Can't get readings from 3458 meter to refresh

[Todd_Lesher]

"NI Approved" drivers are sometimes written by interns (such as at Agilent). NI does not test every intended use. (Well, they didn't a few years ago.) And it is unlikely they keep stacks of equipment to test every driver submitted. Why would they? It is typical for manufacturers to submit their own code.

Such drivers are extremely useful for learning specific uses of the instrument. Once you go outside those specific cases, you either have to poke-and-hope (wire-and-fire?), or figure out what the driver is actually doing. I can't remember using any driver without modifying it - usually by writing it from the ground up. Capture all the SCPI commands from an example, read the programming manual to see what's happening, then modify for your use.

One example is that (a few years ago), the 34401 had no way to change the input impedance from 10M to >10G without configuring the entire measurement. That was fine for some setups, but not fine for other setups. The 3458 has its own intricacies that the driver's submitter can not completely cover.

If nothing else, capture all communications with NI SPY (doesn't that intercept VISA calls?) and you may end up being happy with tweaking the installed driver instead of writing your own. Or use all the config VIs and just put a VISA write and read for the READ?.

[dt1804]

Any real solutions to this yet?  It is something that we have also been facing.  In the past we had to use an old hp dll to accomplish this.

[Jim_Glanz]

Accidently hit accept as solution by mistake, anyway to undo that ?

The solution for me was to use some of the instrument driver and also sendig my own command string using the VISA read and write VIs

[AndyTT]

Did this ever get solved? I am having the exact same problems!

Thanks

[cstorey]

I've spent a lot of time with the 3458.  It has a very intricate trigger model.  Reading this thread I think the OP failed to properly set that up, so the meter never actually took a fresh reading, but just returned the reading in memory.

Perhaps if you describe what measurement you are trying to make DCV?  How fast, how often, expected range, etc.. I can suggest an approach.  For a lot of applications witht the 3458 its best to use the internal scripting setup and then just trigger that and read the results.

Craig

[AndyTT]

There are no special requirments really. The main issue is not being able to code in the standard way. Usually we dont like to initialise and close for each measurment if we are going to be making multiple measurments.

For some reason the meter needs to be initialised each time you want to take a fresh reading from memory. The display actually updates correctly but the data recieved in labview is identicall for each trigger.

My main aim is to be as efficient as possible using best practise coding techniques. This meter and its plug and play drivers dont seem to allow that. 

What would you suggest for DCV measurments if speed is the main objective?

What would you suggest for DCV measurments if accuracey is the main objective?

Thanks