Audio Research D300 Power Amplifier Review

[Feanor]

Reading this review for the first time I was struck by the fact that the D300 shows high levels low order harmonic distortion but almost a complete absence of higher order HD. Amir rightly points out that this is certainly due to the sparing use of negative feedback which, if ampily used, can reduce overall HD to negligible levels.

I'm not so sure, however, that Audio Research engineers are simply misguided or ill-informed. I don't believe that they are so naive as to not know that the D300 has rather high overall amounts of HD, nor that they don't know how to remedy that if they so choose. (Much less do I believe that ARC doesn't have test equipment or know how use it.)

To me it seems altogether more likely that the D300's HD profile is consistent with ARC's design goals for the amp. And, agree with the philosophy or not, the ARC's management and engineers purposely designed the D300 to have highish levels of low order distortion. This would stem from the conviction that most ARC users what to enjoy the euphonic effects of this sort of distortion.

True, of course, that high THD will result in output to the speaker that is less true to the recorded sound, but nevertheless it is what a many audiophiles want IMHO, (not saying I'm necessarily one of them).

 

[milosz]

Reading this review for the first time I was struck by the fact that the D300 shows high levels low order harmonic distortion but almost a complete absence of higher order HD. Amir rightly points out that this is certainly due to the sparing use of negative feedback which, if ampily used, can reduce overall HD to negligible levels.

I'm not so sure, however, that Audio Research engineers are simply misguided or ill-informed. I don't believe that they are so naive as to not know that the D300 has rather high overall amounts of HD, nor that they don't know how to remedy that if they so choose. (Much less do I believe that ARC doesn't have test equipment or know how use it.)

To me it seems altogether more likely that the D300's HD profile is consistent with ARC's design goals for the amp. And, agree with the philosophy or not, the ARC's management and engineers purposely designed the D300 to have highish levels of low order distortion. This would stem from the conviction that most ARC users what to enjoy the euphonic effects of this sort of distortion.

True, of course, that high THD will result in output to the speaker that is less true to the recorded sound, but nevertheless it is what a many audiophiles want IMHO, (not saying I'm necessarily one of them).

I think this is true of many modern tube amps in general, and also some solid state amps where designers have chosen to largely eschew negative feedback.

Many audiophiles believe that negative feedback is ALWAYS bad. After all- it's NEGATIVE, isn't it? (I'm sure that bias operates at some level on buyers, even if they know better.) This stems from early solid state designers who didn't understand transistors all that well and created amps there were quite nonlinear and then (mis-)applied MASSIVE amounts of negative feedback so that these amps would measure OK in reviews. The sonic problems that this misguided approach can engender were discussed in papers by Matti Otala; many believe that these hack designs were responsible for the notion that solid state amps sound harsh, screechy, etc. Making a poor amplifier then using massive NFB as a remedy might result in something that measures OK with steady test signals, but often creates an amp that does poorly with rapidly changing signals like music.

What Mr. Otala was saying was that NFB is not a remedy for poor basic design, but that judicious use of NFB along with an amplifier topology that is linear to begin with is the way to go. What many audiophiles read from this is that "NFB is always BAD." And that myth is perpetuated till this day.

Some audiophiles will listen to an amp with a certain "voicing" in terms of certain amount of euphonic low order harmonic distortion and say "See! This sounds RIGHT!" because that is how they want it to sound, they like that harmonic syrup- and they end up thinking and promoting the idea that an amp with proper NFB, which doesn't have this "tube sound," is just wrong and that NFB must be the culprit. "NFB must be bad, because the amps I like don't use it..."

 

[cistercian]

Some people seem to like a certain type and amount of distortion. I can remember arguments I heard years ago comparing
amps that were "sterile" to amps that sounded "pleasing". I know some designers took this to heart and built what
seemed to be wanted at the time. So @milosz I strongly agree with your post.

 

[Feanor]

I think this is true of many modern tube amps in general, and also some solid state amps where designers have chosen to largely eschew negative feedback.

Many audiophiles believe that negative feedback is ALWAYS bad. After all- it's NEGATIVE, isn't it? (I'm sure that bias operates at some level on buyers, even if they know better.) This stems from early solid state designers who didn't understand transistors all that well and created amps there were quite nonlinear and then (mis-)applied MASSIVE amounts of negative feedback so that these amps would measure OK in reviews. The sonic problems that this misguided approach can engender were discussed in papers by Matti Otala; many believe that these hack designs were responsible for the notion that solid state amps sound harsh, screechy, etc. Making a poor amplifier then using massive NFB as a remedy might result in something that measures OK with steady test signals, but often creates an amp that does poorly with rapidly changing signals like music.

What Mr. Otala was saying was that NFB is not a remedy for poor basic design, but that judicious use of NFB along with an amplifier topology that is linear to begin with is the way to go. What many audiophiles read from this is that "NFB is always BAD." And that myth is perpetuated till this day.

Some audiophiles will listen to an amp with a certain "voicing" in terms of certain amount of euphonic low order harmonic distortion and say "See! This sounds RIGHT!" because that is how they want it to sound, they like that harmonic syrup- and they end up thinking and promoting the idea that an amp with proper NFB, which doesn't have this "tube sound," is just wrong and that NFB must be the culprit. "NFB must be bad, because the amps I like don't use it..."

Yes, I know a few audiophiles who rant & rail against NFB. One for example professes to hate all op amps because the use that feedback. These people insist the higher order harmonic distortion created by NFB is the reason they like low-feedback designs.

In reality what they like is not so much the absence of higher order harmonics as the presence of the euphonic lower order HD.

Of course we know the if a little NFB is bad because it generates higher order HD, then much more NFB is actually better because it suppresses the whole HD spectrum including high order HD.

 

[March Audio]

For interest Bruno Putzeys on negative feedback

https://www.google.com/url?sa=t&sou...FjABegQIARAB&usg=AOvVaw3TQHX7dKj3pUiCoPCahlPA

 

[Sal1950]

Just shows to go ya, taking just about anything to an extreme will usually come out with negative results. (ha ha)
But truthful it is, that there just doesn't seem to be any exceptions to the rule of moderation in all things.

Seriously, it is unfortunate that without some minor design "flaws" done in the name of "voicing" this amp has all the bones needed to build a SOTA ca-1995-2020 amp.

 

[Oklei]

For interest Bruno Putzeys on negative feedback

https://www.google.com/url?sa=t&sou...FjABegQIARAB&usg=AOvVaw3TQHX7dKj3pUiCoPCahlPA

Great paper. I think it is mostly storyline 3 that explains the nfb hatred. It's the "look it measures perfectly but sounds like shit" tale searching for a semi scientific explanation.

 

[Feanor]

For interest Bruno Putzeys on negative feedback

https://www.google.com/url?sa=t&sou...FjABegQIARAB&usg=AOvVaw3TQHX7dKj3pUiCoPCahlPA

I wish I could understand some significant portion of what Putzey is saying; my technical and mathematical skills aren't up to the task, regrettably.

So for example, I wish I knew what he means by, "Open-loop bandwidth is no measure of how fast an amplifier is. Gain-bandwidth product is.". For those who can catch what "gain-bandwidth product" in the course of the article is it is probably quite evident but sadly I'm not one who can.

 

[DonH56]

I wish I could understand some significant portion of what Putzey is saying; my technical and mathematical skills aren't up to the task, regrettably.

So for example, I wish I knew what he means by, "Open-loop bandwidth is no measure of how fast an amplifier is. Gain-bandwidth product is.". For those who can catch what "gain-bandwidth product" in the course of the article is it is probably quite evident but sadly I'm not one who can.

Gain-bandwidth is the product of gain and bandwidth so an amplifier with higher gain and bandwidth will have much greater gain-bandwidth (GBW) product. A low-gain amplifier can have very wide bandwidth but without much gain you cannot apply much feedback. An amplifier with the same open-loop bandwidth but higher open-loop gain can apply greater feedback for higher performance (lower distortion, high input impedance, lower output impedance, and so forth). See simplified diagram below. Both amps have the same open-loop bandwidth, typically very low, and closed-loop (operating) gain, but the high-gain amplifier has much greater loop gain (the amount of gain used for feedback to suppress distortion and such) and also much greater closed-loop (normal operating) and unity-gain (x-axis crossing point) bandwidth.

HTH - Don

 

[Gorgonzola]

H2 at -70 dB. H3 at -98 dB.

With this intentional harmonic profile (to emulate SET tubes sound) the question is how complex recordings like electronic or orchestral sound. Maybe H2 about 84 dB would be the good compromise to ALL recordings, better than 70 - 73 dB.

I've been back to look at the review of this ARC D300 a number of times and feel that I would really like to audition one

I suspect the comment made by @maty the it was, in effect, to emulate a SET amp is dead on. Hopefully at this point there is broad agreement that the 2nd order distortion as in the D300 has a euphonic quality -- if not accuracy -- sot by many. It's noteworthy, IMHO, that all the prominent distortion is low or even order; higher order distortion is very low. Noise is not great but not too bad. The poor SINAD results apparently are entirely due to the high 2nd-4th order distortion, especially H2.

When I compare the D300 to the 'Recommended' Schiit Aegir I see that the latter is lower distortion only in the lower orders. Also, the Aegir is about the same for noise and much worse for higher order HD. Given a straight either/or choice between the D300 and the Aegir -- and apart from the much higher power of the former -- I'd choose the D300 without hesitation.

 

[SIY]

I've been back to look at the review of this ARC D300 a number of times and feel that I would really like to audition one

I suspect the comment made by @maty the it was, in effect, to emulate a SET amp is dead on. Hopefully at this point there is broad agreement that the 2nd order distortion as in the D300 has a euphonic quality -- if not accuracy -- sot by many. It's noteworthy, IMHO, that all the prominent distortion is low or even order; higher order distortion is very low. Noise is not great but not too bad. The poor SINAD results apparently are entirely due to the high 2nd-4th order distortion, especially H2.

When I compare the D300 to the 'Recommended' Schiit Aegir I see that the latter is lower distortion only in the lower orders. Also, the Aegir is about the same for noise and much worse for higher order HD. Given a straight either/or choice between the D300 and the Aegir -- and apart from the much higher power of the former -- I'd choose the D300 without hesitation.

If you "audition" one, see what happens level matched and not peeking versus a more accurate amplifier. I suspect the audibility of the distortion is questionable.

If I want 2HD, I can dial it in by software, giving me the option of turning it off.

 

[Katji]

If I want 2HD, I can dial it in by software, giving me the option of turning it off.

Examples of software I could use to try it?


................AFAIK the only valve amp I ever heard was the Fender Bassman amp I had, not relevant. ........................If there wasn't the "millenial factor", I'd think that all these valve amp fans nowadays are all over 70.

 

[SIY]

Examples of software I could use to try it?


................AFAIK the only valve amp I ever heard was the Fender Bassman amp I had, not relevant. ........................If there wasn't the "millenial factor", I'd think that all these valve amp fans nowadays are all over 70.

@pkane has written some terrific software for this. See for example this.

 

[Bridges]

He's right. late 80s to early 90s there were a ton of Japanese amplifiers with performance well in excess of what this has shown. But Audio Research have always done their own thing, not bowing to fashion or style and they have a loyal following to this day. Personally, I'm glad they have survived and are still in business.



Shall I start at A for Accuphase and go through the alphabet? Pretty much every power amp they produced from the late 1980s onwards would outperform by a wide margin this Audio Research.

There's plenty of Denons, Onkyos, Sonys, Pioneers, Luxmans, JVCs, Kenwoods, Technics and Yamahas that could drive <1ohm loads at stupendous powers and low THD with DC-Daylight responses. Remember, the Japanese were building the fast silicon high powered dedicated audio transistors and MOSFETs (Hitachi/Toshiba/Sanken/Fujitsu etc) Their premium models were often home market or EU and didn't make it to the US.

Here's a few random (early 90s) Accuphase THD vs Po plots I posted in another thread:

View attachment 40824
View attachment 40825
View attachment 40826

Don't forget Carver