Re: [Discuss-gnuradio] Difference in Spectrum Analysis between Wx and Qt?

[Martijn Scale]

Hi Marcus,

Thanks for your reply, I connected a File Sink and later sourced the file into 
both Qt and Wx instruments, with the same result.
I expected this to happen, as for the hardware setup I did not change a thing 
between Qt and Wx measurements, it was just a matter of switching between Qt 
and Wx in GRC and disabling the non-needed instruments.

So I just added SDRSharp into the equation, which showed an AM signal on top of 
the carrier as well, so that indicates that the Qt instruments are right.
Next, fiddled with the Amplitude of the Signal generator, when I reduced the 
amplitude with further 30 dB, the AM signal disappeared.

Although the carrier-level did not reduce a whole lot (maybe some 8 dB), I 
guess the setup was creating a lot of intermodulation distortion. That's fixed 
now.

Still it is interesting that Wx did not pickup that AM modulation... If you 
would be interested I can share the IQ-file....

No more reason to stay with Wx though, I'll keep in track with the community 😊

Thanks, case closed, Martijn

-----Original Message-----
From: Müller, Marcus (CEL) <address@hidden> 
Sent: maandag 24 juni 2019 15:05
To: address@hidden; address@hidden
Subject: Re: [Discuss-gnuradio] Difference in Spectrum Analysis between Wx and 
Qt?

Hi Martijn,

Can you please use a file sink to record the signal, so that you can
verify that it's different when visualized with Qt and Wx?

Because: there might be a scaling difference between Qt and Wx GUI, but
the math should really be the same inside. I'd hence assume that
there's really a different signal causing the different pictures.

Just to pull one tooth right away: staying with WX will force you to
stay on GNU Radio release 3.7 forever, and within that, using a
component that nobody in the GNU Radio developer community knows how to
maintain. So, I'd still really recommend moving away from WX.

Best regards,
Marcus

On Mon, 2019-06-24 at 12:04 +0000, Martijn Scale wrote:
> I’m using GNU Radio v 3.7.11 on Windows 10, 1903 with Intel i5-3550 CPU and 
> on Linux Ubuntu 18.04, with i7-3720 CPU. Both have 16GB mem.
> Further, I have connected a HackRF One with firmware 2017.02.1. To the HackRF 
> One I have connected a Siglent SDG1025 Waveform generator, with external GPS 
> disciplined oscillator. I used a DC blocker and 30 db attenuation between de 
> generator and the HackRF One.
>  
> The waveform generator generates a sinewave of 9.76 MHz, I have setup GNU 
> Radio with a osmocom Source with 2M sps and Frequency of 10.3 MHz.
> To the osmocom Source I have connected first a QT GUI Frequency Sink.
>  
> To my surprise I see an AM modulated signal of around 70 kHz on the 9.76 Mhz 
> carrier. I have tried many things, removed the external oscillator and used 
> the internal one, tried a different waveform generator (HP33120A), but same 
> result, on both machines.
>  
> Today by coincidence I found that there is NO AM modulated signal when using 
> WX GUI and the WX GUI FFT Sink.
>  
> Anyone any thoughts on this? Qt will prevail over Wx in the future, but for 
> me this is unexpected behavior and I would like to stick to Wx.
>  
> 
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/discuss-gnuradio