Closed LarsThunberg closed 11 months ago
No, the Airspy models only support 3e6 and 6e6 when using the mini and 2.5e6 and 10e6 with the R2.
3e6 with an Airspy mini shouldn't be a problem, can you test the following: airspy_rx -r /dev/null -a 3000000 -t 2
I wonder if it will run at 3.000 MSPS.
As a work around if you don't get it to run on 3.000 MSPS is to add the packing option. Then command will be airspy_rx -r /dev/null -a 3000000 -p 1 -t 2
If that is working then add the packing option bitpack=1
here an example:
SATNOGS_SOAPY_RX_DEVICE="driver=airspy,<your options>,bitpack=1"
Thank Jan. I have no problem running in 3.000 MSPS with the command above. My guess is that it is a SatNOGS issue, as I see on the SatNOGS forum that more people having the same problem. It also started after an update I made from the SatNOGS setup on the 20th June.
The ooooooooooooooooo
is an indication of sample problems.
Do you experience this with all observations?
No, it is mostly BPSK 1200.
The flow-graph that has the biggest impact on the CPU, I don't use a Pi so I can't test it. Open a post on the SatNOGS community forum, maybe others can help.
Question: Is it possible to implement support for lower sampling rates? When running the driver in SatNOGS it seams that the Raspberry 3 has problem handling samplings rate at 3e6, I get a lot of lost samples.