Settings for reflevel does not get applied
We are currently using the latest version of Aaronia in a Windows environment for our development efforts. Our process involves initially configuring the settings in the Aaronia UI and then transferring these settings to our application, which utilizes the provided SDK. However, we have encountered an issue with setting the main/reflevel value through the SDK.
The settings are configured in the following order:
- "main/centerfreq", 3.050e+09f
- "calibration/preamp", "Both"
- "main/reflevel", "-25"
- "device/sclksource", "Oscillator"
- "device/receiverchannel", "Rx12"
- "device/outputformat", "IQ"
- "device/receiverclock", "122MHz"
We intentionally set "calibration/preamp" before "main/reflevel" because we observed in the Aaronia UI that it is a requirement. Despite trying various approaches, including different data formats (string, float, int), we have been unable to resolve the issue. Can you please provide guidance on how to address this problem?
EDIT:
When reading back the result of the operation, we get AARTSAAPI_WARNING_VALUE_ADJUSTED
This seems to be a limitation/bug in the SDK in that the value range of the "reflevel" is not adjusted after initialization as would be expected. The development team has been notified about this issue.
Thank you very much for your reply. When could we expect this feature to be available? Without it, most of the use of the SDK is pointless for us.
Just implemented it. Available with build 15799 on.
Thanks a lot. Any estimate of when the build will be available for the general public? 🙂
Hi @AdminTC
Just wanted to hear if you have any update on when you expect this to be available? Very much looking forward to the update.
Please contact is to get a BETA release if needed.
Hi
I have had some time to play around with the provided beta build (3.0.1.15813). It does fix the issues described above, but I have found two new (possibly related issues).
The first is that the reported data flow is doubled when running in Rx12 mode. The data packets, as in the newest public version (3.0.1.15767), each contain 4*65536 floats, representing the interleaved dual-channel IQ samples. With a sampling rate of 122 MHz, this corresponds to roughly 0.54 ms. With the public build, that is also difference between the endTime and startTime field in each packet. However, in the beta build, endTime-startTime is around 0.27 ms, and I get twice as many packets as with the public build. This effectively means that the data flow is doubled, which should not happen, especially since it would exceed the bandwidth of the USB connection.
I have attached two plots, one made with the public build and one with the beta build. I ran the Spectran with a center frequency of 2.44 GHz and a sampling rate of 122 MHz in Rx12 mode. However, I had only plugged an antenna into Rx1, meaning that I would expect only noise on Rx2. The images show power over time for a single data packet. For the beta build, there are several holes in the data. These correspond to the IQ samples being all zeros, making the power undefined.
The second issue, also visible on the plots, is that the data using the beta build seems to not be independent between the channels. As mentioned, I only had an antenna plugged into Rx1, so I expect only noise on Rx2, yet the power readings are clearly strongly correlated between the channels.
I have attached a minimal example demonstrating the double data rate.
Public build
Beta build
Hi
I was wondering if you have any update on this?
Sorry, we didn't really have time to look into this yet as we've found some other issues with the SDK that need to be resolved first.
