PXI 5660 gives spurious signals, niScope Driver Status err

[cdschmit]

Hi,

We recently "re-imaged" our PC and the drivers for the PXI were reinstalled.  When I run the RFSA demo panel with nothing attached to the RF input, I get spurious signals below 20 MHz that trigger the signal overload.  An error also flashes repeatedly to the screen, and although I was not able to capture this in the screen shot below, it reads something like Error Occured at niScope Acquisition Status.vi <err> Driver Status: (Hex 0x3FFA4008), and I am not even sure this code is correct as I can only see the upper half of it.

Just to see if I could remove the overload, I changed increased the reference level.  At 30 dBm (50 dB attenuation), it still appeared.  At 31 dBm (still 50 dB att) it disappeared!

Meas and Automation Explorer show no problems with the instruments.  Am I dealing with faulty hardware or just a bad driver installation?

Chris

[Yardov]

Hi Chris,

This is indeed surprising that you change something in software and the screenshots change that drastically.
I would try it on a different slot to make sure this is not something specific to that slot. Then, I would try to isolate which module is giving this output, the digitizer 5620 or the downconverter 5600. For this, follow the following steps:

1. Connect a signal (50 ohms termination) like a sine tone (fc < 25 MHz) and use the Spectral Measurements Toolkit shipped example called “SMT Power in Band for niScope.vi” (located under the Windows start folder » all programs » National Instruments » Spectral Measurements)
2. Make sure your noise floor is pretty low (about -130 dB + 10log(RBW)). You can see this number on the 5620 specifications.
3. If your spectrum seems correct, then the problem should be on the 5600.
4. Connect now the signal to the 5600 and make sure you have the correct connections between 5600 and 5620 (Getting Started guide).
5. Use the test panel in MAX to set the 5600 to the carrier frequency and use the same niScope example to look at the signal.
6. You might see the noise floor now being about -130 dB + Attenuation + 10log(RBW).
7. Make sure you see this across all frequencies.

If in any of these steps you see something as you saw before, contact National Instruments to send the unit back.
Hope this helps,

[cdschmit]

Geraldo,

Things move slow when you are a one-horse town.

The only odd thing I saw when running your 7 steps below was a decrease in the power from 0 dBm to -15 dBm (the noise floor raised by 10 dB).  The spurious signals did not reappear.  This seems to suggest that the 5620 digitizer is innocent and something is wrong with the downconverter...but I can't figure out why I can't re-produce the spurious signals when controlling the 5600 and the 5620 separately.

Chris

[Yardov]

Hi Chris,

Having 10 dB more noise is expected since your attenuation is 10 dB higher, so that part is normal. If your digitizer is innocent, then we have two options:
A: Use another spectrum analyzer to look at the IF output of the 5600.
B: Use the digitizer to validate the fact and call NI for repair the 5600.

Hope this helps you get going,

[cdschmit]

Geraldo,

I used a second PXI chassis (identical configuration) to bypass the 5600 in question.  All tests were done using the SAME 5421 and 5620.  On each PC, I used the RFSA Demo Panel to confure the 5600s on both PXI-1042 in an identical manner.  The results in the figures compare the suspect (bad) 5600 and the "good" 5600 for two different frequency spans.  The spurious signals appear only in the 5-20MHz span and NOT in the 5-25MHz span.   I fed the output of the 5600 into a benchtop HP89410A Vector Signal Analyzer.  Yes, the frequency spurs are real and disappear when I change the RFSA Demo Panel span.

If you don't suspect a faulty PXI chassis, I'll physically swap the 5600s between the two PXI and see if the problem follows the module.

Chris

[cdschmit]

Gerardo,

I also promise to spell your name correct one of these times! 

Chris

[Yardov]

Hi Chris,

Thank you for your consideration.

So, these spurs are real (IF output of 5600) when no RF input and you change only the span on the demo panel. The software call might be tuning into a bad band. This is really odd behavior and you should return the board (ni.com/support will place you on front of the queue).
One more thing, what is the reference level setting in those screenshots? You have a warning message but I cannot read the Hex code.

Thank you for all this information,

[cdschmit]

Geraldo,

>>So, these spurs are real (IF output of 5600) when no RF input and you change only the span on the demo panel.

Yes, real and exist even with no input.

>>The software call might be tuning into a bad band.

I wish I knew what you meant by this.

>>This is really odd behavior and you should return the board (ni.com/support will place you on front of the queue).

I'm running a conversation in parallel with another AE to get an RMA.

>>One more thing, what is the reference level setting in those screenshots? You have a warning message but I cannot read the Hex code.

The HEX code refers to the ADC overload caused by the spurs.  Here are three more screen shots of the RFSA Demo Panel with NO INPUT, three different reference level settings.

Chris 

[Yardov]

Hi Chris,

You did it again (Geraldo spelling) don't worry about it.

>> I wish I knew what you meant by this.

Basically, the 5600 driver tunes in two steps sizes: 1 MHz and 5 MHz and my thought is that the driver is re-tuning (for different spans). Regardless of this, keep the conversation with the AE and hope you are up and running soon.

[Yardov]

Hi Chris,

One more thing, are you using the latest version of the driver? You can download it here.

Just to make sure we are on the same page.