RMAF show report and pictures
For those interested, I've put a Rocky Mountain Audio Fest show report and a number of pictures up on my web site at www.cordellaudio.com under the RMAF Workshops tab. Peter and Darren and I had a blast there, both giving the workshops and seeing the show. It is a great show to go to, if you ever get a chance.
I also discovered that I had left off Part Two of my paper, "Another View of TIM" under Published Papers, and that is now corrected.
Bob
For those interested, I've put a Rocky Mountain Audio Fest show report and a number of pictures up on my web site at www.cordellaudio.com under the RMAF Workshops tab. Peter and Darren and I had a blast there, both giving the workshops and seeing the show. It is a great show to go to, if you ever get a chance.
I also discovered that I had left off Part Two of my paper, "Another View of TIM" under Published Papers, and that is now corrected.
Bob
Bob Cordell said:
I would suggest this has more to do with the old psychological mechanism: ''newer is always better''.
I have had an OPA627-based preamp compared with an otherwise identical NE5534-equipped version in casual blind listening tests; no difference could be detected.
Re: Folded emitter followers
I am not familiar with it in practice.
What is surprising or unexpected in HF behaviuor of this circuit?
regards
Adam
Bob Cordell said:
I am not familiar with it in practice.
What is surprising or unexpected in HF behaviuor of this circuit?
regards
Adam
lumanauw said:
Here is another, perhaps better way to do the experiment: Take a very good non-feedback design like the Ayre and compare it to a good feedback design of slightly higher power rating with the following changes made to it: add an output resistance to make the damping factors about the same, and add a soft clip circuit to the input of the feedback amplifier and adjust it so that both amplifiers hit, say, 0.3% 1 kHz THD at the same power level. Then do a listening comparison of these two amplifiers where the levels, as read at the speaker's terminals, are matched to within 0.1 dB. Also be sure that the small-signal frequency responses, as measured at the loudspeaker terminals, are within about 0.1 dB of each other for both amplifiers across the band. This latter may or may not be possible if the Non-feedback amplifier's damping factor changes a lot in-band.
We all know, of course, that something sounding different is not necessarily better, and that something sounding better is not necessarily the same as being more faithful. This is where the matter of taste and initial sonic impressions come in.
Whenever I try to think of why classic vacuum tube amplifiers (e.g., Class AB push-pull) may sound different from solid state amplifiers, I think of differences in damping factor and clipping behavior first.
Bob
Bob Cordell said:
I think this damping factor is pretty much overrated. The voice coil is accelerated and stopped by current (hopefuly, and as far as the amplifier is concerned), and an average 8 Ohm speaker has abt. 4-6 Ohm DC resistance. Now let your amplifier have 1 mOhm or 10 or 100 mOhm source impedance, -- it makes no difference. A value as bad as 100 mOhm would count as confession of incompetence.
regards, Gerhard
As a factoid, my B&W 804 S features 4.4 Ohms at DC (just checked), and although not measured myself, it won't drop below that over frequency IIR the data sheet correctly.
Why not take good sounding NFB tube amplifier and add switchable FB path?
In my expirience the sound is worse with FB and I can easily hear it as slight loss of low level details (piano tone decay, cymbal shimmer etc.). 😕
In my expirience the sound is worse with FB and I can easily hear it as slight loss of low level details (piano tone decay, cymbal shimmer etc.). 😕
vuki said:
What do you mean with NFB, NoFeedback or NegativeFeedback?
Anyway, this won't work, because in order to apply feedback you need to have a lot of excess gain and this won't be there in a finished amplifier.
Gerhard
vuki said:
This is the worst way to use feedback ! Of course your amp will sound worse... If feedback, do it correctly and use at least maybe 40db.
A common mistake with feedback is using not enough of it, or too much of it making the amp only conditionally stable.
Mike
Bob Cordell said:
Bob, such clipping factor that causes higher audible IMD on high power is the price we pay for low distortions on low power, especially on high end of the audio spectrum. Well designed single ended class A amplifier may sound "tubey" without such "clipping factor", that is the proof, at least for me and some audio engineers who agree with me, that low order clipping factor is not what characterizes the tube sound. However, diode limiters I use on inputs of my amps may clip similarly, but I mean the sound on levels below the clipping level.
gerhard said:
Maybe not that simple...
Have a look at http://www.mhsoft.nl/spk_calc.asp , go down to "New Qts with series inductor", enter the datas from your woofer, and a series resistor. Observe how quick the Qts changes and messes up the tuning of your speaker.
But please, leat us not start the DF-war here...
Mike
MikeB said:
And even in this case some small details help drammatically, for example 2 mA bias from positive power rail to the output of 5532.
Hi Wavebourn,
-Chris
Not all of us agree with that practice, and certainly not with all op amps. This may help a class B output though.
-Chris
anatech said:
Do you mean some mentioned in my example 5532 chips already have emitter followers loaded by current sources?
I want them! 😀
Hi Wavebourn,
The old uA739 / uA749 IC's used a pull down resistor or current source depending on the model.
Loading the output of an op amp was a fashion a while ago, no doubt there are still people who do that. I don't recommend it unless you have an op amp that suffers from crossover distortion. The 5532 doesn't.
-Chris
The old uA739 / uA749 IC's used a pull down resistor or current source depending on the model.
Loading the output of an op amp was a fashion a while ago, no doubt there are still people who do that. I don't recommend it unless you have an op amp that suffers from crossover distortion. The 5532 doesn't.
-Chris
anatech said:
Sometimes does. 😉
Chris, let's stop offtopics? We don't speak of fashions and don't ask your opinions about to bias or not to bias, we discuss amps and distortions in terms of Bob's interviews.
Hi, Bob Cordell,
Thanks for the info 😀
What is a soft clip cct in input of feedback amp looks like? Do you have a drawing of this?
