Rob, In you message you said:
"Flex really needs to understand that it's current AGC implementation is not acceptable and in fact most likely violates OSHA regulations..." Whether or not the AGC implementation is acceptable or not is a personal opinion, for the sound pressure levels to be violating OSHA (OCCUPATIONAL Safety and Health Administration) regulations would require that the radio be used in the workplace and the levels exceed 115 dBA continuous or 140 dB SPL impulse. I personally have never experienced a problem with excessive noise in my headsets but I live in a rural setting with the nearest KW+ about four miles away. Having been a "professional" CW op in a previous life, I have had my ears blasted and can sympathize with your complaint. 73, Bob, K5KDN -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Robert Dennison Sent: Sunday, May 11, 2008 1:57 PM To: [EMAIL PROTECTED]; flexradio@flex-radio.biz Subject: Re: [Flexradio] AGC Hi Dale, Sorry to take so long getting back.. I also like the AGC-T very much. Here's the problem as I see it. Most modern radios implement AGC as an active feedback control system which maintains signal level at the set point the op picks by juggling AF and RF gain. The operative words are "FEEDBACK CONTROL SYSTEM." Having read the knowledge base, all of FlexRadio's information indicates that their implementation is one of changing gain factors using AGC-T to pick the point at which gain changes. This is a deeply flawed implementation. In fact, it is dangerous to my ears. Consider receiving DX CW at -90dBm, I pick an AGC threshold of 72 and an AF value of 72. This gives me comfortable audio volume and very little noise. Now a local with a beam and amp calls him. At my QTH the local's sig strength is -60 dbm. That means, even if the PSDR AGC gain is unity above the threshold. local's signal is say 30 dBm or 1000 times louder in my ear This is intolerable not to say painful. If you carefully read Knowledge Base Article 10403, it even says as much: "This may not always be desirable however. For example, consider participation in a roundtable. You may find that this large difference between signal volume levels causes you to keep adjusting your AF [gain] constantly, in order to compensate for the very quiet (yet very readable) stations, and the louder stations. In that case, you may want to increase the AGC-T and back off on the AF [gain] so that the volume levels of all signals will be much more uniform although with a higher background level on the weaker stations." The problem in the DX CW case is you're screaming before you can find the mouse to make a change! Flex really needs to understand that it's current AGC implementation is not acceptable and in fact most likely violates OSHA regulations... A simple feedback control loop should not be hard to implement in floating point SW.. very best regards to all Rob AB7CF On Mon, 05 May 2008 18:21:28 -0400 Dale Boresz <[EMAIL PROTECTED]> writes: > Hi Rob, > > A few minutes ago, I was listening to a round table on 20m > (miserable > band condx.!). One station was about S8, one was about S6 and one > was > just at the noise floor at about S2. With my usual AGC-T setting of > 70, > I could comfortably copy the S8 and S6 stations, but the S2 station > was > just a whisper. When I turned up the volume to hear the S2 station, > the > S8 station was way too loud. So, I increased the AGC-T setting to > 90, > which caused me to turn the AF gain down a little bit because the > background noise was a bit too loud, however I could then copy all > three > stations at about the same level. True, the S2 station still had a > lot > of noise accompanying it that the S8 station did not, but their > actual > level in my headphones was about the same. Personally, I don't like > > setting AGC-T that high (PowerSDR's default setting) because I find > the > background noise to be irritating and fatiguing, so I usually run it > at > about the same level as you. However, I do occasionally get blasted > by a > loud station in the passband. > > I think most radios set their AGC threshold similar to PowerSDR when > it > is set to 90, so that all signals are at about the same audio level > in > the headphones or speakers. The only thing that changes is the > amount of > noise that accompanies the signal. I really like the fact that > PowerSDR > allows us to change that threshold (and essentially sneak up on > AGC-OFF > without quite getting there) because lowering the number drastically > > reduces noise level between signal peaks. The downside to that > though is > the increased possibility of getting blasted by a strong station. > I've > learned the hard way, to never turn the AF gain up too high while > the > AGC-T setting is set low. If I know that I'm going to have to dig > for > weak sig's, I bump up the AGC-T to the mid-80 range, and turn down > the > audio gain to make the levels comfortable again. > > vy 73, Dale > WA8SRA > > > > > Robert Dennison wrote: > > Hi Dale, > > > > Thanks for the response. That's about how I thought the AGC-T > control > > works. Still it seems to me either PowerSDR's response is > extremely > > slow, or the AGC compression curve is linear not exponential. > > > > In the prior case the problem may lie in the Windoze audio system, > in the > > other case they have some algorithm work to do.. maybe both... > > > > As it stands, I think Flex is leaving potential liability on the > table! > > I've never had an experience where I had to rip my phones off in > > thousands of hours of air time! > > > > vy 73 > > Rob > > > > > > On Mon, 05 May 2008 08:40:06 -0400 Dale Boresz > <[EMAIL PROTECTED]> > > writes: > > > >> Rob, > >> > >> One of the great things about PowerSDR is that you can so easily > > >> adjust > >> the AGC threshold to accomodate band conditions. Remember though, > > >> the > >> lower the AGC-T setting number, the less the AGC will be > compressing > >> the > >> audio gain, therefore, the greater the audible difference in > >> amplitude > >> between weak and strong signals, and the lower the background > noise > >> > >> between signal peaks. > >> > >> If you increase the AGC-T setting to something more like 75 or > 80, > >> the > >> effective audio gain compression is increased considerably, (at > the > >> > >> expense of hearing more band noise relative to signal levels), > and > >> the > >> differences between weak and strong signals will become much > less. > >> Overall it's a noisier listening experience, but you won't be > jolted > >> by > >> the big signals. > >> > >> > >> 73, Dale > >> WA8SRA > >> > >> > >> Robert Dennison wrote: > >> > >>> Sheesh, > >>> > >>> Just got my ears blasted out by a S9++ sig. Was listening to a > > >>> > >> QSO one > >> > >>> side was about an S5 he was answered by a guy with an S9++ sig > >>> > >> mebbe > >> > >>> -70dbm--. I mean it hurt! Now I thought one of the ideas of > AGC > >>> > >> was to > >> > >>> keep that from happening! Am I wrong?? I as far as I could > tell > >>> > >> the AGC > >> > >>> just didn't respond at all... > >>> > >>> My settings are the ones I normally use with my 1K on 80M: AGC: > > >>> > >> Med, > >> > >>> AGC-T: 64 (seems to have been the sweet spot recently) AF: 80. > I > >>> > >> must > >> > >>> have something set wrong... > >>> > >>> Help?! > >>> > >>> vy 73's > >>> Rob > >>> AB7CF > >>> > >>> _______________________________________________ > >>> FlexRadio Systems Mailing List > >>> FlexRadio@flex-radio.biz > >>> > >>> > >> > http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz > >> > >>> Archives: > http://www.mail-archive.com/flexradio%40flex-radio.biz/ > >>> Knowledge Base: http://kb.flex-radio.com/ Homepage: > >>> > >> http://www.flex-radio.com/ > >> > >>> > >>> > >>> > >> > >> > > > > > > > > > _______________________________________________ FlexRadio Systems Mailing List FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archives: http://www.mail-archive.com/flexradio%40flex-radio.biz/ Knowledge Base: http://kb.flex-radio.com/ Homepage: http://www.flex-radio.com/ _______________________________________________ FlexRadio Systems Mailing List FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archives: http://www.mail-archive.com/flexradio%40flex-radio.biz/ Knowledge Base: http://kb.flex-radio.com/ Homepage: http://www.flex-radio.com/
