Two years ago before we redid the transmitter from the ground up to produce a larger punch audio chain, and have all of the stages metered, the compander was being overdriven by stages in front of it. There was an ad hoc scaling added by me to keep it from overdriving the stages beyond it, in particular, always overdriving the ALC.
Eric and I were working on control problems for the compander in the branch code and Eric did a fantastic analysis of the input and output using the "Phil Harman Spread Sheet Analysis Style". I am completely sold on this. I suggest everyone go the hamsdr download pages and get all of Phil's spreadsheets. They are really instructive. Eric took a page, he plotted the input and output and in five seconds, I knew what the problem was. I removed the scale and the compander now produces peaks that are full scale. The average power should be up by at least 2 dB. That is not insignificant when the lows in the voice can easily be slightly below the noise floor. This should improve intelligibility a lot. SO Dave is right. BE CAREFUL and watch your average power. You may need to do adjustment now that this faux pas has been removed from PowerSDR and dttsp v2.0 (all versions). Alex, VE3NEA, is my favorite software author for a fee in all of amateur radio. Every day he produces something brilliant to be emulated. For example, the new Flex5000 with its ability to generate a swept oscillator in front end, will use the Rocky algorithm in a flash (with huge signals) to almost perfectly balance the 5000 (image rejection). This is based on the simple formulae Alex has given us on his Rocky SDR page. There is much to emulate in Alex's VoiceShaper. I am sure we will be doing that. http://www.dxatlas.com/ and http://www.dxatlas.com/rocky/ The receiver will be able to listen to the transmitter and make nearly perfect image rejection and carrier suppression. When asked, it will be able to do backoff and predistortion to help eliminate IMD. The transmitter and predriver can be tuned (its bias) to run everything from Class A through C. Next, with the impulse generator in the 5000, I can estimate the impulse response. I have been doing some experiments where I correlate the perfect impulse response with incoming signals to see what the type i and type ii error probabilities look like on the noise pulse subtraction front. We have to balance this with effectiveness of course. It appears that Leif Åsbrink has the balance just about right in Linrad, so we will be extracted that work as well as his spur elimination work. You may expect to see this show up in PowerSDR and dttsp v2.0 in the coming days. This is pretty exciting. Bob Dave & Nancy Ridge wrote: > My findings with SDR-X SVN:1209. I work 6 meter weak signal stuff. I > want/need, on SSB, audio that is very penatrating but not distorted. > > I have a Remote Base (SDR) 17 mile from me, thanks to W4MO. I have spent many > hours listening to myself, watch the Panadapter on the Remote Base and > adjusting my audio for penetrating audio. > > I have listened on 20 meters for DX'er audio that has the characteristics > that I like. What I have found is the Icom ProIII and the Yaesu FT2000 has > audio that I like. This is with their compression turned on. I have tried to > measure their band width with the SDR Panadapter. I have found they run about > 200Hz to 2,800 Hz band width for the effect I like. I use no external > processing. > > My setting are: Transmitter band with 200 to 2,800Hz. Leveler On, Compander > on 3, Compressor on 3, Mic Gain 20, TX EQ on, setting of EQ Preamp 3rd mark > down from top...top mark being #1, Low 6 down, Mid 4 down, High 1 (top > mark...15db) and 160Hz Notch on. I did not think the notch would help running > 200 to 2,800 Hz Band Width, but it does. I use no external processing. Using > a Marshall MXL 2003 condenser Mic. > > The lastest change that Bob made (1209) really helped average power out with > very low distortion! Thanks Bob. > > I have tried VE3NEA Voice Shaper software, but much prefer the SDR on board > setting above. Watch your Mic Gain setting! Every rig is different, with a > multitude of Mics out there. > > Let me know what you think after having a friend listen to your audio. > > 73, Dave, W9DR > > -- AMSAT Director and VP Engineering. Member: ARRL, AMSAT-DL, TAPR, Packrats, NJQRP, QRP ARCI, QCWA, FRC. ARRL SDR WG Chair "If you're going to be crazy, you have to get paid for it or else you're going to be locked up." Hunter S. Thompson _______________________________________________ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ FlexRadio Knowledge Base: http://kb.flex-radio.com/ FlexRadio Homepage: http://www.flex-radio.com/
