I'm miffed.
I've been looking at my Keithley 230 programmable voltage source, and it is
almost very good. It generates the right voltages, but not completely reliably.
If I select 10.000V output, and the output is measured at, say, 10.001V then I'm happy.
If I then drop it a couple of cm onto the bench, the voltage shifts out of spec, e.g. from 10.000V to 10.012V. Possibilities:
- the wretched Bourns 3386 single-turn trimpots are shifting. But looking at the schematic, a quick calculation indicates that it would have to be a relatively large shift
- some socketed metal film resistors in the current limiting part of the output driver
- some Coto relays in the same circuit
- another relay which operates to disconnect the output when there is no power
- a switch for internal or external voltage sensing
- the socketed DAC08 inckuding reference
Reseating the obvious components with Deoxit D5 does not improve matters.
If I then change the output from 10.000 (i.e. 10.008V) to 100.00V and back again, the output voltage returns to almost 10.001V. Now there's a range change there, but what is causing the "hysteresis" in output value?
I suppose I'll have to stop now and find some precious treasures to flog on fleabay; thanks bd for pointing out the £1 fees.