It's ok to shift the focus on the GEB positions, I can't debate that. It's only that you mentioned:
I am debating the measurements part only.
I can give a good long break, however know that the "dirty sand" sintagm is Copyright Charles Hansen TM. He was mentioning the dirty silicon resistors in ICs that are nonlinear and hence the IC performance can't be anything but crap.
audiowolf said:All the classic API, NEVE, Quad-8, SSL and other major brands of consoles, EQs, compressors and such uses large scale discrete op-amps because the SOUND BETTER and measure better too than high voltage variants of the vast majority of monolithic op-amps.
I am debating the measurements part only.
audiowolf said:Gimme a break...discrete parts are made of the same dirty sand! and the monolithic are much easier and cheaper to use, meaning higher profit margins and greater production ease!
I can give a good long break, however know that the "dirty sand" sintagm is Copyright Charles Hansen TM. He was mentioning the dirty silicon resistors in ICs that are nonlinear and hence the IC performance can't be anything but crap.
dimitri, the source followers are direct coupled to the grids. That's one huge advantage to using them- no blocking distortion. And it also means that the phase splitter is always driving a symmetric load.
When that happens (as opposed to when it's claimed to happen!), it's because of straightforward causes. Excess noise in carbon resistors in phono stages. High VCR in feedback resistors. All audible and all easily measurable.
yet some people can reproducibly detect the exchange of one type for another.
When that happens (as opposed to when it's claimed to happen!), it's because of straightforward causes. Excess noise in carbon resistors in phono stages. High VCR in feedback resistors. All audible and all easily measurable.
Edmond Stuart said::bs: At this time, all known amps with the lowest distortion operate in class-B or class-AB. Halcro or PGP, for example.
edit: Class-heat is for less talented designers.
See my amp here:
http://www.diyaudio.com/forums/showthread.php?postid=1753120#post1753120
The schematic shown in that thread is a little old (needs to be updated) and I've since made an attempt to "modularize" this design.
Class A and lots of fast output trannies in parallel. The front-end (input/VAS) circuit is working very nicely and I'm currently working on the heatsink. If the triple EF OPS circuit board works out nice I have a preliminary design for a unity gain CFB board/module to stick in middle, which will incorporate the tripe EF OPS and provide much better buffering and lower distortion from the VAS of the front-end. What you you think of that plan? Any hints/tips/suggestions?
Cheers,
Glen
perfect vs optimal
John,
What are you talking about? Which post are you referring to?
Assuming that it is still about that 'PERFECT crossover' thingie, I didn't even use words like global feedback. I really advice you to read my posts AND yours more carefully, because it was you who caused the confusion by talking about perfect crossover, which simply does not exist. Apparently you mean optimal crossover.
As for the latter, please don't start explaining what that means, we know that already for years.
As for fools, aren't you one of them?
john curl said:Of course, Edmond, I am talking about the basic non-linearity of a specific output stage, not one with global negative feedback piled on top of it. Any fool can use feedback.
John,
What are you talking about? Which post are you referring to?
Assuming that it is still about that 'PERFECT crossover' thingie, I didn't even use words like global feedback. I really advice you to read my posts AND yours more carefully, because it was you who caused the confusion by talking about perfect crossover, which simply does not exist. Apparently you mean optimal crossover.
As for the latter, please don't start explaining what that means, we know that already for years.
As for fools, aren't you one of them?
PMA said:Save your flags, Edmond. 'All known amps' .... how funny.
If you know better, enlighten us with an example of a sub ppm class-A amp, please!
(btw, I'm talking about power amps, not baby amps)
The Kleinschmidt 25A
Glen,
Don't ask me for hints/tips/suggestions, as you know already what I think about class-heat and tubes. Nevertheless, good luck with "The Kleinschmidt 25A'
Cheers,
Edmond.
PS: Nice to see you are using THAT chips too.
In my TCP amp (still in the design phase) I will use three of them.
Glen,
Don't ask me for hints/tips/suggestions, as you know already what I think about class-heat and tubes. Nevertheless, good luck with "The Kleinschmidt 25A'
Cheers,
Edmond.
PS: Nice to see you are using THAT chips too.
In my TCP amp (still in the design phase) I will use three of them.
Edmond Stuart said:Glen,
Don't ask me for hints/tips/suggestions, as you know already what I think about class-heat and tubes.
Well that is a pretty lame response I was thinking perhaps something along the lines of a discussion on the bandwidth limitation (of the global loop) imposed by an OPS using its own compensated negative feedback loop.
That would be something to resurrect the Negative Feedback thread for.
Cheers,
Glen
andy_c said:
Okay, I found the article that John was talking about on Walt's site here (PDF download). It describes the capacitance test bridge technique.
Thanks Andy.
Jan Didden
Edmond, what on Earth are you talking about? What is your problem? Why me? Why here?
Why interfere with me giving a little bit of insight of using Class A ?
All that I am saying is that Class B, by its very nature, has NO inherent distortion canceling mechanism. Class A has an inherent distortion canceling mechanism, due to the fact that it can sum two separate outputs with complementary distortion at the same time. Even harmonics are normally present in tubes, bipolar transistors and fets and usually the most dominant distortion found in a single device. When two similar devices are used in Class A push-pull, either by connection with a transformer or directly connected in some fashion, their similar even order distortion that they generate (by the specific laws of their physics) can be almost completely be eliminated as if it didn't exist in the first place. True cancellation.
Now Edmond, would you like to tell us what happens when we use Class B?
Is it similar? Is there any real cancellation of harmonics? Please give examples.
Why interfere with me giving a little bit of insight of using Class A ?
All that I am saying is that Class B, by its very nature, has NO inherent distortion canceling mechanism. Class A has an inherent distortion canceling mechanism, due to the fact that it can sum two separate outputs with complementary distortion at the same time. Even harmonics are normally present in tubes, bipolar transistors and fets and usually the most dominant distortion found in a single device. When two similar devices are used in Class A push-pull, either by connection with a transformer or directly connected in some fashion, their similar even order distortion that they generate (by the specific laws of their physics) can be almost completely be eliminated as if it didn't exist in the first place. True cancellation.
Now Edmond, would you like to tell us what happens when we use Class B?
Is it similar? Is there any real cancellation of harmonics? Please give examples.
class-heat
John,
First, please, don't use words like 'teaching'. Sounds rather arrogant.
Second, no need to explain that class-A (provided that it is proper implemented) performs better. The point is that for high power applications it's just impratical, too costly and a bloody waste of energy. Moreover, we don't need it, as we have an arsenal of other tricks to bring down the distortion to such a low level that it is absolutely inaudible.
And here again some nitpicking: "if the switching is done perfectly" should read as "if the switching is done optimally." You can't do it "perfectly", as already pointed out.
john curl said:Bob, IF you agree that ALL ELSE BEING EQUAL a Class A output stage will measure better than a Class B output stage, even if the switching is done perfectly, then we are saying the same thing.
I am teaching class here, not disparaging anyones opinions or efforts.
John,
First, please, don't use words like 'teaching'. Sounds rather arrogant.
Second, no need to explain that class-A (provided that it is proper implemented) performs better. The point is that for high power applications it's just impratical, too costly and a bloody waste of energy. Moreover, we don't need it, as we have an arsenal of other tricks to bring down the distortion to such a low level that it is absolutely inaudible.
And here again some nitpicking: "if the switching is done perfectly" should read as "if the switching is done optimally." You can't do it "perfectly", as already pointed out.
Edmond -
who 'we don't need it' ?
Have you ever done comparative listening tests of low idle amps against class A, both 'properly designed' ?
If yes, tell us the difference. If not, do not speculate about inaudibility.
You do need to make all power range class A. You may stay at such low idle current like 1A to cover first 10 - 20W in a class A. Still, the difference is high compared to cold amplifiers.
who 'we don't need it' ?
Have you ever done comparative listening tests of low idle amps against class A, both 'properly designed' ?
If yes, tell us the difference. If not, do not speculate about inaudibility.
You do need to make all power range class A. You may stay at such low idle current like 1A to cover first 10 - 20W in a class A. Still, the difference is high compared to cold amplifiers.
Mr. Curl, do you mean the positive effect of classA to supply rails? ClassA will have constant draw of supply rail, like putting a very large constant shunt regulator to power supply. If the supply rail relatively cleaner, with the same PSRR, the one with cleaner supply will have less effect of power supply rail distortion to output node.
john curl said:Edmond, what on Earth are you talking about? What is your problem? Why me? Why here?
blah, blah blah
Read my post again, starting from here:http://www.diyaudio.com/forums/showthread.php?postid=1753976#post1753976
and maybe you will understand what I'm talking about.
Hint: perfect is NOT the same as optimal.
And don't teach me about the difference between class-A and B. Do you really think I'm such an ignorant AH, or any other member of this forum, who doesn't know that since eternity?
Only someone who is blinded by his own brilliancy would think so.
Edmond Stuart said:
If you know better, enlighten us with an example of a sub ppm class-A amp, please!
(btw, I'm talking about power amps, not baby amps)
Maybe the Sassen/Putzey Class-A design could qualify.
Jakob2 said:
Maybe the Sassen/Putzey Class-A design could qualify.
Hello Jakob2
The Sassen/Putzey Class-A needs error correction on the output stage (which is Class-A biased) to achieve its results without the error correction circuitry the distortion is at least 20db worse.
So its performance is not achieved without error correction on the output stage.
Regards
Arthur
semantics
http://www.diyaudio.com/forums/showthread.php?postid=1754057#post1754057
and let's stop this fruitless discussion.
John, read this again:john curl said:What Edmond, changing your tune? Come now, give me some useful input.
http://www.diyaudio.com/forums/showthread.php?postid=1754057#post1754057
and let's stop this fruitless discussion.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Solid State
- John Curl's Blowtorch preamplifier