SSB decoding in SDRs and as procedure (algorithm) in the digital software


Wolfgang Hartmann
 

Hallo Jim,

as far as I understood the technique of decoding SSB, there is a relationship to AM-decoding.

In AM, the software-author takes two signal parts of the total AM-signal after decoding and adds it to an AM-signal. The software seems to take mainly the LSB and the USB signal part and after this, for AM are LSB and USB added in the right phase situation, to "produce" AM.

For this technique, You need not generally a special hardware signal, it could be done very easily by the software.

Some softwares for SDRs have this technique. Because of that, it is only the installation in a software algorithm, to get LSB und USB.

We should wait for coming DSP-receivers. Kchibo mentioned a planned new receiver for SSB too. Have a look in the internet to the not yet produced receiver: KK-S7600L

With best regards

Wolfgang Hartmann, Nürnberg


jim_kr1s <jkearman@...>
 


--- In ultralightdx@..., "Wolfgang H" >
> Hallo Jim,
>
> as far as I understood the technique of decoding SSB, there is a relationship to AM-decoding.
>
> In AM, the software-author takes two signal parts of the total AM-signal after decoding and adds it to an AM-signal. The software seems to take mainly the LSB and the USB signal part and after this, for AM are LSB and USB added in the right phase situation, to "produce" AM.

Hi, Wolfgang,

I was assuming the G3 wasn't a DSP radio. I agree, DSP SSB demodulation would be more stable than using a separate BFO. The control chip would have to do more work, though. For SSB I think you want a tuning resolution of at least 50 Hz. I don't know how they do that on the G3. The Sony 7600GR has a continuous BFO adjustment, plus-or=minus 5 kHz, I think. But I was told it drifts, so it is probably analog. I guess you could switch in a second, "fine tuning" shaft encoder or rocker switch in SSB mode, that would reset to zero offset when you adjusted the main tuning encoder.

73,

Jim, KR1S
http://qrp.kearman.com/ 


lrdheat
 

Jim...
 
The G3, like the 7600 GR, has a clarifier wheel for the BFO...continuous adjustment as opposed to incremental. It does remain stable. ECSS tuning works, and is a good backup when lower sideband sync won't hold a weak signal. Wide filter lower sideband sync sometimes holds better on a weak signal. On moderate or strong signals, lower sideband sync works well. Upper sideband locks and holds even on weak signals. Not using sync, the lower sideband is a little muddier in sound than the more natural sounding upper sideband, but not a drastic difference. As far as mw reception, the G5 is sensitive across the entire mw band. The G3, by contrast, is sensitive below around 1100 KHz, and much less by the time frequencies above 1400 KHz are tuned in. The situation improves when tuning sync or ECSS style, but still remains less sensitive than the G5. What is outstanding is ECSS or sync to hear a station sandwiched between closely packed locals. I have locals on 1340 and 1370. I can hear a 5 KW station 140 miles away (daytime) on lower sideband 1360 KHz on the G3...can't seperate this from the racket on my other sets...even the TECSUN PL-310 on 1 KHz filter can only hear the 1360 station in the null of the 1370 local, and this is not ideal as the radio in that situation is not pointing to maximize the 1360 signal.
 
Heatwave

--- On Thu, 2/18/10, jim_kr1s wrote:

From: jim_kr1s
Subject: [ultralightdx] Re: SSB decoding in SDRs and as procedure (algorithm) in the digital software
To: ultralightdx@...
Date: Thursday, February 18, 2010, 4:38 PM

 

--- In ultralightdx@ yahoogroups. com, "Wolfgang H" wrote:
>
> Hallo Jim,
>
> as far as I understood the technique of decoding SSB, there is a relationship to AM-decoding.
>
> In AM, the software-author takes two signal parts of the total AM-signal after decoding and adds it to an AM-signal. The software seems to take mainly the LSB and the USB signal part and after this, for AM are LSB and USB added in the right phase situation, to "produce" AM.

Hi, Wolfgang,

I was assuming the G3 wasn't a DSP radio. I agree, DSP SSB demodulation would be more stable than using a separate BFO. The control chip would have to do more work, though. For SSB I think you want a tuning resolution of at least 50 Hz. I don't know how they do that on the G3. The Sony 7600GR has a continuous BFO adjustment, plus-or=minus 5 kHz, I think. But I was told it drifts, so it is probably analog. I guess you could switch in a second, "fine tuning" shaft encoder or rocker switch in SSB mode, that would reset to zero offset when you adjusted the main tuning encoder.

73,

Jim, KR1S
http://qrp.kearman. com/ 


jim_kr1s <jkearman@...>
 


--- In ultralightdx@..., Richard Berler
> The G3, by contrast, is sensitive below around 1100 KHz, and much less by the time frequencies above 1400 KHz are tuned in. The situation improves when tuning sync or ECSS style, but still remains less sensitive than the G5.

How does the G3 compare with the G5 on LW? It sure sounds as though the antenna isn't tuning well at the high end of the band. If there were too many turns on the antenna winding, LW and low-MW reception would be good, and perhaps better on the G3 than on the G5. I'd be tempted to do an internal antenna mod in that case. What the heck it's not a legal ULR anyway, might as well. :)

If the BFO is adjustable, then there is no need to align it. I can't explain why one sideband sounds different than the other, but whatever is causing that, combined with the IBOC protocol, would explain the effects you're hearing.

I had an ICF-2010 many moons ago, and never found sync detection that useful on the BCB, with so many signals. Once you get two or more carriers on one frequency, the sync detector doesn't know what to do. ECSS sometimes helps on weak signals, but what I'd really like to see on the next-gen of SiLabs chips is a heterodyne detector that would recognize a continuous tone (I guess it would have to work at audio freqs) and notch it in software. Trans-oceanic listeners would love that!

73,

Jim, KR1S
http://qrp.kearman.com/