Tektronix and Keithley

Hi all,

I am trying to take very small ~5-10pA measurements with my 6485 and record the readings in Excel using LabView. There is also another component where the voltage steps with a Keithley 3390 function generator happening in my program but that part shouldn't be interfering. Right now, I am getting about 4 samples/second with a GPIB connction. My goal is ~200 samples/second.

I have integrated the suggested SENSE:CURR:NPLC 0.01 command to get the 6485 off of the "slow" rate mode but as soon as I reach the "KE6485 Read" function, it goes back to the slow mode. I have tried replacing that function with a "GPIB read command" and using a "decimal sting to number" function to feed into the "Build Array" Function but then I get an error 6 (Generic File I/O Error) on labview and an error 420 (Query Unterminated.) on the 6485. This makes me think it doesn't know what to read on the screen.

Looking back at the - manual (Example E-3 in appendix E) there is a code for a 900 readings/second method but I am unsure how to incorporate it into labview. The code is:

*RST ' Return 6485 to RST defaults.
FORM:ELEM READ ' Return readings only
FORM:BORD SWAP ' Set for swapped byte order
FORM:DATA SRE ' Return single precision floating point binary
TRIG:DEL 0 ' Set trigger delay to zero seconds
TRIG:COUNT 8 ' Set trigger count to 8

SENS:CURR:NPLC .01 ' Set integration rate to .01 PLC
SENS:CURR:RANG .002 ' Use 2mA range
SENS:CURR:RANG:AUTO OFF ' Turn auto-range off

SYST:ZCH OFF ' Turn zero check off
SYST:AZER:STAT OFF ' Turn auto-zero off
DISP:ENAB OFF ' Turn display off
*OPC? ' Operation complete query (synchronize completion of commands)
read back result of *opc?
for i = 1 to 1000
"talk" the Model 6485. ' In 488.1 mode this will automatically do a READ? (see “Trigger-on-talk,” page G-7)
read back the binary data
next i
DISP:ENAB ON ' Turn display back on

When I looked up the trigger on talk section, it seemed like it should be reading whatever is on the screen by default which makes the 420 error more mysterious.

I have attached my VI (Run it with waveform 1 for DC). Does anyone have any suggestions?

Here is a crude Labview 2012 VI that controls the Model 6485 for speed.
This is one way to do this.
If you truly need 200 readings per second rate then I would suggest you use the internal buffer of the 6485.
This attachment does not use the internal buffer. It just uses a loop with a READ? and then read the data, for each point.
The SCPI Command codes that are shown are the important fact.
Use them and experiment to get what you need.
Hope this helps.

The issue is most of my data runs go for around 25 minutes and the internal buffer on the 6485 can only hold 2500 data points. At 200 samples/second, that only leaves me with 12.5 seconds of data. I will work with this VI to see if I can better situate the Read command into my original VI and post it when I get that integrated.

That is the way to go.
Remember the few commands that affect the measurement rate are in the VI I posted.
Dale

Hello Dale.

Do you think I can use this similar labview program for fast measurements using 6485 over RS-232?
I need some 50 readings per second.

That program was using GPIB.
Just change over all the GPIB WRITES and GPIB READS to VISA WRITES and VISA READS.
And then use RS-232, if you want.
Remember with RS-232 you may have to concatenate a LF or CR for each command string with RS-232.

I recently found this thread and found it very helpful. I am trying to use a 6487 to source voltage and measure current as a function of time. I want to pulse the voltage between 0V and some desired voltage for a set amount of time. I have been using the KE6485 labview driver to control the instrument. I have set up the "read" command in a loop that is controlled by a wait timer. The time between readings, however, seem to bottom out at about 300ms. I would like to decrease the delay time to at least 0.1s. I originally thought that I was ultimately limited by the fact that i was using the read command instead of the internal buffer, but then I tried the program that you supplied in the previous thread (6485graphdatafast.VI) and that works with my set up to below 100ms delay. As far as I can tell, that program essentially does the same thing as my program, but uses the GPIB command rather than the keithley driver commands. (I even tried deleting the graph from the loop and it was still reaching the same limitations. Is the sampling rate really limited by the type of command I use or is there something that I am missing here? I attached my program. I am only referring to the i-t measurement folder, and not the I-V scan. Thanks for your help.

Dan

In your VI, where is the Integration rate set?
This would be the NPLC of the A/D converter?
The fastest setting is NLPC = 0.01.
I am thinking the 6487 is taking too long to make the measurement.
This would increase the rate.

the NPLC was set to 1. But even when I set it to 0.01, it is still limiting out at about 300-400ms. I also tried to delete everything from the loop except the read function and an indicator of the measurement, turned auto rage off and set it to 0.002 as you did in the program you provided, and made sure that auto zero was turned off. Even after all of that, it was still limiting out to about 300ms. The only thing as far as i can tell that is different is that I am using the KE 6487 driver program, which ultimately references a .dll. Is it possible that the limiting factor could be referencing the dynamic library where as your program just uses a direct call to the instrument over GPIB?

Dan

With an NPLC of 0.01 and everything else set for speed I would expect somewhere in the range of 200 readings per second or around 5msec.
So I would look at using the SCPI Command codes instead of the driver. Maybe the driver is doing too much "housekeeping".