I’m always looking for opportunities to reduce the costs associated with high quality decoding tasks. The HackRF One is intriguing because it’s not very expensive, has a wide band receiver (10MHz – 6GHz) and an SDR that is capable of sample rates of 2-20 MS/sec (albeit with only an 8-bit sample size). The only way to find out how good (or bad) this SDR was to order one and test it out. Below is my new HackRF One wide band SDR.
Normally, I would hook up an SDR to the IF output to a communication receiver, but this little SDR has a wide band front end that supports 10MHz to 6GHz. So, I hooked it up directly to the Inmarsat POR L-band splitter that feeds many of the receivers here at USA-Satcom.com.
The next step was to just do a quick test to ensure things works. So, I used SDR-Console V2 to get the HackRF One up and running on windows, tuning to the Inmarsat POR L-Band. It’s quite amazing to see almost the entire satellite transponder space on one screen! Again, using the SDR-Console-V2 I set it up for USB demodulation and directed the audio output to one my really old soundcard based decoders. Below you can see the old-school Inmarsat Standard-C decoder running without any immediate issues.
Soundcard based decoding is error prone for many reasons, so now that we have gotten this far, I went ahead and made a quick modification to my Aero-P multi-channel decoder to see how well things would work. This time I took advantage of the UDP Broadcast feature to steam IQ samples to my decoder. Below you can see 3 channels of Aero-P being decoded, again without any immediate issues.
HackRF One is set to the minimum sample rate of 2MHz for this test. As you know, the HackRF One uses just 8-bit samples and so there is some concerns about having adequate dynamic rage. But so, far with the fairly strong Aero-P signals there are no problems, it was running for about 4 hours without a single frame being lost or even a single CRC error.
While the testing so far has been successful I still want to do some additional modifications to allow full Aero decoding, which requires 4MHz of BW and 9 total channels. Voice channels can be weak and 8-bit sampling could introduce problems.
Two issues exist so far; The frequency stability is not so good (i had to correct for 4.5KHz out of the box), however, once set, it’s OK. This should be easily fixed by getting a more precise clock source – but there appears to be a HackRF One firmware bug preventing switch over to external clock source. The other issue is while 2MHz sample rate looks pretty clean, 4MHz has some spurs and images – this could be problematic for wide band use for my applications.
So, next steps are firmware upgrade to fix clock issues, more accurate clock source and finally some modifications to Aero decoder to allow 4MHz of spectrum as input.
If all this is successful, the HackRF One will replace one communications receiver and two SDR (that are limited to 2MHz sample rate). That would be quite amazing.