Pass ACA Class A Power Amplifier Review

[solderdude]

This is all very measurable. There is no mystery to it and still does not explain how amps could measure the same but sound different.
They can only do so when they do not measure the same.
Crossover distortion is measurable and if there is an audible difference there is a measurable one.

NO one says all amps sound/measure the same with all loads. There is no 'missing' signal nor things 'we don't know how to measure yet' when it comes to electric signals.

 

[SIY]

Check at high and low power levels, high and low test frequencies, and importantly, stone cold and up to 85C. I've not found an amp which cannot be provoked into misbehavior in one way or another. Does it matter in listening? Probably not, but that's why we do research and try to stretch the limits.

Typically, I let amps warm up for 10 minutes or so, since any issues when cold are transient (so to speak), and of course, setting quiescent current cold is a recipe for failure. But I haven't seen any issues with power level or frequency once the quiescent current is set properly- could you give me some for-instances? Assume I have a well-equipped test bench.

BTW, my apologies for an erroneous statement: the Cambridge is NOT cheaper than the equivalent ATI, the ATI is cheaper (though neither is cheap, at $2000 and $3000). The website confused me a bit because of the number of channel options.

When you're done this new design, I'd love to have a chance to run it through the ringer for an article, hint, hint.

 

[MakeMineVinyl]

When you're done this new design, I'd love to have a chance to run it through the ringer for an article, hint, hint.

I don't have any power in that department; you'd need to contact ATI directly. In any case, the amp series is not out yet, and the coronavirus thing tossed a spanner into the works. I will certainly talk about what's going on when they're out.

 

[SIY]

Gracias!

 

[MakeMineVinyl]

This is all very measurable. There is no mystery to it and still does not explain how amps could measure the same but sound different.
They can only do so when they do not measure the same.
Crossover distortion is measurable and if there is an audible difference there is a measurable one.

NO one says all amps sound/measure the same with all loads. There is no 'missing' signal nor things 'we don't know how to measure yet' when it comes to electric signals.

I didn't mean to imply that there was some mysterious "X" factor which can be heard but not measured. I conducted extensive listening tests (single blind) at various configurations and differing amounts of crossover distortion were consistently audible - and there was consistently measurable and observable differences on a 'scope.

The listening tests were using music of course; I'm not aware of any measurements which specifically address music of different types, different instrumentation, etc, etc. We can "simulate" music with test signals, but actually coming up with a concrete number of merit with actual music in all its variety as a test signal is pretty tough - at least I've never seen one. One obvious way to "measure" with any signal source is subtraction (nulling). This has its own can of worms, and what number could be ascribed to the result? Maybe DSP will come to the rescue at some point, and maybe Audio Precision is doing just that.

 

[pkane]

I didn't mean to imply that there was some mysterious "X" factor which can be heard but not measured. I conducted extensive listening tests (single blind) at various configurations and differing amounts of crossover distortion were consistently audible - and there was consistently measurable and observable differences on a 'scope.

The listening tests were using music of course; I'm not aware of any measurements which specifically address music of different types, different instrumentation, etc, etc. We can "simulate" music with test signals, but actually coming up with a concrete number of merit with actual music in all its variety as a test signal is pretty tough - at least I've never seen one. One obvious way to "measure" with any signal source is subtraction (nulling). This has its own can of worms, and what number could be ascribed to the result? Maybe DSP will come to the rescue at some point, and maybe Audio Precision is doing just that.

I’ve not looked at simulating crossover distortion, but it should be possible to do this in software for a simple distortion case. I’ll do some research and see if it can be added to DISTORT software.

 

[Speedskater]

Actually cables with high capacitance like the Kimbers can lead to attenuation of the hights and even oscillations/damage if the amp has no zobel network

Can cause ringing or even oscillation, yes
can lead to attenuation of the highs, no

 

[MakeMineVinyl]

I’ve not looked at simulating crossover distortion, but it should be possible to do this in software for a simple distortion case. I’ll do some research and see if it can be added to DISTORT software.

I don't know how simple that would be since there are so many variables and topologies of output stages; it could be very complex I would think for anything other than a handful of scenarios. Programming for a discontinuity at zero crossing would be easy I would think. Even programming variable discontinuity would probably be relatively simple with variable smoothing. That might be enough for a VST plug-in for a DAW. I don't know though what use it might be since crossover distortion is pretty nasty sounding and nothing like a tube simulator since tube class A/B amps are awash in bias current compared to transistor designs.

 

[pkane]

I don't know how simple that would be since there are so many variables and topologies of output stages; it could be very complex I would think for anything other than a handful of scenarios. Programming for a discontinuity at zero crossing would be easy I would think. Even programming variable discontinuity would probably be relatively simple with variable smoothing. That might be enough for a VST plug-in for a DAW. I don't know though what use it might be since crossover distortion is pretty nasty sounding and nothing like a tube simulator since tube class A/B amps are awash in bias current compared to transistor designs.

Right. It looks fairly easy to code for an adjustable amount of cross-over distortion in the transfer function. DISTORT is an app that applies different types of distortion to an audio file of one's choice to let one test the audibility of these distortions. Similar to a plug-in for a DAW, but meant to reproduce common distortions and the effects of various DSP tools, such as non-linearities, jitter, various dither types, etc.

 

[amirm]

Check at high and low power levels, high and low test frequencies, and importantly, stone cold and up to 85C. I've not found an amp which cannot be provoked into misbehavior in one way or another. Does it matter in listening? Probably not, but that's why we do research and try to stretch the limits.

I ran warm up tests for a while on amplifier but did not show anything significant. Here is an example:

I could not provoke misbehavior and eventually got bored and stopped running this test. Thermal feedback better be part of the design criteria.

 

[MakeMineVinyl]

And of course negative feedback is trying to eliminate all nonlinearities within its available loop gain, including crossover distortion.

 

[MakeMineVinyl]

I ran warm up tests for a while on amplifier but did not show anything significant. Here is an example:

I could not provoke misbehavior and eventually got bored and stopped running this test. Thermal feedback better be part of the design criteria.

Maybe its just my luck - or lack of luck - but if something can misbehave, it will do so in my presence. As an amp with a conventional bias circuit heats up, the tendency is for the bias to increase. ThermalTrak trasistors are the usual remedy. From the factory, mainstream amps (not exotic high end ones with massive near class A bias) are set on the leaner side for long term stability and for limiting heat. Out in the world, these amps get stuffed in racks with no ventilation among other abuses. Unnecessary heating is a liability since its impossible to know what conditions an amp is going to find.

 

[amirm]

And of course negative feedback is trying to eliminate all nonlinearities within its available loop gain, including crossover distortion.

Which circles us all the way back to ask why we need the ACA amplifier without:

It is incredible that when it comes to audio, the dirtier the dish is in the restaurant, the more the customer thinks the food tastes better!

We have powerful tools like feedback to bring transparency to sound reproduction. As if being against mainstream is a fashion statement, folks walk around proud that they are ignoring measurements....

 

[MakeMineVinyl]

Which circles us all the way back to ask why we need the ACA amplifier without:

It is incredible that when it comes to audio, the dirtier the dish is in the restaurant, the more the customer thinks the food tastes better!

We have powerful tools like feedback to bring transparency to sound reproduction. As if being against mainstream is a fashion statement, folks walk around proud that they are ignoring measurements....

That amp didn't have a lot of gain (gain stages) from what I gather, so feedback wouldn't have much to work with. Why this amp is so dirty, I can't imagine. but I wouldn't want it.

 

[Sonny1]

"It also had several flaws, one of which did result in a trip to the local emergency room with breathing problems similar to those sometimes experienced by arc welders and caused by extended exposure to ozone."

Wow. That is epic. Noting says high-end audio like having stators running at 10 kV emitting ozone.

Tom

That’s what I call committee to! They say those speakers had sound “to did for”. Or was it “to die from”?

 

[anmpr1]

1) I conducted extensive listening tests (single blind) at various configurations and differing amounts of crossover distortion were consistently audible - and there was consistently measurable and observable differences on a 'scope.

2) The listening tests were using music of course...

3) One obvious way to "measure" with any signal source is subtraction (nulling).

1) What level was distortion 'consistently audible'?

2) In reported tests (the earliest I know of off hand was mid '70s in Stereo Review--Bob Carver modified one of his Phase Linear amps to inject increasing levels of distortion, and a panel listened in order to decide when it became audible. Audibility was initially heard on test tones, and only at higher levels with actual music. The more complex the music the less audible the distortion. I think thresholds were a couple of three percent on music. Might have been four percent on complex stuff.

3) Again, Carver 'demonstrated' this technique to both Peter Aczel and the gang at Stereophile. Peter told me that his comparison (using the Mark Levinson ML-2 class A amp) was probably invalid anyway from an audibility standpoint, because prior to transfer function matching via nulling, the amps were likely indistinguishable if operated within their respective electrical parameters. At the time Aczel was not using matched levels to listen for differences. The Stereophile drama could have been a different thing in the audibility department inasmuch as Bob was nulling against a tube amp. So in that test there was a big difference in the amps from the get-go, due to design topology and intrinsic electrical performance.

Side note: John Saiu's 'white paper' on the amplifier/loudspeaker interface (amplifier damping factor-cable-speaker) is now on-line. John provides a handy Excel formula that one can use to more or less decide about their setup. He acknowledges that FR changes in or around 0.1 plus or minus dB are probably not going to matter. The question of course is whether one's amp can handle the sometimes difficult load variance asked of it.

 

[MakeMineVinyl]

1) What level was distortion 'consistently audible'?

We're talking about differences of 0.xx% verses 0.xxx% which varied across the audio band. Taking full power sweeps of THD vs frequency showed up differences at various biasing schemes consistently. This was correlated in listening tests.

These were research tests; I wasn't attempting to "prove" anything and didn't produce any graphs n' charts. It was purely to guide my development activities. During the listening tests, the subjects had no idea what I was changing, or if in fact anything was being changed at all - they had no idea what the focus of my research was. Sometimes I made changes, sometimes I didn't; I could have been playing video games at my computer for all they knew. But there was correlation. The fact that the listening data was confirmed by measurement data was what was guiding my decisions.

It must be understood that the original goal of this research wasn't to "improve" sound quality per se - it was to solve a practical problem having to do with how bias is handled in class A/B output stages. I found other things I could control, so I controlled them. I'm very pragmatic. I don't chase things which cannot be verified by objective measurements. If some sound change is perceived and I can't measure it, its a figment of my imagination. Any sound quality differences are not something which is going to be marketed, just the practical things we set out to solve. Sound quality things just went along for the ride.

 

[pjug]

I ran warm up tests for a while on amplifier but did not show anything significant. Here is an example:

I could not provoke misbehavior and eventually got bored and stopped running this test. Thermal feedback better be part of the design criteria.

This subject is interesting. As it relates to this discussion, it is a little unfortunate that the M2200 TDH+N is noise dominated so we can't really see if distortion has remained close to the -100dB that you measured. I know you can't test everything, but if you get another traditional Class AB amplifier with very low distortion, it might be interesting to take a few snapshots showing whether distortion changes as the amp warms up.

 

[MakeMineVinyl]

This subject is interesting. As it relates to this discussion, it is a little unfortunate that the M2200 TDH+N is noise dominated so we can't really see if distortion has remained close to the -100dB that you measured. I know you can't test everything, but if you get another traditional Class AB amplifier with very low distortion, it might be interesting to take a few snapshots showing whether distortion changes as the amp warms up.

The goal is to be able to test the amplifier against itself.

 

[pkane]

I don't know how simple that would be since there are so many variables and topologies of output stages; it could be very complex I would think for anything other than a handful of scenarios. Programming for a discontinuity at zero crossing would be easy I would think. Even programming variable discontinuity would probably be relatively simple with variable smoothing. That might be enough for a VST plug-in for a DAW. I don't know though what use it might be since crossover distortion is pretty nasty sounding and nothing like a tube simulator since tube class A/B amps are awash in bias current compared to transistor designs.

Here's my first approximation to simulating cross-over distortion. I decided to use two parameters, width and height of the cross-over region. Each of these defines a fraction of the -1 to 1 transfer function. As you said, the cross-over distortion is pretty nasty and generates a large number of odd harmonics at high levels. Here's an example using a simple 1kHz sinewave:

I zoomed in on the transfer function near zero to show the effect.

And here's what the waveform looks like with a higher level (10x) of crossover distortion to make it easier to see: