Deang Posted October 1, 2003 Share Posted October 1, 2003 True or False? "A large factor as to which is better is really the issue of negative feedback, used to keep solid state amplifiers and pentode tube amps stable and low in distortion. Triodes have a different operating character in that they do not require negative feedback resulting in a usually more transparent sound with greater clarity and better imaging. Lifting the negative feedback off most pentode amplifiers yields a power curve that rises with impedance. Since most speakers have an impedance that climbs with frequency, the pentode will put out more current as the frequency increases. The result of this is a thin dry ultra fast sound with lots of spank. A monster that makes some recordings sound interesting for the first time, and most everything else sound like crap." The statements are made by a well known commercial designer of SE amps. Find the flaws -- there are several. Quote Link to comment Share on other sites More sharing options...
NOSValves Posted October 1, 2003 Share Posted October 1, 2003 I think it might be a much bigger challenge to find what part isn't flawed in some way ! Quote Link to comment Share on other sites More sharing options...
mobile homeless Posted October 1, 2003 Share Posted October 1, 2003 Jesus Dean.... You and your speaker impedance... You read so much yet always take the wrong parts literally and drop the right parts to the wayside. There were some GOOD things in that quote. Why are you so hung up on this stuff? Listen! Lord. NEGATIVE FEEDBACK is not ALWAYS bad. Almost all designers use it in one form or another. Mark and I have gone around with this again but one thing is for sure in my eyes. Negative Feedback was used TOO much in the past to bring distortion measurements down. A lot of these amps sounded like glassy crap. And global negative feedback became a no-no with local versions coming out more. I think a little negative feedback can be a good thing in some amps. On the other hand, in the very low watt single-ended triode amps, I have found that I dont like the negative feedback dialed in at all, almost ALWAYS preferring ZERO FEEDBACK over even 3dB of dialed in Negative. And I have heard negative feedback removed from some PP amps sound more open and natural but if done in the wrong design, lose control of things as well. A little negative feedback can go a long way...too much is a measuring boon and a sonic *****. You seem to be after RIGHT or WRONG or ALL OR NONE answers to all this. Take the stuff you have learned and said in the past and put it to use. There are MANY MANY variables and circumstances and situations, not to mention the combinations of gear and amp/speaker interaction. Saying ZERO NEGATIVE FEEDBACK is WRONG is not right either. It all depends on the design and the implementation. The EITHER/OR ....RIGHT/WRONG is too BLACK AND WHITE. kh Quote Link to comment Share on other sites More sharing options...
fini Posted October 1, 2003 Share Posted October 1, 2003 All I know is, once you leave negative feedback, you can't change it. So you'd better think twice, there. fini Quote Link to comment Share on other sites More sharing options...
mobile homeless Posted October 1, 2003 Share Posted October 1, 2003 heh.... indeed. Sadly enough, I went into this thread thinking that Dean-O was going to have some amusing story involving Negative Feedback via eBAY or Audiogon. I was severely depressed to see the above instead. kh Quote Link to comment Share on other sites More sharing options...
Deang Posted October 2, 2003 Author Share Posted October 2, 2003 "You and your speaker impedance... You read so much yet always take the wrong parts literally and drop the right parts to the wayside. There were some GOOD things in that quote." Not the kind of response I was trying to evoke out of anyone. The specifics regarding feedback haven't been talked about in some time, and there are quite a few new people on the site who I thought might be interested. I wanted some interesting discussion, and my intent wasn't aimed at starting a 'hot' thread. I didn't say a thing about speaker impedance, nor was it specifically brought to mind when I read the comments by the writer. However, since you brought it up -- I do find it ironic that he applies the specific statement regarding the lifting of feedback, and the resulting increase in current at higher impedance/high frequencies, to pentode PP amps -- when in fact the statement applies to ALL tube amps without feedback -- including SE designs (which incidently, can be designed with feedback, and benefit from it as well -- yet he would have one believe otherwise). I think your opening comment to me is off-base. However, I do appreciate your recognition that I read quite a bit -- and I do. So, tell me, what "right parts" did I throw to the wayside? There are no "right" parts. Every sentence in the paragraph is a half-truth. As far as taking "wrong parts literally" -- words are vehicles of thought, and when one writes for the purpose of educating the reader -- should present all of the facts as best as possible. There are no literary devices being employed by this guy -- he's writing literally, and so I read it literally. So what do you mean? That I take the parts that are incorrect and make issue out of them, or that I take the parts that are somewhat unimportant in comparison to the big picture? Either way -- fact twisting and half truths are what they are, and when I see them -- I recognize the content as a whole for what it is (primarily bullsh!t). What I absolutely don't understand is why you don't hold this writer to the same standard that you hold me, or others here. If I were to make these type of "blanket" statements there would be no mercy. At any rate, there is no reason to be "depressed". BTW, with the exception of the insults you heaped upon me -- your post had more truth in it than the writers. Freak. Quote Link to comment Share on other sites More sharing options...
JohnA Posted October 2, 2003 Share Posted October 2, 2003 "Every sentence in the paragraph is a half-truth. " Probably the most accurate thing said, so far. Negative feedback is good, used sparingly. I don't know the difference in pentode and triode sound, but I fail to see how removing it from an amp designed around it would cause anything but sucky results. Quote Link to comment Share on other sites More sharing options...
maxg Posted October 2, 2003 Share Posted October 2, 2003 Maybe it is my simple lack of knowledge but... "A large factor as to which is better is really the issue of negative feedback, used to keep solid state amplifiers and pentode tube amps stable and low in distortion." That's true isnt it - at least that is what I thought negative feedback was for. "Triodes have a different operating character in that they do not require negative feedback resulting in a usually more transparent sound with greater clarity and better imaging." Again true in the first part - triodes dont need negative feedback. This doesnt say that non have it. As for that being the reason they have greater clarity and better imaging I cannot say. Actually I cant even say they always have either of those benefits - my experience is way too limited. "Lifting the negative feedback off most pentode amplifiers yields a power curve that rises with impedance. Since most speakers have an impedance that climbs with frequency, the pentode will put out more current as the frequency increases. The result of this is a thin dry ultra fast sound with lots of spank. A monster that makes some recordings sound interesting for the first time, and most everything else sound like crap." This I have no idea about. It seems to me that as long as that rising current is within the capabilities of the amp (in other words we are not heading into clipping) then the ultra dry, fast sound with lots of spank may not apply. I cant even take issue with the "most speakers have impedance that rises with frequency" although I do know of some exceptions that do the opposite (mainly ESL's). On the whole this read like an extract from some sales literature. Taken as such there may be some poetic licence taken, but that is hardly a unique crime... Quote Link to comment Share on other sites More sharing options...
NOSValves Posted October 2, 2003 Share Posted October 2, 2003 "A large factor as to which is better is really the issue of negative feedback, used to keep solid state amplifiers and pentode tube amps stable and low in distortion." No this is not true because he pretty much saying a amp devoid of feedback is better !! Totally subjective not based on fact. "Triodes have a different operating character in that they do not require negative feedback resulting in a usually more transparent sound with greater clarity and better imaging." No agian any tube can be used without Negative feedback and again "greater clarity and better imaging" Is subjective once again. "Lifting the negative feedback off most pentode amplifiers yields a power curve that rises with impedance. Since most speakers have an impedance that climbs with frequency, the pentode will put out more current as the frequency increases. The result of this is a thin dry ultra fast sound with lots of spank. A monster that makes some recordings sound interesting for the first time, and most everything else sound like crap." Lifting the feedback off any amp designed to run feedback will sound like crap The entire paragraph is complete bull dung Craig Quote Link to comment Share on other sites More sharing options...
mobile homeless Posted October 2, 2003 Share Posted October 2, 2003 I side more with Max's interpretation here, especially with the first part of the paragraph. AGain, more opinion. I happen to prefer low to ZERO feedback on all my tube amplifiers, ZERO with the single-ended triode options, and LOW with the PP amps. AGain, some feedback can be an ok thing dependong on use but I have found it abused in the past, especially when the SPEC wars were going on. Among SET manufacturers, 99% either employ ZERO feedback or have a dial to bring in from ZERO to 9dB or so. kh Quote Link to comment Share on other sites More sharing options...
leok Posted October 2, 2003 Share Posted October 2, 2003 In general, one can't just "lift" or "remove" negative feedback. An element (sub-circuit or entire amp) designed to use negative feedback will have very high open-loop gain. In order to remove the feedback, the gain must be reduced to a useable level, making it a different amp to begin with. leok Quote Link to comment Share on other sites More sharing options...
leok Posted October 2, 2003 Share Posted October 2, 2003 Mark, The distortion of a triode circuit with no feedback drops as amplitude goes down. I'm pretty certain that distortion in a feedback circuit hits a floor and depending on design, may begin to rise at very low levels. I am a good example of a SET fan who will tolerate some distortion at higher levels and impedance sensitivity, but I do so because the very low level details, sibilances, ambience, harmonics, bow sounds, etc, are likely to be less distorted, through my 101dB speakers, using a SET amp. The whistling and smearing of sibilances that I have heard on amps with feedback is simply not lower distortion than what I hear with the SET on those sounds. I accept that the feedback examples (Heath W5M and Dynaco ST-70, and even my own P6) may not be examples of what you consider properly executed feedback. But I simply haven't heard an amp that employs feedback, handle low level detail without adding noise and distortion that my SET does not add. leok Quote Link to comment Share on other sites More sharing options...
leok Posted October 2, 2003 Share Posted October 2, 2003 Actually, Mark, the improved low level performance extends to my pp/no-feedback implementation also .. not just the SET. leok Quote Link to comment Share on other sites More sharing options...
leok Posted October 2, 2003 Share Posted October 2, 2003 Mark, Until quantum effects kick in, triode distortion does approach zero as "power" apprpaches zero. I will see if I can find the KT66 plot to which I am refering and post it. Actually, it may be the ultra-linear plot. Well, I'll find it. Even including that, yes, tradeoffs. leok Quote Link to comment Share on other sites More sharing options...
leok Posted October 2, 2003 Share Posted October 2, 2003 Please see the last page. I believe all these are no-feedback plots. Note that only in triode mode does distortion continue to drop as power drops. I believe this has a lot to do with the appeal of amps based on no-feedback triode technology. leok KT66.pdf Quote Link to comment Share on other sites More sharing options...
sunnysal Posted October 2, 2003 Share Posted October 2, 2003 my thanks to both leo k and mark m for their posts, I am learning a lot and that is why I hang around here...regards, tony Quote Link to comment Share on other sites More sharing options...
leok Posted October 2, 2003 Share Posted October 2, 2003 I think it would be more like 90dB at 100mW (or 0.1W). When I had a scope to look at the signals going to the Fortes (99dB @ 1 W), I found the sound comfortably "loud" with signals that looked like about 1Vrms which is 1/8 Watt, which is 90dB with those speakers. With the Chorus-IIs this works out to more like 1/10 Watt for 90dB. That's why I always say I use about 1/10 Watt for loud stuff. 90dB isn't all that quiet. But, it also means that the low end details I like to have right, are in the 10s of milliWatts. From 1/10 of a Watt down, even that little KT66 is starting to look pretty good (and the pp amp uses 2 of them), while a lot of other amp topologies are starting to get pretty messy. I guess the nice thing about tubes is the sweet spot can be scaled so that bathtub distortion curve doesn't have to start rising until a very low power. However, that will probably restrict the max power. leok Quote Link to comment Share on other sites More sharing options...
NOSValves Posted October 2, 2003 Share Posted October 2, 2003 Leok, To me and maybe I'm nuts but your sound levels for watts make no sense what so ever. If a speaker is rated at 100db 1 watt 1 meter this is 3 feet from the speaker where exactly do you sit ? There is no way a 1/8th of a watt is going to make 90dB in any descent size room with even a 104 dB speaker. I have no Idea where your getting these figures from. I have often shrugged off your millawatt range for klipsch speakers statements but I just couldn't resist this one. Your math and your scope work just do not jive with mine at all. Craig Quote Link to comment Share on other sites More sharing options...
Soeren Basboell Posted October 2, 2003 Share Posted October 2, 2003 mdeneen writes, that NFB has some advantages, but in my mind several of them are more than doubtful, not in order: - it reduces the distortion; not really, it reduces the amplitude of the distortion, but it becomes of higher order and will thus be disharmonic to the fundamentals and/or behave like noise, much smaller quantities are percieved. This is even more true when we are looking at the intermodulation distortion, there you can have two frequencies, f1 and f2, and if the distortion products are near the f1 or f2 , they can be quite strong before percieved (wrap mask curve effect with a bad translation) -it decreases the output impedance and increases thus the damping factor; first the damping factor is a very doubtful thing as the DC resistance of the voice coil normally is much bigger than the internal resistance of any amplifier. In many 'normal' bass speakers it is even worse, as the the voice coil is several times higher as the flux gap. When we are introducing NFB the resistive part of the output impedance is being smaller, which is not som important, but the reactive part can change in a little bit unpredictible ways, which is certainly not an advantage. The other things are true in well designed equipment, but is not important as you can get it well enough with good components and a good design; but such a design will not tolerate errors as there is no correction. ---------------------- To make a photographic analogy: It is something like the difference between Leica and Hasselblad: The little format in Leica has far the best lenses and is therefore unforgiving; you get exactly what you deserve. Hasselblad is very forgiving, thick emulsions and soft 'flared' lenses: everything looks nice - and the same, something like a Bose speaker. Quote Link to comment Share on other sites More sharing options...
leok Posted October 2, 2003 Share Posted October 2, 2003 Mark, The Dynamic Range of 60 dB just as you stated. I misunderstood and thought that was the listening level. Yes, I generally listen at 100mW. When I want a few Watts I usually change to the P6. leok Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.