Can I configure niScope Fetch Binary 16.vi to grab I,Q data from PXI-5620 ?

[cdschmit]

I posted this by accident in the general LabVIEW forum yesterday.  Here it is again, hopefully directed to the right people!

Hi,

I am trying to build an RF front end that grabs (< 1.25 MHz BW so DDC is involved) baseband I,Q data from the PXI-5620 *FAST*.  I thought that the niScope Fetch Binary 16.vi would return unscaled, 16 bit (14 actual), interlaced I Q data, but it is now apparent that I am only getting one channel...I assume the I data.  Can I configure it to send both channels?

Chris

[TheWoost]

Hi Chris,

I noticed that this message was posted before my reply on your general LabVIEW forum post.  Here is my post again:

I believe you just may be missing the property node to configure retrieval of the interleaved data.  See the attached image for the location of the property in the niScope property node.

Did this answer your question?

Regards,
Andrew W
National Instruments

[cdschmit]

Andrew,

That didn't seem to work.  I've attached both a transmitter and a receiver for you to try out.  The transmitter sends BPSK at 12.5 kHz and the receiver samples at 400 kHz complex, so I expect 32 samples per symbol.  After the de-interleave, I am only seeing 16 samples/symbol and the IQ constellation plot (locked at + or - 45 degrees) confirms that I am merely dividing the samples of a single channel.  You can change the "Fetch Interleaved Data" control on-the-fly if you hit the Update Scope Settings button.

Chris

[TheWoost]

Chris,

By looking at your code, it does not look like you are enabling the DDC or configuring any settings.  I used the shipping example that we have for Modulation Toolkit (MT ni5660 IQ.vi), and there are many settings for the DDC case (less than 1.25MHz) and even more VIs for the non-DDC case.  Do you know if you have the Modulation Toolkit available?  It will probably be much more complicated to do this without Modulation Toolkit, but I'm checking into this to see if it can be done.

Regards,
Andrew

[cdschmit]

Andrew,

Yes, the MT ni5660.vi works fine, but that wasn't my issue.  I want to use non-5660 VIs to investigate the difference in maximum acquisition rates.  The ni5660 Read IQ.vi in that example returns, if my memory serves me, 32 bits per channel of (scale + offset)-corrected data.  Assuming I don't care about the absolute power or dc offset of the signal, it should be must faster if I can acquire 16 bits per channel of unscaled data.

Chris

[TheWoost]

Hi Chris,

I looked into this a little further...

In your application you are using separate NI-Tuner and NI-Scope calls.  Unfortunately you must use the ni5660 VIs to make use of the DDC, and programming the DDC directly via NI-SCOPE is not officially supported.  Let me know if you have any other questions.

Regards,
Andrew

[cdschmit]

Andrew,

Are you sure of that?  How is it that the signal I obtain with BW<1.25 MHz doesn't show any sign of the 15 MHz IF?  The IF is obvious for higher BW settings.  If the DDC isn't at work, I don't know what is. 

Chris

[Andy_Hinde]

Hi Chris,
The PXI-5620 does not automatically enable its DDC so unless you explicitly enable it you will be acquiring real data. I'm attaching the low-level subVI used by the ni5660 Configure for IQ and ni5660 Configure for Spectrum routines to configure the DDC. It takes in the Digitizer instrument handle and will program the DDC with one of 8 discrete filter designs based on the bandwidth setting. You should be able to integrate this subVI into your Digitizer conf routines to turn the DDC on. The data should be interleaved I and Q samples (16 bit each) by default. If you want to skip the conversion from binary codes to scaled voltage values, you can use the NI-Scope Fetch routine (1D I16 version of the polymorphic routine).Just replace the Read with a Initiate/Fetch pair of routines.

Regards,
Andy Hinde
National Instruments

Message Edited by Andy Hinde on 03-08-2007 01:28 PM

[cdschmit]

Hi Andy,

I have integrated the mod_Config DDC subVI.  Now, I can see the I and Q channels interleaved.  There is still a problem...after configuration, I request the niScope property "Actual Sample Rate" and, with the bandwidth set to 1.24 MHz it returns 1.255E+6 but the niScope Fetch Binary 16 returns an xIncrement value of 2.5500E-5 seconds (3.9216e+004 Hz).  Both numbers seem odd considering the DDC, but the latter seems to agree with my observations...about 3 Samples per Symbol with a BPSK data rate of 12.5k Symbols/sec.  Am I incorrect in thinking that these two numbers reflect the true sampling rate?  What am I still missing?

Chris

[Yardov]

Hi Chris,

There are lots of things going on after this step and probably this number is changing when the scope gets configure with the "record length" and "sampling rate".
I believe the best way to ensure that the output data is correct is by using the ni5660 driver. At the end, this is why there IS a driver to control it. This will not only give you time data accurate but it will correct it with temperature, attenuation, etc.
I have attached the hierarchy of the VI you are using that is actually called when in the ni5660 Initialize.vi.
Hope this helps,