Page 1 of 1

Spektrum logger with BladeRF

Posted: Thu Dec 24, 2015 9:40 am
by pe2bz

at my website ... _power.png I have a spectrum from 400 - 406 MHz showing the last 8 hours of activity in the EU radiosonde frequency band. It's running with an RTL-SDR stick and the rtl_power tool and it gives the 30 second average signal values on each frequency. Those (csv) values are converted by a script to the png file.

I also want to run this for the 1680 MHz sonde band. However the RTL dongles are not very good on those frequencies.

I try to find a tool which can create such images with the use of the BladeRF. I tried the bladerf-cli and can indeed (running a script) create an CSV output file, but running the bladerf-cli within Windows does not give me the possibility to make a fifo as is described on this page ... and-Tricks

Is there anyone who can help me with this ?

Thanks in advance,

Ben - PE2BZ

Re: Spektrum logger with BladeRF

Posted: Fri Dec 25, 2015 12:04 pm
by jynik
If you're not running at a particularly high sample rate, you should be able to output samples to disk. If the CSV output seems to become a bottleneck, you could try binary output and convert it later. Was there a particular reason you wanted the output written to a FIFO, other than just avoiding a slow disk?

If the CLI doesn't provide the functionality you need, it might be best to dive into the libbladeRF API and write a program to do what you need.

By the way, I believe staticfloat is working on exactly what you're looking to do, with the pybladeRF python bindings, in his project bladeRF_power. We had recently chatted some approaches he could use to improve the quality of his receptions (e.g., offset tuning to avoid the DC bin), so note that this is a work in progress. Just wanted to mention this, as having a couple brains on a similar problem is generally useful! :)

Re: Spektrum logger with BladeRF

Posted: Sun Dec 27, 2015 3:19 am
by pe2bz
Hi Jon,

thanks for the excellent link ! I have been googling but did not find this script.

For the "not realtime" part of what I want to achieve this indeed fits my wishes. Have made a comment about one line which should change in the script because now the end_freq is ignored and the script always tunes from start_freq up to the highest BladeRF frequency receivable.

It now gives me images like this:
which look much better than the dongle reception.

Now I'm gonna try to use the CSV in a PIPE and have another script update the image every 5 minutes.....

Kind regards,