Keithley 2410 Labview driver/trigger problem

[Witch2000]

Hello,

I'm trying to set up a Keithley 2410 sourcemeter to perform just the following task: Output a specified voltage in the range 0-600 V when it is given an external trigger. The device gets a 'arming trigger' every 5 seconds or so, and a 'trigger trigger' to advance to the next voltage from the list every 1, 10, or 20 milliseconds.
In order to do that, 2410 is run by a custom Labview driver made by using the NI driver library for the 2400 series.
The problem is that the voltage output is always 200 microseconds late after the trigger for advancing to the next voltage. This is too much for our application.
Is there a way to make the device output the voltage exactly when it gets the trigger?

Just to make things clear, the 'trigger delay' is set to zero, and the 'timer delay' in the trigger configuration VI is set to the lowest allowed value of 1 millisecond (if the latter is relevant at all). Also, I've tried to give the device just the arming trigger, and then set the voltage advance trigger to Immediate, with a trigger delay parameter set to, say, 20 ms. This also gives a mismatch between the expected advance every 20 ms, and the real advance (as checked on the scope) which comes every 20.8 ms.

Thanks for any ideas!

[johnsold]

I wonder if you are not seeing limitations of the instrument.  From the data sheet it appears that the slew rate is 0.5 V/us, the command processing time is 7 ms, and the output settling time is 100 us typical.  I did not see a specification for response to triggers.

If you chage the voltage by 400 V, the slew rate alone would cause the output to take 200 us to reach the new value.

Are you using GPIB, RS-232, or digital triggering?

Lynn

[Witch2000]

The triggers are connected through the Keithley 8502 trigger link adapter.

The interface is GPIB, but nothing is triggered by the software or through GPIB - the instrument drivers just set up the 2410 to receive triggers through the 8502 box. There are no commands processed continuously, the Labview program is just run once to set up the device, which is then triggered only through the box.

Sorry, I forgot to mention that this mismatch of 200 us is measured on a sequence 0-10 V, where each trigger increases the output by just 1 V, so the slew rate shouldn't cause this.

Anyway, after writing the paragraph above, I found the problem. The 8502 box has a trigger delay of 100-350 us!!! It never occurred to me that a simple trigger box could have such a huge delay.

Thanks for the help!

[johnsold]

I am glad you found the problem.

Many power instruments were never designed for speed. I suspect that the designers tend to think that anything faster than a few milliseconds is esoteric.  Plus, it is hard to design circuits with the voltage and power ranges of the 2410 which will work at microsecond timescales and maintain accuracy.  Measurements are relatively easy, but programmable sources are not.

Lynn