On Sun, 18 Oct 2009 20:09:38 -0400 w2xj <[email protected]> wrote: > Leif > > Actually I think you are over thinking this. The signal is really DSB > with one sideband being only a couple hundred hertz so it is not a new > mode. This would permit a Costas Loop to lock the received signal upon > reception. I already have speech processing so it really boils down to > two audio channels and two Hilbert transforms ouput as a single I/Q so I > dont know where all the power and filtering, etc comes in.. You make many assumptions and you have a particular application in mind. What you describe seems to be a DSB voice signal with one sideband having reduced bandwidth.
> It just feeds > a Softrock the same as any other mode of operation.. It could be > included one of many existing SDR programs. ISB is and has been a real > mode for many decades. As I mentioned, ISB is already on the list of > features to be added to the PowerSDR software if one wanted to wait. > Actually after my last post yesterday I did it with one Hilbert > transform and barely any more CPU processing than for SSB. After I > finish my experiments using a computer test bed, it will be > professionally coded into an ASIC. Now knowing what you want to do I can say for sure that it does not motivate "ISB" in Linrad. If there will be a request for DSB with different bandwidths I might consider it. (But a discussion of the mode would be required. See below.) Linrad has an advanced speech processor (which does not yet have any particularly userfriendly interface) and it would obviously be possible to add a couple of hundred Hz of the opposite sideband. That would however destroy the very good peak to average power ratio of the signal and degrade the DX properties. The correct way to generate the signal you ask for would be to generate it as an audio signal with the suppressed carrier at e.g. 1 kHz. That is trivial. The signal can then be sent into an ordinary SSB transmitter such as Linrad which would apply RF clipping followed by a filter to make the peak to average power ratio good. The filter would allow the full sideband on one side and the desired part of the sideband on the other side. There would be intermodulation products between the sidebands produced by the RF clipper and one might like to filter those out also. May I say I hope that this mode will not become popular among amateurs. Spectrum is valuable and this mode does not use it well. (Provided that I understood your cryptic description correctly. You or others might convince me that there are advantages.) Using ALC to regulate the power of the signal you describe is not a good idea. The initial peak to average power ratio is too large and ALC (which is wideband AM modulation) would create wideband splatter. 73 Leif / SM5BSZ
