E4440A driver compatible with 8.0

[MFPetty]

Is NI's driver for Agilent's PSA E4440A Spectrum Analyzer compatible with LabView 8.0?

[Dennis_Knutson]

That's what it says. Are you having a problem with it?

Message Edited by Dennis Knutson on 10-16-2007 02:24 PM

[MFPetty]

Dennis,

            I'm having a lot of trouble with it, or at least the sample vi, "...Configure and Query Marker.vi", that was included with the download.

When I first executed the vi, it failed, returning error code 1074000000, instrument reports data out of range. I "fixed" that by connecting a value, I choose the start frequency value, to the X(0) input of the Configure Marker sub vi. Now the Configure and Query Marker.vi executes, but it returns incorrect values for both the X and Y values of the marker. With an approximately -20 dBm tone inputted into the spectrum analyzer, the Marker Position output in the vi reads the start frequency, and the Y position reads 0. The input frequency is place in the middle between the start and stop frequencies.

If I individually execute the sub vi's the Marker Search.vi correctly positions the marker on the peak, and the Query Marker.vi correctly outputs the marker's frequency and amplitude. But, if I run the entire vi again, the outputs are incorrect.

Since LabVIEW has been revised at leaast twice since the version 8.0 was installed on the laptop I'm using, and the drive was copyrighted in 2007,  I thought I'd ask the simple question first.

Regards,

Mark

[Dennis_Knutson]

I doubt that the version of LabVIEW would have anything to do with a problem. What I am always a little skeptical about though, is seeing a large number of model numbers supported by a single driver and only a single model tested. Often, the author of the driver will create the compatiblity list from information from the instrument vendor or from the manuals. There could be subtle differences. You mention that individually perfoming the subVIs you get the correct results. Try rearranging the subVIs in the example in the same order that you run them individually and make sure you pass the same values in the example.

[Matthew_Kelton]

If your subVI's are working, you may also look at the timing between them in the example.  Perhaps the VI is running faster than when the user wrote it.  I recently had a problem when a customer had to update NI-VISA to support a new serial card, that my GPIB Keithley driver was failing to accurately report values.  The update to NI-VISA either ran much faster, or some default value changed so the delay that used to exist in the driver went down enough that my delays were no longer long enough.

[MFPetty]

Dennis and Matthew,

             I've reordered the example a few times. I've inserted waiting periods between several subVI's. I've increased the delays inside some subVI's  I've modified some of the subVI's to make sure that the device ID was read correctly in the subVI, and brought that value to the top level. I've removed a subVI, "Set Instrument From Marker.vi", for which I have no need. I've added a second "Marker Search.vi". This last modification got me closer, with one in three executions resulting in a correct output.

I've tried lots of things and spent lots of time, much to the amusement of a colleague who in 5 minutes with 15 lines of Perl was able to configure the spectrum analyzer and correctly read the marker position.

Thanks,

Mark

[Matthew_Kelton]

As Dennis pointed out, the driver claims to support a family of instruments, however, it only has been tested with the E4407B.

I still have a hard time believing that you can run all of the subVIs seperately, but when you run them in the main VI you cannot get them to work.  Are you sure every input you change in the subVIs has a control in the main VI wired to the subVI?

If you turn on highlight execution, or add breakpoints between each subVI, you can identify which subVI has not set the insturment up as you expect.

This example has numerous subVI inputs which are not wired to controls, so their default values will be used.

If you had taken your friend's approach, you may too be done with the task.  It's easy to look the commands up and write a VI that sends all those commands to get your marker values.