Gents, Bill wrote: "example is in the difference in the sound and behavior of the various versions of the common 5532 and 5534 opamp. Originally a TI device and also Signetics mfg." Comment: Below is a link about modyifing CD players where the op amp after the D-A has a tremendous effect on the perceived audio quality. One explanation is that many op-amps have very noticeable cross over distortion, which becomes more significant as the signal level drops into the lower portion of the available dynamic range. One solution for some op-amps is to use a pull up, or down resistor from the output of the op-amp to a power buss. The idea is to require the output stage to sink, or source a little current to in effect keep only one half of the output running, or to put it in other words: make it act as a single ended amplifier, and therefore no cross over distortion. You will also like the use of tubes following the D-A converter. Neat stuff! Jim WD5JKO The reference here is at the bottom of this web page:
http://home.netcom.com/~wa2ise/radios/tubedac.htm Jim candela <[EMAIL PROTECTED]> wrote: Gents, While going through my parallel connected 6L6 heising plate modulator in my Gonset G50, I too ran into gain compression whenever the plate voltage swing went below about 200 volts (had B+ at about 400, and G2 slightly below with a small series R, about 100 ohms). I wanted to maximize the dynamic range of the modulator to get more +/- modulation. In stock form the G50 modulator goes into limiting at about +50, -60 % modulation. Instead of fixing this, Gonset featured it as a high level limiter (that's marketing for you ---MBA's!). I got the + modulation up by running the 6L6's at maximum plate dissipation (decrease cathode resistor on TX, increase it on RX), but the (-) modulation suffered the gain compression prior to clipping. I found something similar to what Patrick did with the 6GH8's. Adding a screen dropping resistor to the paralleled 6L6's seemed to do the trick. It is important to note that keeping the screen unbypassed was key. I found that the symmetry of a sine wave was maintained to about 80% +/- this way. For me a value of 7K was the compromise between maximum power, and max symmetry. Bypassing G2 brought back the asymmetry. My theory here is that whenever the plate voltage got too low (during audio swing), the screen current rose, and progressively took over as the plate; only the plate was connected to the heising reactor (top end) so hence the gain compression. The series R, and unbypassed G2 kept the plate in charge because the screen voltage (instant by instant) stayed below the plate voltage. Maybe this is all hog wash, but heck it works for me! I can see how changing a tube can alter the DC operating point, and this will have an effect on the distortion products of an amplifier, especially if the tube is driven such that the available dynamic range is being used. Hmmmm. By a changing to a Chinese 12A_7, shift the point, gain compress on one side, and then start writing stories on how wonderful the thing sounds! BTW gain compression on one side can be a valuable tool in a super modulated AM rig...... Regards, Jim -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of [EMAIL PROTECTED] Sent: Wednesday, October 01, 2003 10:24 PM To: [email protected] Subject: Re: [AMRadio] Testing tubes OK Guys, I forgot I was addressing Radio ops, LOL ! I should have said most Audio tubes. I did real and subjective tests on the standard audio preamp tubes using a Sond Technology 1700 Distortion analyzer and Variable HT and Bias supply several years ago. Since it was just personal interest in finding good sounding tubes I really didn't document anything, just chose the best testing lowest noise tubes then did a subjective listening matchup in my amps. Now the test gear has been in storage since and I am setting it up again to do more advanced and documented testing including square wave and pulse behavior. Will post some actual figures before too long. But I agree that really there should not be an audible difference within tubes of the same type operating in the same identical voltages. My ears and experience with tube audio tell a different story tho and I also would like to see why there are noticable fidelity differences in various types of plate structure and mfg. I have a few suspicions that some of the difference may be caused by a change in plate resistance caused by differences in cathode composition, quality of mfg., and deterioration which would explain why NOS from the "Glory Days" of tube mfg. seem to hold up the best and are the most consistent. An interesting example is the now no longer available China Mfg 12AX7's that were highly regarded in audio but when used in guitar amplifiers for a year would still test almost as new but would be so dead sounding they were useless. I will include some of those that I still have in my testing. A little non tube issue but similiar. Published spec and actual real world function of many devices do not necessarily agree with each other as a verifiable example is in the difference in the sound and behavior of the various versions of the common 5532 and 5534 opamp. Originally a TI device and also Signetics mfg. I have yet to find a currently made version from EXAR, NEC, etc that exibits the same characteristics in regards to output impedance and current capability, i.e. sound. Even tho these are "supposed" to be the same device. Sorry for the subjective length of this but I will follow up with hard data as soon as I can derive it. Thanks for listening ! Bill KB3DKS --- StripMime Report -- processed MIME parts --- multipart/alternative text/plain (text body -- kept) text/html The reason this message is shown is because the post was in HTML or had an attachment. Attachments are not allowed. To learn how to post in Plain-Text go to: http://www.expita.com/nomime.html --- _______________________________________________ AMRadio mailing list [email protected] http://mailman.qth.net/mailman/listinfo/amradio _______________________________________________ AMRadio mailing list [email protected] http://mailman.qth.net/mailman/listinfo/amradio --- StripMime Report -- processed MIME parts --- multipart/alternative text/plain (text body -- kept) text/html The reason this message is shown is because the post was in HTML or had an attachment. Attachments are not allowed. To learn how to post in Plain-Text go to: http://www.expita.com/nomime.html ---
