Help reading buffer in Keithley Electrometer 6517B through SCPI

[gr2]

Hello,

I've been struggling to make buffer voltage readings on a Keithley electrometer 6517B. I want to do it with SCPI commands through a GPIB interface.

I have used the following commands on the Visa Interactive Control:

*RST
:SYST:ZCH ON
:FORMAT:ELEM READ
:FORMAT:DATA SRE
:SENSE:FUNC 'VOLT'
:SENSE:VOLT:RANGE 20
:SENSE:VOLT:NPLC 0.01
:SENS:VOLT:AVERAGE:TYPE NONE
:SENS:VOLT:MED:STAT OFF

:DISP:ENABLE OFF
:SYST:ZCH OFF

:TRACE:FEED:CONT NEVER
:TRACE:CLEAR
:TRACE:POINTS 100
:TRIG:COUNT 100
:TRIG:DELAY 0
:TRACE:FEED:CONT NEXT
:INIT

:TRACE:DATA?

My problem seems to be how slow it is. If I increase the number of points from 100 to (for example) 1000, I get a timeout expired error (for a timeout of 30 s). I have used a similar code with the 6514 model and no such issue happens even for measures of 2500 points. Any help would be appreciated.

I have seen the Labview drivers but they don't seem to work either.

[santo_13]

When you say, "My problem seems to be how slow it is" , do you refer to the data fetching or the time the instrument takes to capture the points?

The former depends on the data format and communication bus, whereas the latter depends on instrument capability and settings.

[gr2]

I believe it has to do with how it captures the points which should mean there is something wrong with how I configured it.

[santo_13]

---

@gr2 wrote:

I believe it has to do with how it captures the points which should mean there is something wrong with how I configured it.

---

Have you tried exploring the programmer's or user manual of the instrument to see if there are any best practices to optimize measurement speed? or to verify if your implementation is the optimal one?

[gr2]

I have tried, yes, but I did not find anything that I was doing wrong. I think I will have another look at it, though.

Any other suggestion is appreciated, still.

[gr2]

Recently I noticed it works at the desired speed (that is around 400 measures per second) while measuring current to the buffer. Is there any reason why the device would take so much longer (around 16 measures per second) to make voltage readings?

[Albert.Geven]

The timing depends on measuring accuracy and filtering enabled.

So select the fastest mode possible.

[gr2]

As far as I understand, I have disabled all filters and the setting NPLC 0.01 should make for the fastest readings possible. In practice, however, it takes a long time.

[gr2 · Accepted solution]

I have figured it out: I needed to add the command ":SYST:LSYNC:STAT 0" after ":SYST:ZCH OFF".I have also added ":TRACE:ELEM NONE".

I had tried this command before yet it did not work. I believe it was because VISA Interactive Control sometimes cuts the commands short when you click on them from the command "history" panel and I only noticed that now. I have to say this is a rather counter-intuitive "feature".