Using Spectran VSG 120X to generate custom waveforms
Hello,
We want to use the Spectran VSG 120x to generate a custom radar waveform generated in a MatLab/Python script.
Since multiple people are involved in this project we decided to go for a File Source/File Reader mission architecture instead of using the HTTP connection. We already have a script that generates the IQ raw file and its XML, but since it is not outputting the signals correctly we had to step back and debug the behaviour with Aaronia blocks first.
I am trying to use both input blocks (File Reader and File Source) as depicted below:
If we use the 'IQ Signal Generator' directly connected to the Spectran v6 block we have a perfect sine at its output if the 'USB Compression' is set to Compressed. Also, when writing the 'IQ Signal Generator' output signal via 'File Writer' block set to rtsa extension we see the same signal with few spurious.
However, the problem starts when we use the IQraw format with the 'Raw IQ File Writer'. The output signal behaviour is very erratic as you can see below:
I can't understand why this is happening and I need help to fix such issue. The CPU usage seems low.
Is it a low resource approach to use a HTTP connection to send data instead of File Readers?
Thank you
Did you check out https://v6-forum.aaronia.de/forum/topic/simple-iq-file-to-tx/ ?
Also what are you actually looking at there? Is that the output of the Raw IQ File Writer recording the IQ Signal Generator output, or something else? It isn't quite clear from your description.
For troubleshooting it is usually best if you can provide the mission file that generated the problematic output. And use the proper visualization blocks for analysis, not just the preview of the File Source.
Thank you for the help.
@AdminTC: About the 'Simple-iq-file-to-tx' post, yes I did check it before. I am using all the provided cables and power supply. Please note I am not using the boost cable, because I have the 120MHz model. At least, I didn't have any boost cable included in the box.
@mm_dev: I started with a very basic example provided here in the forum 'Simpel-IQ-FileReader' to read both .rsta and .iq files to their respective blocks. I send in attachment this mission with the small changes I made with both read blocks. The generation of the files was performed with the mission I send in attachment named x.rmix. There I have both the 'Raw File IQ Writer' and the 'File Writer'.
I can also send other visualization blocks result:
1) File Reader 2) File Source
They both look very similar to me. I hope I further clarified your questions. Thank you for the help once again.
Uploaded files:Well, I see a few things in those missions that could cause issues/confusion:
- In your filereader mission, the "Emulate Realtime" option is enabled on the File Source block, but disabled on the File Reader block. This can make a critical difference when streaming IQ data to the V6 generator.
- In your first post, you're comparing very different datasets (look at the time axes)
- The previews of File Reader and File Source will be different even for the same file and are therefore not comparable (as said, use the regular Waterfall block for that)
- The "Max Folder Size" option of the File Writer block doesn't work the same way as the "Max File Size" option of the Raw IQ File Writer block: It won't stop the recording when the defined value is reached, but will execute the "Max Folder Action" action to free space (it is for a very different use case). This was probably the reason why you looked at different datasets.
Some recommendations:
- If you want to ensure that you're recording the same data, better remove the size limits and enable the "Auto Start" option instead. Then both blocks will start/stop recording exactly when you start/stop the IQ Signal Generator.
- Ensure that "Emulate Realtime" is enabled when streaming to the V6 generator.
- Don't record uncompressed IQ data at 92 MHz sample rate if you don't actually need it, the bandwidth can exceed what some storage systems can handle (it is fine for small samples of up to 2-3 seconds).
- When comparing records, use the same block type to process them instead of mixing File Source and File Reader. In the same way, if you want to compare two different block types, use the same record (so if you think the File Source block is causing you problems with the .iq record, check if it does the same with the .rtsa record).
- For comparing data, only use the proper visualization blocks (e.g. Waterfall or Spectrum) and check the settings.
