to measure some spots in a synthesiser I am repairing (Korg PolySix, RAM battery area for those interested) and am getting some weird readings with my Keithley 2000 that do not match expectations nor measurements with a handheld multimeter (Agilent U1231A). Pls refer to the diagram with an overview of the problem.

I am lazy, so when measuring often I do not care about polarity: reversing the + and - probes should only change the sign, right?. So when I check with my Keithley 2000 the voltage across the RAM chip that should hold internal settings when power off , I measure with reversed probes -4.1 volts huh? Now put on the probes correctly: +3.9V. (should be between 2.1V and 3V)

After some fooling around noticed, that the 10V range of the keithley has 10Gohm input impedance.... tried the 100V range (10Mohm) and all voltages measure as expected and are corroborated by the hand multimeter. Back to the 10V range: still weird values, also when reversing probes.

Started measuring my lab power supply: measures ok, also when reversing probes. Measured another spot in the circuit (across the PSU caps, they still hold some charge in power off state) and that measures also ok in the 10V range,  also when reversing probes.

I did a self test of the Keithley 2000, failed initially at the 300.1 test. After cycling the front/back switch multiple times this fault did not appear anymore.

Q: Is my 2000 faulty or am I misunderstanding something on how the 2000 measures in 10Gohm impedance mode?

regards,

---Reinard

[The Soulman]

Interesting, could there be any noise/ripple on the 3.0V rail?

You could try measuring a battery with the keithley, if the reading is then still off compared to the 100V range and the agilent, something is likely wrong with the 10V range...

edit:

didn't read this:

Started measuring my lab power supply: measures ok, also when reversing probes. Measured another spot in the circuit (across the PSU caps, they still hold some charge in power off state) and that measures also ok in the 10V range,  also when reversing probes.

So I'd expect noise on the rail or your wiring setup.

[Kleinstein]

If the voltage source measured is really high impedance and nonlinear (e.g. like through diode leakage current), it can happen that the high Z mode (10 V range) can give odd looking readings. This is than the fault of the circuit. In high Z mode there can still be some input bias current of several 10 pA and with a near open circuit this can make a difference on reversing the leads.

Things can get worse than normal if the DMM has higher than normal input bias, e.g. from contamination of maybe a slight damage from ESD of similar.  One can relatively simple check by looking at the voltage drift with a good (e.g. PP or PS film type) capacitor of some 10 nF at the input. The slower the voltage drifts in the 10 V range the better. 10 pA and 10 nF would be some 1 mV per second.