Instrument Control (GPIB, Serial, VISA, IVI)

cancel
Showing results for 
Search instead for 
Did you mean: 

Keithley 2400 does not source the desired voltage

Hello all,

 

I am using LabView to perform a sort of software-controlled linear voltage sweep with a Keithley 2400. The problem in short is that the Keithley does not always output the voltage it claims to be sourcing in the display. It has _not_ reached its compliance value when the problem appears.

 

The procedure is as follows: suppose a sweep from -10 to +10 volts in 21 steps (-10,-9,-8...9,10)

 

Using the Keithley 24XX driver: 

First I send reset and initialization commands

Then I set the source to generate the initial voltage (-10V)

Then I set the device to measure current

I configure it to expect 1 software arm and trigger immediately upon receiving it

Next I turn on the source and wait  several seconds.

 

After few seconds, i send a software trigger and acquire the measurements.

I adjust the source voltage to -9 volts and re-arm the sourcemeter.

Next I wait several seconds again and repeat this process until 10 Volts is reached.

 

Now, the problem: The first half of the sweep, the sourcemeter outputs the right voltage, but after crossing 0 volts, the sourcemeter tells me it is sourcing 1 Volt, but it actually outputs only about 300 mV!  Then, 2 and 3 volts are sourced fine, but 4 volts results in a too low output again!

 

Does this sound familiar? I have checked the low-level text commands that are sent by the driver and they seem to be okay (no decimal separator problems or anything). I have also mailed Keithley this same question.

 

Thanks - DJ 

 

 

 

 

 

0 Kudos
Message 1 of 3
(5,715 Views)

AH- you are auto ranging and crossing quadrants (-V to +V) changes from I source to I sink.

 

The 2400 doesn't like that- specify the range and the compliance current (+/-) for each step.  

 

Alternately, isn't there a step functions available with the 2400?  (mines not right in front of me)  check the examples in the driver to see an implementation that Keithley has made work right.


"Should be" isn't "Is" -Jay
Message 2 of 3
(5,712 Views)

Ah, I expected this was a ranging problem, tried different things to solve it, but I hadn't tried to specify the range and compliance each time.

 

Performing a programmed sweep is not possible: I want to acquire the data after each measurement for a 'real-time' display. It also enables me to abort easily without losing already recorded data.

 

Thanks for now! 

0 Kudos
Message 3 of 3
(5,693 Views)