A lot of times people refer to "a wire with gain" so I thought it was a good idea. And, very important, one that we all can implement and use to calibrate and compare our ANA gear.
Re the TI amp - by just introducing a controlled distortion generator and add it to the TI output, it could measure like the Quad 😉
//
Re the TI amp - by just introducing a controlled distortion generator and add it to the TI output, it could measure like the Quad 😉
//
Yes. The definition was this piece of wire 1cm above a ground plane. Thats a 2-pole 🙂 !?
OK, check it.. it didn't say "plane" - for rev B 😉
//
OK, check it.. it didn't say "plane" - for rev B 😉
//
A change in FR of existing frequencies is a linear distortion. Same for phase shifts. Same for DA that is essentially equivalent to an RC ladder network. Phase shifts can be audible to the average ear if they exceed the threshold of audibility for group delay.
Production of new frequencies is a nonlinear distortion.
Regarding two pole, poles are places where a function blows up to infinity. A "two-port" is probably the correct term your wire with gain. https://en.wikipedia.org/wiki/Two-p...etwork has,has only one independent parameter.
Some poles:
Last edited:
But changed your mind...!?
Why, it really is contradictory to:
"For assessment and scoring, the amplifier under test (AUT) is treated as a black box except for the protection features scoring where the AUT specification is consulted."
//
Mine will be good.
Modified version of Valery's Stark mosfet amp. (2 stage comp, different fet drive )
80w Rms 0.005% or less
Unity gain stable
110 dB s/n at 1w
Slight 1% clip hold at 20khz at rail
2 stage frequency comp > 200v slew rate
25w idle bias set a bit high.
Dual differential / dual VAS mosfet outputs
Distortion Comes from 6922 tube preamp
Modified version of Valery's Stark mosfet amp. (2 stage comp, different fet drive )
80w Rms 0.005% or less
Unity gain stable
110 dB s/n at 1w
Slight 1% clip hold at 20khz at rail
2 stage frequency comp > 200v slew rate
25w idle bias set a bit high.
Dual differential / dual VAS mosfet outputs
Distortion Comes from 6922 tube preamp
Last edited:
If you measure all amplifiers the same you end up with only solid state high feedback amplifiers in the top ranks.
This is exactly the issue you see at the other site. Looking at the Topping L60, it’s clear their focus has been on gaming the measurement criteria. We know nothing about this amp other than it has low distortion and low noise (remote PSU).
That’s another reason why amplifiers scoring >70 get 5 stars. At >70, the designer/manufacturer is getting the fundamentals right
Last edited:
This is my view - different circuit topologies have different pros and cons, strong points and issues. They need deep investigation in specific measurements, depending on the amplifier circuit design, power supply design. There is no way to cover such options in a chart system based on score points. Specific investigation is needed for almost any amplifier. Just my view.
That’s why I’ve held off class D criteria for the time being. There is serious gaming of the measurements criteria in that part of the audio industry eg LPF, load drive capability etc.
What is really the issue with "solid state high feedback amplifiers in the top ranks"? Is there something wrong with these amps? Comments seem that of jealousy rather then proper facts...
I think that if the answer is: Yes, they forgot measurement/criteria xy and if that had been included, the amp have had much worse score - then I think that the xy measurement should be included in the suite rather then placing it in a special category.
It must be possible to define a set of measuremnts that weed out a proper amp irrespectively of technology i.e. that is "fair"... this is what is needed. The argument that only "solid state high feedback amps" "win" is just not valid.
If "LPF, load drive capability etc" is "gamed", measure for it and the game is over - this must be the only proper way forward - no?
Please find my comment on the revised document below.
//
Ok, good. What about black box? Do you believe limiting measurements and or tests to black box is sufficient for deep investigation?
BTW and as an aside, wanted to mention to you that sometime ago I stopped using foobar2000 completely. Found it was a source of distortion in my system, possibly depending on settings and or digital media properties, and or use of the ASIO driver component. There were times I heard something was wrong and one day started going through the system, device by device. Eventually tried removing foobar which fixed the problem. Don't know if it was Windows related, do know Window sometimes interferes with ASIO drivers supplied by Thesycon (which supplies pretty much all of them for Windows). Now use PlayPCMWin and WASAPI Exclusive Mode instead.
Last edited:
It is not enough during design phase and it is not enough if you want to find a root of the issue. It might be enough when restricted to a mere test result.
The way amplifier types are assessed was changed to treat technology classes separately because it was clear non-global feedback and tube amps would be penalised.
You can only really tell if a parameter is gamed if it sticks out in comparison to other measured aspects. If that is the case, you don’t reward it beyond a certain level. That’s the reason for example why distortion below 50 ppm is not rewarded with more than 5 points.
The quickest way to destroy any credibility with a tool like this is to tell non-feedback linear designers they are jealous because they can’t compete and their efforts are sub par. There are many fine tube amps that deliver excellent performance that will never meet the distortion or noise floor of a good linear amp with feedback. Nevertheless, these are very good amplifiers in their own right. Same for ZGFB amps where none other than Bob Cordell has pointed out the good performance of some of these within their chosen technology remit.
AAPAP should not be a weapon. Ideally it will be a useful tool that helps designers of any amp raise their art to higher levels.
🙂
Last edited:
What about some figure of merit for head room or dynamic range?
Or is this adequately defined by the max power available?
Maybe discussed already.... sorry I have not read the entire 12 page thread yet.
I mention it because of an observation made during prototyping a few years ago.
I was testing a design with music playing into speakers and a scope connected to the output.
I noticed during heavy bass passages -kick drums, bass guitar, etc. it was easy to drive the amp briefly into clipping.
All it took was a listening level of about 90db and a little EQ with the bass control.
I thought any clipping, even brief transients, were not acceptable, since they will produce transient harmonics right in the middle of the audio band.
At that point, my design had a PS with a 500VA transformer but the rail voltage was +/-30V, so the max power was roughly 40Wrms.
I purchased new transformers so I could get the rails up to +/-47V which eliminated the transient clipping at those listening levels.
Since then I've built several amps optimized for headroom by selecting PS rail voltages as close as possible to the circuit/chip limits and they do sound good.
Or is this adequately defined by the max power available?
Maybe discussed already.... sorry I have not read the entire 12 page thread yet.
I mention it because of an observation made during prototyping a few years ago.
I was testing a design with music playing into speakers and a scope connected to the output.
I noticed during heavy bass passages -kick drums, bass guitar, etc. it was easy to drive the amp briefly into clipping.
All it took was a listening level of about 90db and a little EQ with the bass control.
I thought any clipping, even brief transients, were not acceptable, since they will produce transient harmonics right in the middle of the audio band.
At that point, my design had a PS with a 500VA transformer but the rail voltage was +/-30V, so the max power was roughly 40Wrms.
I purchased new transformers so I could get the rails up to +/-47V which eliminated the transient clipping at those listening levels.
Since then I've built several amps optimized for headroom by selecting PS rail voltages as close as possible to the circuit/chip limits and they do sound good.
OK. But all these statement and tales are subjective.
For sure they are "fine" amps, but there are finer. If one can hear it or not is as I see completely irrelevant. What you have done now is to turn the assessment system into an objective, weighted one. I hope that you can see this. And that this fact is described in the assessment declaration.
And with this, its credibility is destroyed for me. I just want the least signal destructive amp for my system for a cost I'm willing to pay. With this approach is not possible/hard - why, because there will probably be a ZNF amp that got 65 points and an other that got 64, but the 64 is actually better measuring on at least a few parameters. Maybe 10 dB less distortion and less noise... Is this accepted by the inventors? Apparently yes.
//