On Tuesday, December 20, 2011 10:15:07 PM Fons Adriaensen did opine: > On Tue, Dec 20, 2011 at 06:24:13PM -0500, gene heskett wrote: > > On Tuesday, December 20, 2011 06:22:02 PM Fons Adriaensen did opine: > > > Mmm. An IP3 of 40dB means that at e.g. -10dB (already high as > > > an average level), IM is at -110dB. That is nowhere near to > > > 0.5%... > > > > Picky Fons :), but I swear I can hear it, or maybe I stated the > > terminology wrong? Or have I been listening to SSB radios too long? > : > :-) Have you ever listened to AM radio on headphones with > > LSB in the left channel and USB in the right one (and the > local oscillator synced to the AM carrier) ? It's absolutely > wonderful. All the interference is stereo, and the AM station > is exactly in the center which makes it very easy to ignore > the interference. Its been yonks, as in decades, and not in stereo, but a locked oscillator, and yes, even then I was surprised at the quality. Which I felt at the time could be largely attributed to the fact that the usual diode detector was being pumped by the locked oscillator, getting rid of the very non-linear detection characteristics from only having 3 or 4 volts of drive at dead carrier levels in the average radio. I think much of the same advantages can be had by a 4 quadrant multiplier type of detector, which separates the signal from the LO but gives what may be mathematically provable to be essentially identical except for the noise being in stereo in your example. Something built around the MC-1495 would be a good starter point, although I don't think its as linear as can be designed in discrete parts running in an oven. I'd consult with Bob Pease at National Semi., but his penchant for driving an old VW bug eventually wrote his obit about a year back. > Regarding IM distortion, I've found that in many cases the > cause is not the active analog electronics but non-linear > contact resistance in switches, pots and faders. A few years > ago I did a routine check on my analog mixer. It had horrible > IM on one of the channels wired to the soundcard. And the > reason was simply that the mic/line switch on that channel > hadn't been used for years. Just toggling it a few times > made the distortion go away... Oh my yes, lost track of the times over the last 60 years. And yet its rare enough to catch me up sometimes till I've slept on it. One other instance comes to mind, one of my employees had built an audio board for the tv station, using 'raysistors' for the switching elements ostensibly because the switching they could do was silent, unlike the relays & switches of the day (1984). But at nominally +4dbm through them, the intermod was pretty obvious. I pleaded for & bought a 12 + 4x a channel input multiplier (48 in, 12 of them mics) board from Logitech that used lamps & cds cells, much quieter intermod specs. But it was loaded with enough 5532's that a cop of coffee sitting on it was nothing but 1/4" of mud in the bottom in 2 hours. That heat killed the cds cells by the grocery bag full. A couple 5" rotrons to suck out the hot air cut that down to 1 or 2 a year, and some clip-on heat sinks helped the 5532's last for years. We ran that board till the digital switchover in mid 2008, 5.5 years after my watch that officially started in Oct '84 was officially over. I think Russ (the owner) grabbed it and sent it to one of his radio stations that needed a board. It was too big for our satellite uplinking production truck so it still has a 4 channel toy board in it IIRC. > Ciao, Cheers Fons, Gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) My web page: <http://coyoteden.dyndns-free.com:85/gene> Brain fried -- Core dumped _______________________________________________ Linux-audio-user mailing list Linux-audio-user@xxxxxxxxxxxxxxxxxxxx http://lists.linuxaudio.org/listinfo/linux-audio-user
