eBay Fluke 8842a

[bigtexun]

I just purchased a Fluke 8842 DMM from eBay.   It is functional, and seems to produce good readings.  I'm working on getting some calibration standards together to give it a good test, but in general it seems fully functional...

But I'm noticing strange behavior, and I wonder if it is a common/known/normal behavior.  When powered up with an open corcuit on the test leads, it comes up in DC volts auto ranging mode, and displays a small positive voltage, typically around .4mv DC.  The display updates at about .5hz, and the values change in negative increments until it reaches zero, then starts reading negative voltages.  If you short the leads, it jumps to near zero (.005mv dc typically)  But left open, the display updates every half second, drifting slowly toward more and more negative values, slowing down on a gradual basis.  After a couple hours it is around -50mv and after 20 hours it is at .800 mv with the drift slowing down to less than .001 mv change every 30 seconds or so. 

It is acting like there is an internal voltage source just above .8v slowly charging a capacitor over the course of 20 hours.  I've seen DMM's that drift around when the probes are open, but they usually just display random low values, they don't usually have a directional drift, and .8v is an awfully high and specific value.  If I place it in a circuit, the measurements are rock solid and look accurate.  So I'm happy enough with the meter, until I get my reference standards in to see how well calibrated it is... 

But my question is this open circuit slow drift to -.8vdc, is that normal?  Or is that something someone has seen before?


Thanks!
George

[z01z]

Meters with high input resistance behave like this. If you check the specs, 20V or below ranges have >=10G, while 200V and over have 10M.
Try the same on the 200V range.

[Kleinstein]

The behavior is normal.  In the high Z mode there is still some residual input bias current from the input amplifier - usually in the +-50 pA range. his small current slowly charges the input capacitance (typically some 100 pF to 1 nF).
The current can depend a little one the actual voltage so that there may be settling to a final values, but this is more like unit dependent and many meters continue drifting all the way to the range limit.

With some extra capacitance (e.g. 10 nF - should be low loss like PS or PP film type) one can use the drift rate to measure the input current. This is one of the test one can do to check the meter input for too much input current.

[BobRyan]

I have the same meter and it behaves the same way.