Audio Science Review = "The better the measurement, the better the sound" philosophy

ASR hasn't measured an AGD amp (probably too expensive). 

@fleschler , if AGD provided the amplifier on loan, I expect they would measure it.

We should all expand our horizons and got out for Taro ice cream. My local Walmart of all places has it.

I shall look for it tomorrow! Talking about the mochi or the standard fare?

@thecarpathian 

I had to look that up. I have never had Mochi stuffed with ice cream.  This is standard, in a tub fare in the freezer section. I was trying to find it on their website, but it is not there.

 

https://www.biline.ca/audio_critic/mags/The_Audio_Critic_20_r.pdf (page 16 on)

Italics mark content contributed by crymeanaudioriver

A lot of trash in this collection of articles unfortunately.

I see it differently. Sure, we learned quite a bit since these articles were written. Yet, the time during which they were published marked presence of many amplifiers designers and manufacturers in North America. They knew what they were talking about.

Interesting that you linked to the specific article on the power cube. They say in that article,

We do not perform such separ-
ate IM distortion tests here because they characterize the
same nonlinearities identified by the THD tests. A non-
linearity that gives rise to a high 20 kHz THD will also
cause inband distortion products in a multitone test. A
full-scale 20 kHz test has the advantage that it has the
maximum rate of change of any inband test signal and it
characterizes both even- and odd-order nonlinearities
[Borbely 1989], [Jung 1979]. Transient intermodulation
effects [Otala 1970] are also covered in this test.

I assume you agree with all that written above as well, or only what suits you?

A typical “moving the goalposts” / “straw-man” rhetorical manipulation. That discussion of IM correlation with THD is still conducted in the context of the amp’s linearity under best conditions, not in the context of the linearity of system comprised of amp+speaker.

The article you link talks, often, about the performance of the current limiter inside the amplifier. If the amplifier is running into a current limit, it is clipping. If you are running your amplifier into clipping, then you are beyond the limits of the amplifier.

Current limiting occurs in amps without specific current limiting circuits too. As an example, a heating up power transformer coil may in effect serve as a current limiter. Another example is insufficiently sized capacitive power bank.

Look at the curves of THD vs output power characteristics of amplifiers, and you’ll see that typically, there is a rise in THD (and by extension in IM) long before the amp clips. The degree of such deterioration is typically frequency-dependent too.

This is indeed one of the mechanisms explaining the phenomenon of some of the amps distorting significantly more while they are connected to a speaker compared to when they are connected to a dummy resistive load.

Note the only example they show of oscillation, the issue yielded by non-resistive tests, shows oscillation occurring at 2 ohms, 60 degrees, and 1 ohm 30 and 60 degrees. This is important as it relates back to this article on ASR you linked:

Yes, occurrences when a commercially sold amplifier becomes unstable and turns into a generator while connected to a specific speaker are rare. Even though, the thread referred below has a description of a surprisingly common-case instance of that.

However, just like with the discussion of THD and IM, we need to take into account that the amp-speaker system can “ring” for some time, instead of turning into a self-supporting generator. Some of the replies in the thread below describe precisely such occurrences.

Complex Load for Power Amplifier torture testing

This specific issue is discussed, as they talk about how many speakers have both very low impedance and very large phase shift. The conclusion is very few.

Yes, this is one of the aspects discussed there.

The “conclusion”, if we were to derive any, is that in certain segments of the world market, predominantly preferring smaller speakers with simpler crossover topologies, there are indeed fewer opportunities for a given amp to exhibit ringing, and even fewer opportunities to exhibit runaway self-generation.

However, some of the replies highlight the fact that in some other  market segments, including that of affluent audiophiles, larger speakers employing exotic transducers and much more sophisticated crossovers are more prevalent, and thus the events of ringing and self-generation are much more probable.

Hence why the consensus that resistive testing into low enough impedance is sufficient.

There is no such consensus in that discussion. Interested readers can go there and see for themselves. I would roughly split the multitude of members posted there onto three categories:

1. Designers and restorers of amps from Western countries. They are for comprehensive testing with non-purely-resistive loads.
2. Designers of amps from China. They are for limited testing with non-purely-resistive loads.
3. Vendors selling amps made from pre-built blocks, Amir, and some of Amir’s followers. They maintain that testing on purely-resistive load is not ideal, yet good enough for predicting amp’s performance in 99.99% of cases.

Elsewhere there is a call to include 2 ohm testing which I think I have seen on some more recent tests.  It is probably important to identify from the articles linked that the worst issues are with tube amplifiers, lauded by audiophiles and rarely tested by ASR. When they are, the result is not positive.

Behavior of most amplifiers, including tube ones, does depend on the value of purely-resistive load, yet the change in behavior is much more predictable with the change in just the resistance value.

Thus, while testing on 2 ohm has its merit, it appears from the discussion that testing on non-purely-resistive loads is of more interest to people with practical experience in designing and repairing amps.

The other threads you posted from ASR are primarily not technical discussions about testing, but more banter from what appears to be the less technical members. Not everyone on ASR is technical.

Unless a member is an unscrupulous dealer pushing some version of snake oil, most of the “banter” deserves consideration, in my opinion. I give full credit to Amir for filtering out vast majority of such snake oil salesmen. However, the story doesn’t end there.

Similarly to doctors giving zero credence to patients describing their symptoms, and only relying on the results of locally available objective tests, amp designers and restorers only relying on simple measurements aren’t likely to keep their clientele for long.

Also, quite a few replies there were redacted: one can see quotations from them and references to them, but not the original replies in their entirety.

If you are going to participate in a thread putting down a web site you should probably learn how that site works, or at least the "Click to expand" button.

Ignoring your not so subtle attempt of slighting me, a person whom you obviously don’t know much about.

There is nothing redacted.

Once again, interested readers can go there and see for themselves. They’ll find the traces of reductions exactly as I described them.

The forum has a very good quote and reply system unlike another one I am thinking of.

Indeed, ASR runs on a more modern discussion platform than Audiogon. The “reply system” in practice also includes so-called moderation subsystem, or, in simpler terms, censorship features.

A discussion site without actively working moderation quickly devolves into pointless incomprehensible mess, mostly frequented by spammers.

Yet a discussion site with overreaching moderation generates its share of issues, both for regular members and site owners. I maintain that the ASR moderation has been such since about 2021.

Here we differ too. As technical as the dedicated ASR discussion thread was, it didn't touch on stochastic behavior of non-linear time-dependent systems, of which a practical multi-transducer loudspeaker is a prime example.

Attaching such a system to an approximately linear, approximately time-independent power amplifier leaves the combined system still non-linear and time-dependent.

The math describing non-linear time-dependent systems is far more sophisticated than the one underlying the simple measurements that Amir uses.

At first, that appears to be a lot to unpack. However, it can quickly be taken as a deflection. The topic at hand is the test of amplifiers. Specifically in this case resistive testing.

Nope, topic at hand is relevance of the testing Amir does on specifically power amplifiers to the subjective perception of audible distortions contributed by amp A vs amp B when connected to a specific speaker.

Note that I don’t discuss Amir’s testing of DACs, which I personally find entirely satisfactory and providing huge value to the community.

Neither I discuss Amir’s testing of loudspeakers, which Amir positions as in effect partial, only providing about 70% of information needed for a purchasing decision.

My position, as is the position of majority of ASR members with practical experience in designing and repairing power amplifiers who cared to express their opinions, is that the testing Amir has been conducting is marketed as more definitive than it shall be based on scientific understanding of the limited nature of the tests.

Attaching such a system to an approximately linear, approximately time-independent power amplifier

This negates all the other words used in the last paragraph. By your admission, the amplifier is time independent, approximately at least. That a speaker is not, is not relevant to the discussion.

OK, let me give you another analogy. Imagine if we assigned championship titles in boxing based on tests involving a boxer and a punching bag.

Boxer would be performing prescribed sequence of moves of varied amplitude and frequency, and we would be measuring, let’s say, acceleration of the bag’s center of mass in relation to the ideal acceleration expected at the execution of a specific move.

Case (A). A very heavy bag - let’s say weighting 8 times the average human weight - would approximate a linear time-invariant system pretty well. That’s an analog of 8 Ohm purely resistive load. Whatever the boxer does would translate pretty unambiguously into the bag’s acceleration.

Case (B). A lighter bag, let’s say ½ of the average human weight, would sway under the punches, thus failing to remain a linear time-invariant system from the boxer’s point of view, yet it would be still somewhat predictable. That’s an analog of ½ Ohm purely resistive load.

The geometry of the bag and the rope it is suspended on would now influence the system dynamics stronger. An audiophile analogy would be thermally induced deviations of the load resistance value, and parasitic capacitance of the cable leading to the resistive load.

Case (C). Now imagine that instead of the single heavy bag, the boxer would be punching a system of four bags: one of them, still somewhat heavy, used for punching, and the other three with differing weights interconnected with the first bags and between each other with a system of springs, ropes, and pulleys. In addition, the ropes would be threaded through a system of friction pads, with friction coefficients strongly dependent on the pads temperatures.

The goal of the boxer remains the same: the center of the mass of the bag he is punching must exhibit a specific pattern of acceleration. Only now the bag is also pushed and pulled by other swinging bags via the springs and the ropes threaded through friction pads and pulleys. This is analogous to how an amp must work when attached to a practical speaker.

Obviously, the task of boxer in case (C) is more complicated that in case (A). And the champion of case (A) won’t necessarily be the champion of case (C). The testing of power amps Amir is doing is analogous to case (A). What audiophiles are interested to know is analogous to case (C).

The only relevance would be if speakers drifted from a maximum of 30 degrees phase shift to 60 degree when they got hot and this is not identified in the ASR discussion linked. Is that what you are claiming? 

Indeed, thermal drift of a transducer coil resistance value due, to ,say, a loud music passage, is a factor that a good amplifier must somehow compensate for. Yet even if we remove the friction pads in the system of bags described above, its behavior will remain pretty sophisticated, and very different from a behavior of a single heavy punching bag.

While being a crude analogy, the visuals of the system of bags give insights of where an amp testing close to perfect on a 8 Ohm purely resistive load could fail miserably on auditioning involving a specific sufficiently large and sophisticated loudspeaker.

One of the cases is simply running our of amp’s power supply current capacity. A music passage may be such that at some point all the bags will be moving towards the boxer, overwhelming him with the combined impulse.

Such deficient behavior is usually exhibited by amplifiers of all classes with power supply sized insufficiently relative to the speaker and music characteristics.

Audibly, such deficiency usually manifests itself as “lack of dynamics”. Once again, it may be not outright clipping, but rather increased distortions as the amplifier approaches its power limit.

Another case is quick stochastic oscillations of the bags,  excited by a complex music passage. It can overwhelm ability of the boxer to punch quickly enough to counteract the resulting irregular oscillations of the bag he is punching.

This type of deficient behavior may be exhibited on some music passages by certain class A, A/B, and especially class D amplifiers, with their open loop bandwidth insufficient to deal with such combination of the speaker and music passage.

Audibly, such deficiency may manifest itself as a lack of transparency, and timing errors, especially in music produced by dozens of instruments playing at once.