how to send a GPIB command at 1500Hz?

[lolopuf]

Hello,

I would like to set a step +-3dB in a UMTS signal.
I just have to send :POW:STEP 3/-3 via the GPIB port but every 0.666ms so at 1500Hz.

I think the GPIB communication is enough fast but how to implement a loop or whatever which sends 1500 times this command in 1sec and in a regular way (every 0.666ms)

Thank you for your answers

[Jochen]

Hello,

first of all this question raises some major concerns in me:

1. It's correct that the GPIB should be able to easily carry the data rate that you want to send. On the other hand in your case this won't be a continuous data stream but 1500 single write commands per second and I don't think that neither the driver nor the GPIB-board will be able to initiate so many write commands so quickly.
   
   
2. Let's assume that the driver and the board are fast enough, it's very questionable that your GPIB device is capable of processing commands at this rate. Please note that there is also some handshaking going on on the GPIB when initiating a transfer.
   
   
3. On a Windows PC there is no way to run a loop with a stable timing of 0.666 ms. If you are using LabVIEW, you could think of using a timed loop with an external hardware timing, e. g. derived from a DAQ device but still you would need an additional DAQ board to time the loop but still on Windows the typical jitter is significantly higher than 0.666 ms causing a very bumpy timing for your signal.
   To improve this situation you could run your PC or a PXI system under LabVIEW RT, but as GPIB itself is not deterministic (timing is impacted by the handshaking of the device), this wouldn't solve these issues truely.
   

With all this said, I don't think that the approach you are aiming at will work. You better should try to find out if your device provides means to change the 3 dB steps by either an internal timing or by an external trigger signal. A stable 1500 Hz trigger signal could easily be generated with the counter of an NI data acquisition board.

Best regards,

Jochen Klier
National Instruments

[lolopuf]

Hello Jochen,

thank you for your developped answer.

I ve estimated the highest frequency managable by the hardwares, by making a while loop, send 1500 times the commands +-3dB, and in fact it needed 2.5sec, so that s seem unfeasable.

I had two choices, either use labview as I described before, or use an external square signal set to 1500Hz. I will use an external signal.

I ve uploadead my .VI just in case you find out this is not the optimal way to do it.

Best regards,
Laurent

[Ray.R]

I agree with Jochen,

Can you describe what you are trying to achieve (the global picture), because if you can simply use a square wave at 1500 Hz, then that would be the way to go.  I suspect that you do not need to send a command at such a rate.  

By knowing more about your application, we can provide better suggestions.

RayR

[lolopuf]

Ok I can try to describe it.

I generate UMTS signals (1.9GHz) thanks to a  signal generator, and I also need to add to this signal  +-3db step at a rate of 1500kHz.
And in fact, all the material used for the experiment will be taken 2 years by a university to test the physiological effects of UMTS signals to the brain.
That's why the less hardware I use, the better.

But never mind now, I found a new pretty old device to generate my square wave easily. The problem is solved.

Thank you guys.

Laurent

[Ray.R]

Hi Laurent,

Nice research project.  It would be nice to see the results.  We were to carry out similar experiments many year ago for a TET (basically a through-skin-transformer), at ~750MHz (I think).  I was never carried out 😞  

Glad to read that you found a solution.  Let us know if there's a web-link to find out progress / results of this experiment!  🙂

[lolopuf]

Hi,

yes no poblem.
Here you can already find the subject of this project, and other projects on this webiste lead by the National Research Program (in Switzerland)

http://www.nfp57.ch/e_forschung_epidemiologie.cfm?command=details&id=33

Nice reading

Laurent

[Ray.R]

Thank you.

As I said, this is an interesting project.  I'll be visiting the site regularly.. 🙂