[Hackrf-dev] TX Trouble

[Dominic Spill]

On 17 October 2017 at 08:36, <wiemeyer at nehp.net> wrote:
>
> I'm still having trouble transmitting from file, a car door clicker
> waveform that I saved, similar to one of the tutorial videos.
>
> For one thing, the TX light comes on when I start the flowgraph, but when
> I close it, the light stays on.  Is that normal?

It's normal, but not intentional.  There's a bug somewhere in the process
of tearing down the flowgraph and stopping the HackRF.  It's on my todo
list to hunt this down and fix it.

> And, generally speaking, what are things to look for when debugging TX
> problems?  I can think of two problems:  I'm not transmitting with enough
> power, or I messed up the waveform somehow when I recorded it.

> Regarding the first, I've tried setting the RF gain as high as it will go,
> and I've tried different antenna lengths.

Those are likely to be your biggest issues, along with incorrect antenna
length for the frequency of the signal.  What's the frequency of the
signal? 315 MHz? 433 MHz?

In terms of your signal, make sure that it's being transmitted on the
frequency that you think it is, and try to reduce noise if you can (i.e.
filter it in GNU Radio).  If you ahve another SDR, such as a RTL-SDR
dongle, you could take a look at the transmitted signal by receiving with
that.

> Regarding the second, I'm worried that I may have saturated HackRF when I
> recorded the waveform.  Is there a way to tell, if I view the recorded
> waveform in the QT GUI time domain plot?

I like to look at it using gr-fosphor or the FFT sink.

> Anything above 1 V or below -1 V?  I guess the waveform would look
clipped,
> but it appears so quickly, that it is hard to tell.

If you set trigger levels you can have the display hold when the part that
you want is in view.  You'll want to filter to avoid noise interfering with
this.

> Same question on transmit: how can I tell if I am saturating?

Again, having another SDR to view the signal is handy, but there are other
options.  You could clean up the file with some filtering and view it using
gr-fosphor or inspectrum.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist9.pair.net/pipermail/hackrf-dev/attachments/20171018/03f6634f/attachment.html>