Keithley 2015 Stray Voltage

[sparknmike]

I just picked up a Keithley 2015 to add to my collection of useful junk.  It was supposed to be working but it wasn't put through the paces to prove it works fine before I took it home.  I am finding that so far it does seem to work but it does something odd that I haven't seen before.  When there is nothing connected to the input the DCV reading slowly runs from zero to about -11.8V.  If I set the scale to a lower range that doesn't over -11.8 it will eventually show an overload.  If the inputs are connected to each other it does read out at zero.

Is this normal for these meters?  It seems like it is it just a floating input slowly building a charge.  Since I have never seen a meter do this before it worries me a little that this is a minor fault and not an odd side effect of the high accuracy input circuitry.

[don.r]

I think someone else here had the same problem with their 2015 or it could have been a 2000 and solved it by placing a high value (30Mohm?) resistor between the inputs. If I can track down the thread I'll post a link to it here.

[ve7xen]

I don't have a Keithley 2015, but my bench meters do this on low ranges. With input impedance of 10gigohm, they quickly accumulate charge and have no path to bleed it off. I suspect the meter is probably working fine.

If you put the meter on its 100V range (10megohm input impedance, per the datasheet), do the spurious readings still appear?

[amspire]

When there is nothing connected to the input the DCV reading slowly runs from zero to about -11.8V.  If I set the scale to a lower range that doesn't over -11.8 it will eventually show an overload.  If the inputs are connected to each other it does read out at zero.

Sounds right to me. You are just used to meters with a  10MOhm input impedance. For the 20V DC and lower ranges, the input impedance of the Keithly is extremely high - you would have a lot of trouble trying to measure it. This means it can be used to measure the output precisely from sources such as resistive dividers without causing any significant error so it is a very good thing.

[EEVblog]

Is this normal for these meters?  It seems like it is it just a floating input slowly building a charge.  Since I have never seen a meter do this before it worries me a little that this is a minor fault and not an odd side effect of the high accuracy input circuitry.

Yes, completely normal for meters with Gohm level input impedance.

[sparknmike]

Thanks!

The 100V scale doesn't quite settle on zero.  Scales above 100V zero out though.  It looks like all is well.

I didn't realize the input impedance was in the GOhm range.  It wouldn't be hard to build up a charge with that much resistance.  To have a little fun I blew across the input terminals and indeed the reading went down temporarily.