• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Review and Measurements of Benchmark AHB2 Amp

Joined
May 4, 2020
Messages
192
Likes
206
An application note by John Siau (VP at Benchmark) on damping factor published today, which also describes Benchmark's approach to damping factor and discusses the AHB2 and Benchmark/Canare cables related to the note's topic, is available here.
Interesting stuff, especially the effect of the cables thank you for posting.

After successfully asking google to convert your annoying American units, I got a total error of 0,19dB for my system.
Replacing my 230€ budget AVR with a 4K€ AHB-2 would push that down to 0,12dB.
I'd wager both values are inconsequential in normal listening.
 

samsa

Active Member
Joined
Mar 31, 2020
Messages
283
Likes
255
An application note by John Siau (VP at Benchmark) on damping factor published today, which also describes Benchmark's approach to damping factor and discusses the AHB2 and Benchmark/Canare cables related to the note's topic, is available here.
There is something a bit messed up about that blog post.

Siau says that his 10 feet of his 11 AWG cable has a return resistance of 0.0252 Ω. If you look up the resistance/foot of 11 AWG copper cable and multiply by 20 feet (for the return resistance), you indeed get 0.0252 Ω.

But he says that the same length of 12 AWG cable has a return resistance of 0.04 Ω, 1.6 times larger than his 11 AWG cable. From the same table, the return resistance of a 10 foot length of 12 AWG copper cable is 0.0318 Ω. Only 1.26 times larger than his 11 AWG cable. For the purposes of his discussion, this is not a small correction! Maybe, in citing the 0.04 Ω figure, he's talking about copper-clad aluminum, rather than copper. But, if so, he should say so.
 
Joined
May 4, 2020
Messages
192
Likes
206
Siau says that his 10 feet of his 11 AWG cable has a return resistance of 0.0252 Ω

But he says that the same length of 12 AWG cable has a return resistance of 0.04 Ω, 1.6 times larger than his 11 AWG cable.
From the same table, the return resistance of a 10 foot length of 12 AWG copper cable is 0.0318 Ω.
Hmm the spreadsheet they linked, uses the correct values for both. Maybe it was just an honest typo?
 

John_Siau

Active Member
Technical Expert
Industry Insider
Joined
Jul 17, 2018
Messages
117
Likes
905
Location
Syracuse, NY USA
thx @pjug.
So in normal language a well designed and engineered amplifier working not in clipping or distortion have inaudible differences? Can we say this?
Here is an A/B/X test that we did in our listening room that clearly identified differences between an AHB2 and another amp with reasonably low distortion specifications. The difference was produced by zero-crossing distortion when playing a single tone at a fraction of a watt. The other amplifier produced harmonics that could be clearly identified in the A/B/X test. Test results show a perfect score, meaning the differences were very clearly audible:
https://benchmarkmedia.com/blogs/application_notes/power-amplifiers-the-importance-of-the-first-watt

Also, my latest application note explains why damping factor can produce audible changes in the frequency response of the system:

https://benchmarkmedia.com/blogs/application_notes/audio-myth-damping-factor-isnt-much-of-a-factor
 

John_Siau

Active Member
Technical Expert
Industry Insider
Joined
Jul 17, 2018
Messages
117
Likes
905
Location
Syracuse, NY USA
There is something a bit messed up about that blog post.

Siau says that his 10 feet of his 11 AWG cable has a return resistance of 0.0252 Ω. If you look up the resistance/foot of 11 AWG copper cable and multiply by 20 feet (for the return resistance), you indeed get 0.0252 Ω.

But he says that the same length of 12 AWG cable has a return resistance of 0.04 Ω, 1.6 times larger than his 11 AWG cable. From the same table, the return resistance of a 10 foot length of 12 AWG copper cable is 0.0318 Ω. Only 1.26 times larger than his 11 AWG cable. For the purposes of his discussion, this is not a small correction! Maybe, in citing the 0.04 Ω figure, he's talking about copper-clad aluminum, rather than copper. But, if so, he should say so.
Thanks for catching that error. It is good to see that some people are reading carefully! I fixed the error. BTW, I used the correct number when calculating the resulting damping factor of 200 (due to the cable resistance). The 0.4-Ohm number was for a 12 foot 12-AWG cable. We don't sell 12-foot cables, so I had changed the calculations to reflect 10-foot cables, but missed that number.
I am very impressed that you found the error!

Five gold stars to samsa!
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,078
Likes
2,449

John_Siau

Active Member
Technical Expert
Industry Insider
Joined
Jul 17, 2018
Messages
117
Likes
905
Location
Syracuse, NY USA
One thing I wasn't expecting, and haven't seen before - I bought a pair of AHB2s for a system I'm setting up in Colorado, and the (excellent) manual says they're unsafe to operate above 2,000 meters, which is about 5,600 feet. Not a problem for their eventual destination, which is at 5,000 feet, but while I'm waiting for construction to finish, I had them delivered to my place in Wyoming, which is at 8,400 feet. I'll call Benchmark after the holiday, because I'm interested in the reason, but in the meantime, what are your best guesses?
Amplifier RMS power ratings are based on the ability to run a 1/3 power continuously for one hour, before testing at full output. The 1/3 power preconditioning usually brings an amplifier up close to its thermal limits. High altitudes reduce the cooling and the amplifier may reach the thermal shutdown limit in less than an hour. Above 2000 meters (6562 feet), the maximum temperature will be reached in a little less than an hour. This has no impact on normal listening and and no impact on safety. The amplifier has multiple thermal sensors that will shut the unit down before an unsafe temperature is reached.
 
Last edited:

Bombadil

Active Member
Forum Donor
Joined
Jan 28, 2017
Messages
195
Likes
149
Location
Davis CA
Here is an A/B/X test that we did in our listening room that clearly identified differences between an AHB2 and another amp with reasonably low distortion specifications. The difference was produced by zero-crossing distortion when playing a single tone at a fraction of a watt. The other amplifier produced harmonics that could be clearly identified in the A/B/X test. Test results show a perfect score, meaning the differences were very clearly audible:
https://benchmarkmedia.com/blogs/application_notes/power-amplifiers-the-importance-of-the-first-watt

Also, my latest application note explains why damping factor can produce audible changes in the frequency response of the system:

https://benchmarkmedia.com/blogs/application_notes/audio-myth-damping-factor-isnt-much-of-a-factor
Thanks for this info. Wouldn't it be more applicable if you did an ABX comparison using actual recordings of music at low power levels, rather than a test tone?
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,078
Likes
2,449
Here is an A/B/X test that we did in our listening room that clearly identified differences between an AHB2 and another amp with reasonably low distortion specifications. The difference was produced by zero-crossing distortion when playing a single tone at a fraction of a watt. The other amplifier produced harmonics that could be clearly identified in the A/B/X test. Test results show a perfect score, meaning the differences were very clearly audible:
https://benchmarkmedia.com/blogs/application_notes/power-amplifiers-the-importance-of-the-first-watt
What is the THD of your SMS1 speaker at 1kHz at 0.01W?

What was the noise level in the room in which you were doing ABX test with 0.01W and 1kHz test tone?


Quote from your paper: "If amplifier distortion is audible with a test tone, it may also be audible while playing music. "

That really is a bit of a stretch and should be proved by at least one music material with ABX test.
 

John_Siau

Active Member
Technical Expert
Industry Insider
Joined
Jul 17, 2018
Messages
117
Likes
905
Location
Syracuse, NY USA
Here's a comparison I did for using an AHB-2 to my now sold Emotiva XPR-1 monoblocks I used to drive some Klipsch P-17b bookshelf speakers. I don't have any speaker data for the P-39f drivers or I'd do similar calculations for them. The bookshelf speakers still have the factory crossover in them and sit in my office.

From: https://hometheaterhifi.com/reviews...fier/emotiva-xpr-1-monoblock-power-amplifier/
View attachment 65452

View attachment 65447


Emotiva XPR-1 monoblocks (max output and 1 watt output):
View attachment 65464
View attachment 65465

AHB-2 bridged mode (max output and 1 watt output):
View attachment 65466
View attachment 65467

Alternate method of calculating pressure for a given voltage to prove to yourself that Benchmark's rule of thumb makes sense (which it should provided the speaker specs are measured at 2.83Vrms):
View attachment 65468

Note that for the noise SPL figure, it will likely depend on your environment.

A while back I did an accurate measurement of my room in this thread:
https://www.audiosciencereview.com/...hat-level-is-noise-heard-in-your-system.1013/

The spectrum below is for my room (measured with Earthworks M30, Benchmark MPA1, Benchmark ADC1 sending AES to my RME soundcard). Note that the room's inherent noise varies with frequency over the audible range by A LOT! My room is actually very quiet in the md-range.

View attachment 65469

@RichB
I love the photos of your hand-written calculations!

One correction:

Don't assume that something is inaudible if it is lower than the room noise. We have the ability to hear a 3 kHz tone that is 30 dB lower than white noise. I often do this demo for visitors to Benchmark. Our ears are amazing.

If you do an analysis, you will find that many of the musical details that you are hearing in a recording, are playing at levels below that of your room noise. This is especially true of natural reverberant sounds.

Distortion does not normally sound like random noise and this makes it easier for our ears to detect underneath background noise. Most of the masking of distortion is due to the fact that musical instruments already have significant harmonic content. IMD is not harmonic and is not masked by the harmonic content of the instruments.

I find that solo piano recordings are very good for exposing harmonic distortion. The reason for this is that the overtones of the piano are slightly non-harmonic. The overtones are stretched to a slightly higher than integer ratio above the fundamental. Harmonic distortion from the electronics are always spaced by exact integer ratios. The electronically-generated harmonics audibly beat against the piano's overtones, making the piano sound like its strings are out of tune with each other. I first noticed this effect after tuning a piano for the first time. I am not a professional piano tuner, but I tune our home piano from time to time. It is a good learning experience.
 

pjug

Senior Member
Forum Donor
Joined
Feb 2, 2019
Messages
393
Likes
259
Here is an A/B/X test that we did in our listening room that clearly identified differences between an AHB2 and another amp with reasonably low distortion specifications. The difference was produced by zero-crossing distortion when playing a single tone at a fraction of a watt. The other amplifier produced harmonics that could be clearly identified in the A/B/X test. Test results show a perfect score, meaning the differences were very clearly audible:
https://benchmarkmedia.com/blogs/application_notes/power-amplifiers-the-importance-of-the-first-watt
Do decent Class AB amplifiers really have distortion of the magnitude shown in that app note? For example, Amir's test of the Outlaw M2200 showed distortion below -100dB (5W).
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,078
Likes
2,449
Don't assume that something is inaudible if it is lower than the room noise. We have the ability to hear a 3 kHz tone that is 30 dB lower than white noise. I often do this demo for visitors to Benchmark. Our ears are amazing.
Sure we do. But question here is if we would be able to hear 2nd and 3rd harmonic distortion component of the 1kHz or 3 kHz tone which is 40 or 50 dB lower than the base tone
 

John_Siau

Active Member
Technical Expert
Industry Insider
Joined
Jul 17, 2018
Messages
117
Likes
905
Location
Syracuse, NY USA
When 2 x AHB2's are used in mono presumably the damping factor is different? Am I correct in saying it's halved?
Yes sir, the damping factor is cut in half when running bridged mono. But, in bridged mono, you can, and should, place your amplifiers next to the speakers so that you can use short cables. The shorter cables make up the difference.
 

RichB

Addicted to Fun and Learning
Forum Donor
Joined
May 24, 2019
Messages
793
Likes
811
Location
Massachusetts
Quote from your paper: "If amplifier distortion is audible with a test tone, it may also be audible while playing music. "

That really is a bit of a stretch and should be proved by at least one music material with ABX test.
It's a truly odd phenomenon, the veracity of those making the case that products that measures orders of magnitude better than other can have audible better performance.

So tell us, what is the SPL and frequency of distortion that above the 0 SPL that is tolerable to remain transparent?
Please, include the level and character of room noise, the type of music, the listening levels, and the distortion profile and sensitivity of the speaker.

If you know better, please enlighten us.

- Rich
 
Last edited:

GXAlan

Addicted to Fun and Learning
Forum Donor
Joined
Jan 15, 2020
Messages
506
Likes
700
Here is an A/B/X test that we did in our listening room that clearly identified differences between an AHB2 and another amp with reasonably low distortion specifications. The difference was produced by zero-crossing distortion when playing a single tone at a fraction of a watt. The other amplifier produced harmonics that could be clearly identified in the A/B/X test. Test results show a perfect score, meaning the differences were very clearly audible:
https://benchmarkmedia.com/blogs/application_notes/power-amplifiers-the-importance-of-the-first-watt
1) @John_Siau , without naming the amplifier can you at least say something like the MSRP or approximate weight? In the automotive world, people are happy to say that they benchmarked their car against the Porsche 911 or “insert competitor here”. It would be nice to understand what is going on.

2) I think what you are saying is that there are audible differences between amplifiers, and that the science of it is a) crossover distortion b) damping factor. But with all of the caveats you speak of.
 

John_Siau

Active Member
Technical Expert
Industry Insider
Joined
Jul 17, 2018
Messages
117
Likes
905
Location
Syracuse, NY USA
Do decent Class AB amplifiers really have distortion of the magnitude shown in that app note? For example, Amir's test of the Outlaw M2200 showed distortion below -100dB (5W).
The problem with traditional class AB amplifiers is the thermal stability of the bais circuit. If the biasing of the class-AB output stage is ideal, the crossover distortion can be nulled to a relatively low level, but it is very difficult to maintain this over temperature range and over time. This nulling is also impacted by the load impedance and phase angles. Remember, most tests are performed into a resistive load.

When crossover distortion does occur, it raises havoc with the amplifier's feedback network. The transient steps produced by the zero crossing, get extended in time by the slow response of the global feedback network.

The ABH2 avoids this by using high-bandwidth feed-forward error correction combined with a minimal amount of feedback. This is part of the reason why the AHB2 can have a bandwidth that exceeds 200 kHz, while being stable into any phase angle within the audible band. Feed-forward correction is inherently stable.
 

RichB

Addicted to Fun and Learning
Forum Donor
Joined
May 24, 2019
Messages
793
Likes
811
Location
Massachusetts
Yes sir, the damping factor is cut in half when running bridged mono. But, in bridged mono, you can, and should, place your amplifiers next to the speakers so that you can use short cables. The shorter cables make up the difference.
If the distance remains 10' and using bridged mono is there a gauge of wire that can make up the difference. At some point, does thicker gauge cause other issues?

- Rich
 

pjug

Senior Member
Forum Donor
Joined
Feb 2, 2019
Messages
393
Likes
259
The problem with traditional class AB amplifiers is the thermal stability of the bais circuit. If the biasing of the class-AB output stage is ideal, the crossover distortion can be nulled to a relatively low level, but it is very difficult to maintain this over temperature range and over time. This nulling is also impacted by the load impedance and phase angles. Remember, most tests are performed into a resistive load.

When crossover distortion does occur, it raises havoc with the amplifier's feedback network. The transient steps produced by the zero crossing, get extended in time by the slow response of the global feedback network.

The ABH2 avoids this by using high-bandwidth feed-forward error correction combined with a minimal amount of feedback. This is part of the reason why the AHB2 can have a bandwidth that exceeds 200 kHz, while being stable into any phase angle within the audible band. Feed-forward correction is inherently stable.
Thank you for the reply. We are fortunate to have experts like you posting here.
Here's another question; maybe you won't want to answer. If you couldn't have an AHB2, would you prefer a good traditional Class AB amplifier, or a good Class D amplifier?
 
Top Bottom