AB(x) testing of Power Amplifiers

[Plcamp]

Interesting.

I would like to see the coincident voltage and current waveforms of the two amps responding to a low frequency square wave signal in time domain. (When I designed switching power supplies in the 80’s this nature of test was quite revealing).

I am guessing such a test would show significant differences.

 

[pma]

The measurement was not straightforward, because both output terminals of the class D amp are floating and the shunt (0R22) voltage of small value is thus floating with high CMV. I will probably try transformer isolation later, this was a "standard" balanced input, however the resolution was limited due to input divider scaling with respect to CMV.

I repeated the measurements of speaker current and speaker current distortion, with a setup with higher resolution and lower noise floor - the shunt + signal transformer to get rid off the CMV. The measurements of class AB and class D are almost identical re speaker current and distortion and no new insight was brought. The audible difference seems to be almost certainly because of the difference in voltage frequency responses.

Is this AB commercially available or is it your own design?

Sorry it is not commercially available, I have not been selling audio components since 2012.

 

[Blumlein 88]

85 percent of hifi is frequency response......again!

 

[boXem]

To finish your demonstration, and show that there is no magic behind the treble boost, I would suggest plotting the voltage at speaker terminals for both class D out of loop filter and class AB. From this an impedance curve could also be calculated.

 

[boXem]

85 percent of hifi is frequency response......again!

What is funny here is that the inaccuracy of one element (the class D with out of loop filter ) improves the accuracy of the system.

 

[pma]

To finish your demonstration, and show that there is no magic behind the treble boost, I would suggest plotting the voltage at speaker terminals for both class D out of loop filter and class AB. From this an impedance curve could also be calculated.

You are right and it was already done in post #10, first plot
https://www.audiosciencereview.com/...testing-of-power-amplifiers.20464/post-678583
(just shift the class AB plot 0.2 dB upwards in your mind)

What is funny here is that the inaccuracy of one element (the class D with out of loop filter ) improves the accuracy of the system.

Yes, it compensates for the HF roll off due to the 6m speaker cable inductance. This is one of many reasons why I insist on testing of the complete audio chain.

 

[JSmith]

*watching thread*





JSmith

 

[boXem]

You are right and it was already done in post #10, first plot
https://www.audiosciencereview.com/...testing-of-power-amplifiers.20464/post-678583
(just shift the class AB plot 0.2 dB upwards in your mind)

My bad, I understood it was the speaker FR response.

Yes, it compensates for the HF roll off due to the 6m speaker cable inductance. This is one of many reasons why I insist on testing of the complete audio chain.

Are you sure this is the cable inductance and not the speaker inductance?
The plot of the speaker output with the two amplifiers would achieve the demonstration. Unless I missed it - again - .

 

[Blumlein 88]

What is funny here is that the inaccuracy of one element (the class D with out of loop filter ) improves the accuracy of the system.

Same thing happens on my ESL speakers.

 

[pma]

Wiped off the dust from my old A/B relay switch box to be able to make the tests in stereo ....

I was able to match the level by 0.02dB at 1kHz

 

[Spkrdctr]

What or who is controlling the switch box? I am wondering how blind the test was and how many times you guessed correctly on the amps. Thanks!

I'm editing this post. I think I missed that it was an A/B test only? No blind testing. I messed up my reading of your posts.

 

[pma]

What or who is controlling the switch box? I am wondering how blind the test was and how many times you guessed correctly on the amps. Thanks!

I'm editing this post. I think I missed that it was an A/B test only? No blind testing. I messed up my reading of your posts.

Your question is correct. It may be either A/B or ABX, because the operation is by long remote control cable. This time (yesterday) it was A/B. Several times before it was ABX, with another person doing switching and the record of switch positions sequence was made and compared to tested person guessing record. I plan to add a random "coin-flip" circuit driven from PC to the control cable in future.

In addition, similar ABX test with recorded files was made earlier and posted here in ASR

Class D AIYIMA A07 amplifier (ABX) listening test

I was curious if inserting AIYIMA A07 TPA3255 class D amplifier in the D/A – A/D loop would be audible. I tested the amp in another thread and found that the frequency response depends, at higher frequencies, on speaker complex impedance, due to amplifier's output LC filter - that is outside...

www.audiosciencereview.com

another ABX test with recorded files from 2 class AB amplifiers posted

Listening test of 2 power amplifiers - files recorded for download - disclosed

There are requests to record outputs of two different power amplifiers and prepare a listening test with such recorded files. I have decided to answer the challenge again and recorded files from 2 different class AB amplifiers, loaded with real speakers. One of the amplifiers has about 6 times...

www.audiosciencereview.com

 

[UsableThought]

If this thread is still alive . . . pma, I'm currently in the middle of making some simple A/B switching gear for a stereo amplifier comparison of my own - on a much smaller scale, just a cheap class D ($40 or so Amazon) versus a tiny desktop amp, the class A/B AudioEngine N22. I am also going to compare both these guys to my elderly but still pleasant-sounding Kenwood KR-4070. I won't be doing any waveform analysis or scoping or whatever - just using my aging, highly unreliable ears. I guess you would say it's a DIY project for fun, done at a crude level.

So far I have just built a mechanical switch for the amp/speaker connections. Works fine if I sit close enough to the amp setup - puts me right in the magic triangle at moderate to low volumes. I may go to relays just because it would expand my seating & listening options.

However what I am including, but what I don't seem to see in your setups, is switching over the input source to the amplifiers at the same time that I switch the speakers. I want the same exact input in other words - whether it be my iPad with its built-in DAC or my Mac M1 Mini connected to either a Schiit Modi 3 DAC or a Focusrite Scarlett 2i4 DAC. Thus I have an additional switch that has stereo RCA input jacks & swaps a pair of RCA stereo outputs.

The other option I can think of is to not bother with a switch for changing the inputs, but simply manually changing the input. This is easy to do given that aside from simply going straight from a computer or tablet 1/8" plug output, there are RCA-to-1/8" cables to make it quick and easy even with a snobby DAC like the Schiit which offers only RCA outs.

So anyway I don't see anything like an input switch in your design. Are you doing the input switch manually, which would suggest that you listen to fairly lengthy segments of music in your A/B'ing, rather than doing quick switching from one to the other? I don't trust my own memory to handle lengthy listening before I switch over - too likely I will fool myself. Or do you have an input switch somewhere & it's blindingly obvious and I'm just missing it?? I'm no sort of audiophile, just a DIY middlebrow, and quite often I do in fact miss the obvious. So illuminate me if you would - thanks.

 

[chrispdx]

If this thread is still alive . . . pma, I'm currently in the middle of making some simple A/B switching gear for a stereo amplifier comparison of my own - on a much smaller scale, just a cheap class D ($40 or so Amazon) versus a tiny desktop amp, the class A/B AudioEngine N22. I am also going to compare both these guys to my elderly but still pleasant-sounding Kenwood KR-4070. I won't be doing any waveform analysis or scoping or whatever - just using my aging, highly unreliable ears. I guess you would say it's a DIY project for fun, done at a crude level.

So far I have just built a mechanical switch for the amp/speaker connections. Works fine if I sit close enough to the amp setup - puts me right in the magic triangle at moderate to low volumes. I may go to relays just because it would expand my seating & listening options.

However what I am including, but what I don't seem to see in your setups, is switching over the input source to the amplifiers at the same time that I switch the speakers. I want the same exact input in other words - whether it be my iPad with its built-in DAC or my Mac M1 Mini connected to either a Schiit Modi 3 DAC or a Focusrite Scarlett 2i4 DAC. Thus I have an additional switch that has stereo RCA input jacks & swaps a pair of RCA stereo outputs.

The other option I can think of is to not bother with a switch for changing the inputs, but simply manually changing the input. This is easy to do given that aside from simply going straight from a computer or tablet 1/8" plug output, there are RCA-to-1/8" cables to make it quick and easy even with a snobby DAC like the Schiit which offers only RCA outs.

So anyway I don't see anything like an input switch in your design. Are you doing the input switch manually, which would suggest that you listen to fairly lengthy segments of music in your A/B'ing, rather than doing quick switching from one to the other? I don't trust my own memory to handle lengthy listening before I switch over - too likely I will fool myself. Or do you have an input switch somewhere & it's blindingly obvious and I'm just missing it?? I'm no sort of audiophile, just a DIY middlebrow, and quite often I do in fact miss the obvious. So illuminate me if you would - thanks.

I'd be really interested to hear the results of your evaluation of the cheap class D vs the Audioengine N22.

 

[UsableThought]

I'd be really interested to hear the results of your evaluation of the cheap class D vs the Audioengine N22.

Well . . . my testing was pretty quick & dirty and I didn't do a lot of it. I don't know how people manage it, but if I were to try it again, I would want to loop a particular musical passage, keep it looping, and have someone else do the switching so I could listen blind. I found that even the minimal attention needed to manipulate the switch back & forth got in the way of putting my focus on listening. And my listening memory just cannot hold what a passage on amp A sounded like versus what it sounds like when I switch to amp B.

So if I were really going to spend time on this, I would spend more time reading & learning on comparison theory & practice, and zero time thinking about the hardware (e.g. switches) until I knew what kind of comparison setup might actually result in success - vs. just fooling myself or else being totally uncertain.

Anyway, using a fairly crude switch to flip between 2 amps, I did come away one impression of the N22 vs. the class D, which I will pass on for what it's worth. First of all the class D is a $45 "Kinter K2020A+ Limited Edition Original Tripath TA2020-020 Class-T Hi-Fi Audio Mini Amplifier with 12V 5A Power Supply," as purchased on Amazon - product page is here. I ran into into some decidedly ordinary 3-way speakers which I had rescued from the dump & upgraded with new tweeters & mids, plus improved crossovers. But still not what you would call audiophile grade speakers.

My impression? That for most music, the two amps sounded the same to me. The only passage I could find where I thought the N22 did better was with one of the Beethoven late sonatas, the "Hammerklavier." Since it's a Beethoven piano piece, it has a lot of sudden jumps from very quiet chords to loud chords. And my impression was that the Kinter class D amp was slightly harsher in the mid-to-high end, whereas the N22 sounded smoother & "more musical" (less harsh). I didn't bother doing any waveform analysis etc. etc. - I wouldn't know what to look for, nor how to connect something on a graph with subjective listening impressions. Again I really don't know enough about what constitutes useful comparison vs. just playing around.

I also compared the N22 to my old Sherwood KR-4070 when playing my digital piano keyboard through Pianoteq software and thence into a DAC and thence into the amp - for DACs I sometimes use a Schiit Modi 3+, other times a Focusrite Scarlett 2i4; I can't tell the difference between those two. For my money, the N22 does just as well as the Kenwood, despite being a much lower-watt amp. In fact I was surprised that even though I have to turn the volume knob up on the N22 nearly all the way to max - say 4/5th of the way - I don't hear any unpleasant distortion whatsoever.

So that's it for my very minimal testing. Pic below of the 2 boxes I built to switch 2 amps between the same DAC and speaker set.

 

[Capitol C]

I think I have enough data to explain audible differences.

1) FR at speaker terminals
View attachment 113524
Interestingly enough the FR elevation of the class D amp (LC output filter) compensates for decay of highest frequencies due to inductance of speaker cable.

2) distortion at speaker terminals (behind the speaker cable)

View attachment 113525
class AB distortion at speaker terminals, 2V. THD = 0.0062% is at 400Hz.


View attachment 113526
class D distortion at speaker teminals, 2V. THD = 0.0034% is at 400Hz.

Conclusion

As always, we MUST measure the whole audio chain and not only a single component of the chain.

The frequency response must be it, it is pretty bad for the class AB. Still curious if it was actually double blind, or was it really just an AB test?

 

[escksu]

I am not surprised by your results. Amps dont sound the same. Some are designed to to have certain characteristic which is their "house sound". Some likes it, some dont. Its up to individual preference.

 

[SIY]

Some are designed to to have certain characteristic which is their "house sound". Some likes it, some dont. Its up to individual preference.

Name one, specifically.

 

[restorer-john]

Or do you have an input switch somewhere & it's blindingly obvious and I'm just missing it??

Feed the signal into both amps (splitter) and just switch the outputs.

 

[escksu]

Marantz (esp. KI signature ones) has it own house sound. Krell has it. Mcintosh has it too.

Marantz sound is by design. Its not neutral. Its characteristics was created on purpose. Good or bad, i am not here to discuss about it.

Krell is again known for its bass and on the cold side (esp. for early ones). Its not neutral as well. Only newer gears are on the neutral side.