RF Measurement Devices

cancel
Showing results for 
Search instead for 
Did you mean: 

Playback mode (NI RF Phase-Coherent Multi-Channel version 1.1)

Solved!
Go to solution

Hello!

 

This is about the Playback mode in the NI RF Phase-Coherent Multi-Channel version 1.1 software.

 

I would just like to ask if I’m expecting identical plots when I view real-time versus viewing a recorded file.

This is because when I playback the recorded xml file, I get plots which are different to those I see during real-time viewing.

 

Videos of the actual/real-time plots and the recorded plots could be viewed on the links below.

The phase offset has already been compensated for both cases.

 

Actual/Real-time: http://www.youtube.com/watch?v=W14RZwFR8MA&feature=youtu.be

Recorded: http://www.youtube.com/watch?v=GxI1STAPt2o&feature=youtu.be

 

 Thank you very much! 😄

0 Kudos
Message 1 of 16
(6,184 Views)

After some off-line troubleshooting with Betty Boop, we learned the following:

 

  • The reason for the strange behavior is that the regenerated sigmal was being output at a much stronger level than it was recorded at.
    • We recorded a -20 dBm signal using a -20 dBm reference level on the analyzer
    • We then played back the signal, leaving the reference level @ -20 dBm.
    • The regenerated signal was actually being generated with a -4 dBm power level
    • Because the ref level was too low, the analyzer was being overloaded causing clipping of the ADC and distortion in the downconverter, leading to strange results.

The next question is why the signal was being output at too high of a level. The software attempts to regenerate the signal at the exact same power level, minus some cable insertion loss.

  • The analyzer gain setting, which is proportionate to the reference level/RF attenuation of the analzyer, seemed too large compared to a similar system/settings here at NIC.
  • More suspiciously, when the reference level was then changed to 0 dBm, the reported analyzer gain value was unchanged.

The gain setting acquired and stored into the Record/Playback XML file is obtained from the NI-RFSA driver function niRFSA Get Scaling Coefficients. As a test, I compared this to the gain value contained in the wfm info output cluster of the NI-RFSA function niRFSA Fetch IQ (I16). They were different, and the gain value from niRFSA Fetch IQ (I16) was correct, and did change with a change in the analyzer reference level.

 

So for now, we are looking more closely at the niRFSA Get Scaling Coefficients subVI.

 

Regards,

Andy Hinde

RF Systems Engineering

National Instruments

 

Message 2 of 16
(6,158 Views)
Solution
Accepted by topic author Betty Boop

Betty what version of NI-RFSA do you have installed? This may be an issue (incorrect reported gain values by niRFSA Get Scaling Coefficients) that is fixed in the latest version of NI-RFSA, version 2.5.

 

http://joule.ni.com/nidu/cds/view/p/id/2666/lang/en

 

Regards,

Andy Hinde

RF Systems Engineer

National Instruments

Message 3 of 16
(6,154 Views)

Hello, Sir!

 

I'm using NI-RFSA 2.4.5. Do I need to uninstall this before installing the latest version?

 

Thank you! 🙂

0 Kudos
Message 4 of 16
(6,145 Views)

Hi Betty,

If you update to NI-RFSA 2.5, your issue will be resolved.

 

You do not need to uninstall NI-RFSA 2.4.5 before installing NI-RFSA 2.5.

 

Regards,

Andy Hinde

RF Systems Engineer

National Instruments

0 Kudos
Message 5 of 16
(6,140 Views)
Hello, Sir Andy! 🙂
First, something weird has happened. Before installing NI-RFSA 2.5, I tried recording and playing back CW. It worked! I used the original acquisition VI, not the revised version. I really don’t understand why it is suddenly working. However, I prefer installing the latest NI-RFSA just to make sure that the problem is solved.
Second, I think I’m having problems with the installation. Attached are screen shots of the errors I’m getting. It is only the first time that I encounter errors during installation. I have downloaded the file from the link you gave me so I don't know where the errors are coming from. 
Thank you very much! 🙂
Download All
0 Kudos
Message 6 of 16
(6,127 Views)

Hi Betty,

I would recommend going back to the NI-RFSA 2.5 download page:

 

http://joule.ni.com/nidu/cds/view/p/id/2666/lang/en

 

and selecting option 2:

 

Standard Download: NIRFSA250.zip (2118 MB).The Standard Download:
  • downloads directly to your PC
  • can be a less stable experience for downloading files should the download be unintentionally interrupted due to dropped connectivity
  • does not provide ability to "pause and resume"

 

When you unzip this file, run setup.exe and see if this works.

 

Regards,

Andy Hinde

RF Systems Engineer

National Instruments

 

 

0 Kudos
Message 7 of 16
(6,111 Views)

Here is a recent question from Betty Boop:

 

Hello, Sir Andy!

 

This is what I'm trying to do.

 

I generate CW signals and then record them for offline processing.
After which, I use the "Read PC Binary Data File and Write to Excel" VI.
From the Excel file, I need to get the I and Q data of the generated signals.

That is, I need to produce a total of 4 arrays:
For signal 1: one array for I data, one array for Q data
For signal 2: one array for I data, one array for Q data

These arrays will be used for baseband processing and would serve as input arrays to the code I made for my intended application.

 

The referenced example code "Read PC Binary Data File and Write to Excel" is located here:

https://decibel.ni.com/content/docs/DOC-21389

 

This code serves as a demonstration on how to read recorded binary data from a file and use the metadata stored in the Record XML file to scale the binary values to scaled voltage values. It then demonstrates how it is simple to then use basic LV File I/O VIs to write these values to an Excel file.

 

The data read from a single channel's data file is written to Excel in a column format, with the first column being the I data and the second column the Q data. The number of rows is the number of samples read from the binary file.

 

If you want to do this for two signals at the same time, you could modify this example code to do this, or you can simply run this example once for each data file, and then copy and paste data from each Excel file into a single Excel file.

 

Regards,

Andy Hinde

RF Systems Engineer

National Instruments

 

 

0 Kudos
Message 8 of 16
(6,101 Views)

Hello, Sir Andy!

 

In getting the I and Q data of a generated and recorded signal, how do I know which samples to extract so that I can completely represent the signal?

 

For example, in the case of the periodic CW, I’m trying to find a pattern on the numbers listed in the Excel file. I’m assuming that repeated numbers on the file correspond to the repeated periods of the CW signal. And then I just need to consider one period to capture/represent the CW. Is this correct?

 

Attached is the Excel file.

 

Thank you! 😄

0 Kudos
Message 9 of 16
(6,098 Views)

Hi Betty,

In the case of a sine wave being recorded to disk, a sine wave is represented by DC or static levels for I and Q vs time. The values themselves are in units of Volts.When you say:

 

in the case of the periodic CW, I’m trying to find a pattern on the numbers listed in the Excel file. I’m assuming that repeated numbers on the file correspond to the repeated periods of the CW signal. And then I just need to consider one period to capture/represent the CW.

 

I think you are misunderstanding what you should be expecting here. You won't be getting a sinusoidal pattern representing the values of the instantaneous voltage of a sine wave, which when graphed would look like a sine wave. Because the signal is downconverted to baseband IQ data and represented with complex values, you will get an array of I data and an array of Q data, where arctan(Q/I) is the phase of the sine wave and sqrt(I^2 + Q^2) represents the RMS amplitude (in V) of the sine wave. As stated above, you will see two DC values essentially, as shown in the attached JPEG which shows graphs of your 1000 samples of I and Q vs sample number.

 

Regards,

Andy Hinde

RF Systems Engineer

National Instruments

 

0 Kudos
Message 10 of 16
(6,094 Views)