You can guess the feedback from the simplified schematics. Do you see any global feedback loops? There is a small local feedback in the VAS/output stage to stabilize DC, and it results in lowered gain for the DC signal. Output stage is in high-biased AB. Even harmonics cancel. Odd harmonics have very fast decay with their order.
Last edited:
Mike, the problem with people's opinion here regarding good sound and bad sound is, they don't always refer to the same thing. When they say "good", it doesn't really matter. It must refer to the goodies we all know about. But "bad" has broader meaning. Even worse, the existence of certain kind of badness can be undetectable by others.
Do we really need to choose between transparent system and enjoyable system?
There's no perfect black and there's no perfect white (those who think that theirs is black, or white, are color blind. There are many of them). We choose the right blend of black and white. And it is different from person to person.
When we increase transparency, we use techniques that may risk the existence of unenjoyable sound. But never assume that these unenjoyable aspects always come from bad recording.
It is necessary to understand the various kinds of badness, as of to understand where they come from. One may put the blame on the bad recording on defense of his transparent system.
PMA, it is good to hear your input. Charles and I are pretty much in agreement on this as well, even though we are serious competitors, and we race each other for our place in audio society.
From a purely topological point of view, a complementary differential folded cascode design is the best I have found. Charles has moved on to proprietary current mirror technology instead of cascode, so far as I can guess, and does equally well, or better.
When it comes to power amps, open loop is tough, because the MEASURED DISTORTION is very high, actually embarrassingly high at high power with low Z loading. My speaker load is from 0.5-8 ohms for my Wilson Audio speakers, so open loop is really, really, tough to do comfortably. Even my feedback designs, such as the Parasound A21 is not up to it for my personal use. I just hate to see higher order harmonics AT ALL on the FFT print out. Charles lives with much more distortion than I like to do, but he often beats me in the LISTENING tests. That is the trade-off.
Of course, the CTC Blowtorch preamplifier, the reason for this thread, uses complementary differential folded cascode for both input and output, and nothing more, with no global feedback, and it is the ONLY piece of audio equipment that I would not exchange for anything else.
I do think that mosfets on the input are too problematic, now the Linear Systems is FINALLY making enough p channel jfets to even sell to amateurs.
From a purely topological point of view, a complementary differential folded cascode design is the best I have found. Charles has moved on to proprietary current mirror technology instead of cascode, so far as I can guess, and does equally well, or better.
When it comes to power amps, open loop is tough, because the MEASURED DISTORTION is very high, actually embarrassingly high at high power with low Z loading. My speaker load is from 0.5-8 ohms for my Wilson Audio speakers, so open loop is really, really, tough to do comfortably. Even my feedback designs, such as the Parasound A21 is not up to it for my personal use. I just hate to see higher order harmonics AT ALL on the FFT print out. Charles lives with much more distortion than I like to do, but he often beats me in the LISTENING tests. That is the trade-off.
Of course, the CTC Blowtorch preamplifier, the reason for this thread, uses complementary differential folded cascode for both input and output, and nothing more, with no global feedback, and it is the ONLY piece of audio equipment that I would not exchange for anything else.
I do think that mosfets on the input are too problematic, now the Linear Systems is FINALLY making enough p channel jfets to even sell to amateurs.
Yes Jay, but if one has a transparent system that can handle very well one classical recording with lots of very fine detail and ambiance etc, it is not fair to assume when one plays another classical recording where the fine detail and ambiance and been degraded into a mushy, rough, burbling mess that it's probably not the system that is to blame ?
Last edited:
I think my first step towards a balanced & symmetrical design will be a balanced version of my single ended amps. They are a 3 stage design but at present they only need 15pF miller caps for good stabilty so I think they are fast enough to put in a good performance.
I don't just want to jump to the "ideal" design immediately, I also want to understand what changes bring what sonic benefits in the way that John, Charles and some others here probably did quite some time ago.
Just enjoying the catching up process
I don't just want to jump to the "ideal" design immediately, I also want to understand what changes bring what sonic benefits in the way that John, Charles and some others here probably did quite some time ago.
Just enjoying the catching up process
I did everything I could to reduce higher order distortion to minimum. I don't think I'm hearing an edge.
I hate that edgy sound but normally this amp makes me smile - except with a few bad recording
p.s. - don't like veils either
Simple logic is like this:
1) Single driver will not give top level transparency, so we use crossover technology, which gives evils as compromise. But every ones system has different amount of evil.
2) Rigid cone is necessary for top level transparency, or ideal pistonic behavior, but it gives compromise at many areas.
3) You want a flat response at crossover frequency? You can opt for LR4, but watch out for 360 deg phase lag. You want less phase lag? LR2 will do it but 3dB bump will not give accurate sound, and so absolute phase issue. You want perfect transient? You may get it but compromise is at off axis listening.
Many things can be said. Thinking that one's system is perfect, is blind.
Cascode and current mirror is a sure way to lower THD, and surprisingly transparency is commonly attributed to low THD amplifiers.
I have never liked cascodes and current mirrors. But there is this unique current mirror (I don't know if it is still current mirror). The input to the next stage is taken from the supply side of the "current mirror transistor". This sounds good, and I think Ayre uses this too.
OK, so you say, the system is not transparent. This is good to know, because simple logic says that if the system is transparent, it cannot sound bad UNLESS the source recourding is bad.
Now we are back to the usual program: which parameters of the system determine the 'badness' - or, the deviation from transparency.
All the points you mention above change the sound in one way or another - frequency response, speaker directivity, distortion, amplifier damping factor, etc. Yet, none of these things can turn a bad recording, with no soundstage, lots of compression, bad miking, into a good recording.
But I agree that a bad system can turn a good recording with nice soundstage, lots of dynamic range, good voice/instrument separation, good 'rythm' etc in a bad listening experience.
Also, a 'bad' sytem can make the difference between a good and a bad recording much smaller than a good system would show.
That is the reason that when you hear little difference between a good and bad recording, you know something is not right with your system. In fact, I keep a few bad CDs for this - if they sound pretty much the same as my best CDs, I've done something seriously wrong!
jan
- Status
- Not open for further replies.