How to clear RFSA WARNINGS

[serge_malo]

HI all,

I'm using RFSA to acquire GPS signals (basically, noise signal) on a NI-5663.

My code is all in C++, using MS VS 2010 Express.

Signal acquired: centered at 1.57542Ghz, 50Mhz bandwidth, RefLevel -50dBm.

I start the acquisition correctly (RFSA_Initiate was called successfully), and the I/Q data is fine.

I call niRFSA_FetchIQSingleRecordComplexI16  60x per second.

I know that my input signal might "peak" too high for the current Reference Level (-50 dBm), but it's Ok for me if this doesn't happen too often (once every 5 sec would be acceptable).

I know this "too high peak" occurs when I get the next warning from the RFSA Driver:

"... overflow outside the valid vertical range...".

The problem I have is that, as soon as I have this warning once, I will always get it on ALL my next calls to the RFSA driver until I stop the acquisition. It seems that I can't clear this warning message once it occured.

When I got the warning, I have tried to call 

niRFSA_ClearError(0);
niRFSA_ClearError(m_rfsa_session);

But I keep receiving the warning message.

Is there a way to clear the RFSA warnings without stopping the data aquistion?

Best regards,

Serge

[Anthony_F]

Hello serge_malo,

What is the return value of the niRFSA_ClearError() call?  If the return value is zero then the error was cleared successfully and it is possible that what you are seeing are multiple errors.

Do you have any amplifiers inline with your antenna? Can you describe your setup?

Can you post a picture of the Soft Front Panel running/configured with these settings?

[serge_malo]

Hello Anthony,

Thanks for your answer.

I receive the warning from either "niRFSA_GetFetchBacklog" or "niRFSA_FetchIQSingleRecordComplexI16" calls. In both cases, my next calls are "niRFSA_GetError" and "niRFSA_ClearError", and they both return 0.

My setup has an Amplifier between the antenna and the RFDC. It is set at +11 dB.

Here's what I see:

When I increase the Amplifier's gain by 0.5 dB, I get a the next warning after ~5 sec:
RFSA status: 1073364993, function: RFSA_GetFetchBacklog, details: The digitizer returned the following warning:
An Onboard Signal Processing feature has caused the input data to overflow outside the valid vertical range. This can be corrected by reducing the input signal amplitude with digital gain or configuring a larger vertical range.
Status Code: 213401
Status Code: 223532

But as I said, once I get this WARNING, I keep on getting the same warning over and over, even if I diconnect the RF Cable connected to "RF IN" of the RFDC! If I stop and restart the IQ acquisition after I have disconnected the cable, I obviously don't get the warning anymore, as the input power is much lower.

You had a good idea of checking with the RFSA Soft Panel:
When I set all the parameters at the same values as in my program, we see an orange "WARNING" at the top-right of the spectrum window. But this warning dissappears when I disconnect the RF cable (obviously, the input power is much lower).

So, I only want to implement the same thing as in the RFSA Soft Panel: Clearing the warning without stopping the IQ acquisition.
The problem I have is that I seem to not be able to "clear" that warning once I get it.
There is no "niRFSA_ClearWarning" function.

Any ideas on how I can do this?

Regards,
Serge

[serge_malo]

More details:

My program uses the Acquisition type "NIRFSA_VAL_IQ".

I have set the NI RFSA Soft-Panel in IQ-Mode, and checked with NI-Spy what was going on, in order to make the Warning disappear.

I realized that the Acquisition was aborted and re-initiated each time there was that warning.

So, it seems I can't clear that warning without stopping the Acquisition.

See the attached NI-SPY capture

[Anthony_F]

Hello Serg,

I have tried this on a 5663E device with RFSA driver version 2.5  When the input signal goes over the range the Warning is displayed but when the signal amplitude is lowered the error goes away.

My NIspy capture does not have any ClearError or Abort function calls.

Can I ask what version of RFSA driver you have?

[serge_malo]

Hi,

I have RFSA 2.4.

I'll try the same test with RFSA 2.5 and post the results here.

Thanks again!

Regards,

Serge

[serge_malo]

Hi,

Well, too bad, I still have the same problem with RFSA 2.5.

One important thing I have noticed with RFSA Soft-Panel is that we don't see the "Abort/Initiate" calls if we use it in Spectrum Mode. But we do if we use it in "IQ Mode". See attached "CaptureRFSA2.5.tct".

I lowered the power of the input signal so we don't get any Vertical Range Warning, and I still see the "Abort/Initiate" calls. So, I understand that this is simply the way the Soft-Panel is implemented for the "IQ Mode".

My Application uses the "IQ Mode", but we continuously fetch IQ from the Digitizer, without calling "Abort/Initiate", unless the users wishes to change the frequency, the ref level, etc...

As I explained, once I get the "Vertical Range Warning", I always get it when I call RFSA, even when I have lowered the power of the input signal.

So, is this a bug in the RFSA driver, or is there another way around?

Regards,

Serge

[Anthony_F]

Hi Serge,

I looked at the IQ mode for the soft front panel at it looks like it always executes the same 17 steps in a loop for "continuous" mode -- but it includes abort and initiate calls.

A better test in this case will be to use the example RFSA Acquire Continuous IQ where we can control the sequence. I know that you are not using LabVIEW and while I tried to recreate your particular problem I noticed that the RFSA Clear Error function is not available from the LabVIEW Environment.

I will need to speak with the product support engineer for the RFSA driver on this issue. (Clear error not working, Clear error no available in LV and this error being persistent)

[serge_malo]

Hi,

Thanks again for your answer on this issue.

I will try to have a compiled version of the "RFSA Acquire Continuous IQ" example, and I will look with NI-SPY if I can see anything that would give me a hint on how to clear those warnings.

Regards,

Serge

[Anthony_F]

Hi Serge,

Forgive me but I may have not explained myself properly. The Acquire Continuous IQ program also has the same problem you are experiencing, this is yet another reason for talking with the product support engineer since the demo is somewhat incomeplete I would say