spurious 96kHz on output

General discussion and support for the Peaberry SDR V2.

spurious 96kHz on output

Postby ks1h » Sat Mar 14, 2015 8:03 pm

Been playing around with Peaberry CW and trying to track down a buzzing sound on the carrier. Then I noticed (with HDSDR) that when the tune freq is offset quite a bit (40kHz, for instance) from the LO, there is a 96.000kHz modulation of the carrier. The modulation amplitude (but not the frequency of modulation) increases with distance of the tune freq from the LO. My scope indicates a modulation amplitude that can be higher than 50%. If I pick a tune freq close to the LO, I can't see the spurious component at all on my scope but I can still hear it in a separate receiver so it is still there. I get this with Peaberry CW and with HDSDR in AM or FM mode. I'm not sure where there is any 96kHz signal in the Peaberry that would cause this. Has anyone else experienced this or have any idea what could be causing it?

Ken KS1H
ks1h
 
Posts: 47
Joined: Sun Dec 22, 2013 6:50 pm

Re: spurious 96kHz on output

Postby W4MMP » Sat Mar 14, 2015 8:15 pm

Hi,
Any particular band or all bands? I will check but please let me know what version of Peaberry you have (mine are 80/75/60/40 and 30/20/17/15)
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: spurious 96kHz on output

Postby ks1h » Sat Mar 14, 2015 11:40 pm

I made the 60/40 and 30/20 version but the effect I was was on 40 mtrs. Haven't tried another band yet.
ks1h
 
Posts: 47
Joined: Sun Dec 22, 2013 6:50 pm

Re: spurious 96kHz on output

Postby ks1h » Sun Mar 15, 2015 3:30 pm

More poking around gives this:

It shows up on all bands and, in fact, is on the output of the pcm3060 so it is band independent.

The modulation frequency varies with the HDSDR Input Sampling Rate (under the Bandwidth control). The Output Sampling Rate doesn't affect it.

I get the same result on a different computer.

Ken KS1H
ks1h
 
Posts: 47
Joined: Sun Dec 22, 2013 6:50 pm

Re: spurious 96kHz on output

Postby W4MMP » Sun Mar 15, 2015 3:58 pm

Hi Ken,
Just to be sure, you are referring to VOUT(L/R,+/-), correct? I will break out my (less than good) USB scope and take a look.
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: spurious 96kHz on output

Postby ks1h » Sun Mar 15, 2015 4:33 pm

Hi Ron,
Yes, although I found it much easier to probe R25-R28 so as to avoid destroying something.

73, Ken KS1H
ks1h
 
Posts: 47
Joined: Sun Dec 22, 2013 6:50 pm

Re: spurious 96kHz on output

Postby W4MMP » Mon Mar 16, 2015 1:25 am

Hi Ken,
I scoped my 30/20/17/15 on 20M. Yep, starting at approximately at a 40KHz LO differential from Tune frequency the audio waveform becomes distorted. The measurement point was at R25 and I am assuming the remaining Op Amp outputs are reacting similarly. At a 10KHz differential, the waveform is perfectly clean. I don't know for sure why this happens other than I suspect 40KHz is out of spec for the design of the PCM3060 and/or the Op Amp. This is a yet another new one for me, but as I have HDSDR configured for a fixed LO of 10KHz it does not present an issue for me.
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: spurious 96kHz on output

Postby ks1h » Mon Mar 16, 2015 1:55 pm

It would appear that a transmitting freq of ~ 40kHz from the LO is too close to the Nyquist limit (96kHz/2=48kHz) for the pcm3060 DAC to deal with. It also doesn't look easy to increase the sampling rate for transmit since the rate for both receive and transmit comes from a single clock source in the PSoC. This is all somewhat important since it is possible to transmit out of band if one isn't careful. For example, if you chose a LO of 7.050 and a freq of 7.010, then there will be sidebands at 7.010 + 0.096 = 7.106 and 7.010 - 0.096 = 6.914. I have checked these (transmitting into a dummy load!) with a separate receiver. With only a 10kHz separation of LO and transmitting freq then the sidebands are almost undetectable. I think it is prudent to keep the LO and transmitting freq to within this limit (or to keep a 96kHz buffer from the upper and lower edges of the bands).

Ken KS1H
ks1h
 
Posts: 47
Joined: Sun Dec 22, 2013 6:50 pm

Re: spurious 96kHz on output

Postby ks1h » Thu Mar 19, 2015 11:35 pm

Spent some time trying to get the tx sampling rate up from 96kHz but the combinations of clocks that might work violate some constraint within either the pcm3060 or the I2S module. I suppose David already found this out and that is why the rates are what they are. Haven't given up yet, though.

Ken, KS1H
ks1h
 
Posts: 47
Joined: Sun Dec 22, 2013 6:50 pm

Re: spurious 96kHz on output

Postby ks1h » Fri Mar 20, 2015 3:23 pm

This is a calculation a 45kHz signal sampled at 96kHz.
45kHz_signal_96kHz_sample.jpeg
45kHz sinusoid sampled at 96kHz
45kHz_signal_96kHz_sample.jpeg (47.36 KiB) Viewed 4918 times
It is very similar to the scope trace of the output of the pcm3060 for a 45kHz signal so it looks like hardware is working as it should. Clearly highly distorted so it is not surprising that there is spurious output. This effect seems to be inherent in any SDR that uses the same sampling rates for Rx and TX, as HDSDR/Peaberry does. It seems to me that one would want to use an output sampling rate of at least 192kHz if the Rx sampling rate is 96kHz. That would mean at least 4 Tx samples/period up to 48kHz.

73, Ken KS1H
ks1h
 
Posts: 47
Joined: Sun Dec 22, 2013 6:50 pm

Next

Return to Peaberry SDR V2

Who is online

Users browsing this forum: No registered users and 3 guests

cron