HackRF test. Broadcasting FM silence

Discussion in 'Software Defined Radio (SDR)' started by VU2TVE, Dec 5, 2019.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: abrind-2
ad: l-BCInc
ad: Left-2
ad: Left-3
ad: QSOToday-1
ad: l-gcopper
  1. VU2TVE

    VU2TVE Ham Member QRZ Page

    Consider this very simple flowgraph below. It should broadcast FM silence at 104.5 MHz. There is no station in my vicinity at 104.5. When I tune my receiver at this frequency, I do hear silence, but when I turn the volume up all the way, there is some static that I can hear. I can hear it on multiple receivers.

    Now, I understand that in radio, there will always be static, but this seems a bit more than "normal". Is there any way I can dig deeper into this to reduce the static...

    Is it a function of the fact that the hackrf is an 8 bit SDR, and perhaps using an SDR with better resolution might make the resulting signal quieter?

  2. VU2TVE

    VU2TVE Ham Member QRZ Page

    When I compare it to a very cheap chinese FM transmitter broadcasting on the same frequency, the cheap chinese transmitter is quieter.

    What could be reasons for this?
    Last edited: Dec 5, 2019
  3. AG6QR

    AG6QR Premium Subscriber QRZ Page

    My guess is that it is phase noise/jitter in the VFO. The frequency is synthesized, and is probably pretty accurate on average, but with some rapid short-term random variation. Any variation in the frequency/phase of the transmitted signal is seen as modulation by an FM receiver.

    To get it perfect in a digitally synthesized VFO, you need infinite precision, both in timing and amplitude. Of course nothing achieves perfection if you look closely enough. More bits, properly implemented, would help.
    VU2TVE likes this.
  4. VU2TVE

    VU2TVE Ham Member QRZ Page

    The hackrf comes with a build in internal oscillator that generates the clock. I did also purchase an external tcxo clock, but havent yet plugged it in. Do you think it might help?
  5. VU2TVE

    VU2TVE Ham Member QRZ Page

    The reason I'm doing this test is because when I am inputting actual audio, it seems there is a fair bit of background noise in the produced FM.

    Upon further experimentation, it seems to me that the FM signals that are being produced are low in volume. What I mean by that is that when I tune to a commercial FM channel, and then re-tune to my channel, the audio volume on my channel is quite low. Any thoughts on how to boost that within GRC?

    I'm guessing there will always be some level of noise in radio. What is that noise with respect to the signal is what's going to determine is usable or not.
    Last edited: Dec 5, 2019
  6. W6RZ

    W6RZ Premium Subscriber QRZ Page

    Here's some test files for you to try. The first one is the flow graph I'm using for a single channel. It has the correct parameters for the WBFM Transmit block. You'll need to replace the UHD block with your hackRF osmocom block.

    The other files are test audio files. ssbaudio.wav is full fidelity 48 kHz audio. ssbaudioprocessed.wav is a compressed (in audio level, like a commercial FM station does) version of ssbaudio.wav. silence.wav is just silence.




    VU2TVE likes this.
  7. VU2TVE

    VU2TVE Ham Member QRZ Page

    I tried your GRC block with the osmocom sink and the ssbaudio file. My setup is the HackRF one in transmit mode, connected directly via SMA cable to an RTL-SDR. I didn't change any parameters on the GRC block except the osmocom sink.

    This is the resulting output as captured on cubicsdr software recording from the rtl-sdr.


  8. W6RZ

    W6RZ Premium Subscriber QRZ Page

    It sounds good to me. It's a little lower in volume than the original file, but it's difficult to know where that's happening. Could be CubicSDR. If you like, I can give you an exact gain difference number. Here's a sine wave file that can be used to get an accurate number. Just post another recording of it, and I'll analyze it for you.


    You can always add some audio gain. However, you'll run the risk of over deviating on peaks.


  9. N2DTS

    N2DTS Ham Member QRZ Page

    I think the lower the sample bits the higher the noise and distortion.
    The clock phase noise also comes into play.
    All that at lower signal levels.
    My 14 bit sdr-iq has a good amount of hiss at all times, a good 16 bit radio has much less, as do some sound card based (down converting) radios.
    My home brew single conversion tube superhet has a VERY low noise level and works just as good or better then most other radios as it has almost NO
    background noise with no signal.

    Terminate the antenna port and turn the audio gain all the way up.
    If its very quiet, its good. If its a loud roar,, that will cover up weak signals.

    A strong signal will reduce the gain and/or cover up built in noise.
  10. VU2TVE

    VU2TVE Ham Member QRZ Page

    I apologize for the silence on this thread. I'm very much committed to getting to the bottom of this but caught up with other work taking up my bandwidth.

    Shall return to this in the coming week.

Share This Page