Review and Measurements of Benchmark AHB2 Amp

[blueone]

The bass section is crossed at 105 Hz. It's truly unlistenable and nothing produced would lend one to use the term "speed"

You mean 150Hz.

 

[RichB]

You mean 150Hz.

Yes. Corrected.

- Rich

 

[Habu]

Matter of taste. I prefer the cleaner look of the Benchmark.

The cleaner look and the small size of the Benchmark !
11.04" W x 3.88" H x 9.34" D or (280mm W x 98mm H x 237mm) including feet and binding posts.

 

[Sal1950]

To my old experienced eyes they look too small to be real and expensive. It will take another generation of "philes" to know and appreciate the small form factor that these Class D type amps are capable of and give back stellar performance to boot.
Give me a big honkin pair of Krell monoblocks to ease my worried mine.. @RayDunzl has to invite a couple of big strong friends over any time Mrs. Dunzl wants to do the good housecleaning.

 

[waynel]

. It will take another generation of "philes" to know and appreciate the small form factor that these Class D type amps are capable of and give back stellar performance to boot.

Not Class D

 

[Wes]

I appreciate John's explanation, and the time he spends here discussing these matters. And I love my AHB2 (and DAC3). If only he could make it look as nice as an Accuphase, for Benchmark dollars...

Buy a busted up junked Accuphase case and put the AH2B inside it...

 

[QMuse]

Would this mean that DC stable amps like the old Citation 7.1 or the Kenwood Laboratory series or when the 80s era Sony’s have less phase response error?

Why bother when phaser error at 20Hz of any speaker is far greater then that of an amp?

 

[mocenigo]

Not Class D

He said "class D type" and the benchmark has a SMPS which is often associated with class D ;-)
But, yeah, you are correct. The Benchmark is an AH class amplifier, not AB, not D.

 

[Sal1950]

Not Class D

Yea I know but close enough. A exotic "tracking" power supply design more in the Class H style.
But when they get this small and efficent they're all class D to me.

 

[GXAlan]

Why bother when phaser error at 20Hz of any speaker is far greater then that of an amp?

Well, we're trying to figure out why there is perceived added bass with the Accuphase P-300 versus the AHB2. John Siau, from Benchmark, has suggested the phase error as being part of the problem.

He's also mentioned damping factor. This is one of those measurements that often is said to be not critical once you're past a certain point
http://www.lansingheritage.org/html/jbl/reference/technical/damping-factor.htm

But John Siau has also stated that this is a reason for the perceived difference in bass between the P-300 and AHB2.

350 at 20 Hz, 8-Ohms = Benchmark AHB2
60 at 40 Hz, 8 ohms = Accuphase P-300

It would be interesting to get a sense of what his opinion is the best practices. You have amps like Parasound JC1 with >1200 @ 20Hz or older Kenwood's with Sigma Drive that achieved 15,000 at 55Hz thanks to NFB sensed from the speaker terminals of the speaker itself.

 

[QMuse]

Well, we're trying to figure out why there is perceived added bass with the Accuphase P-300 versus the AHB2.

Is there a measurement done with the same speakers playing in the same room with 2 different amps which supports such difference or was it somebody's subjective perception?

 

[Kal Rubinson]

Yea I know but close enough. A exotic "tracking" power supply design more in the Class H style.
But when they get this small and efficent they're all class D to me.

.......and I thought that I was old.

 

[trl]

Was using an app called Decibel X pro - yes measuring spl with an iphone is not ideal, but I'm only interested in getting the magnitude of spl difference below the obvious discrepancy in the observed and heard differences. Used a 1kHz test tone with phone 1m away from speaker for each test. Similar levels for both tests, but very dissimilar results.

This should do, as long the position of the phone's mic and the distance between phone and speakers remains constant and you're using same sinewave generator and same audio file and amplitude.

 

[Coach_Kaarlo]

There are several important differences between these two amplifiers that can change the way the low frequencies are reproduced. The Accuphase response is -3 dB at 5 Hz while the AHB2 extends all the way down to 0.1 Hz (-3 dB). The effect of this is that Accuphase delivers frequencies near 20 Hz slightly late (due to the low-frequency phase response errors which are a direct result 5 Hz lower limit of the Accuphase amplifier). A second issue is that the damping factor is much higher in the AHB2. The Accuphase cannot damp the drivers as quickly. The third issue is that the Accuphase may have output current limitations that are contributing additional distortions. In contrast, the AHB2 output stage is drawing current from tightly regulated power supply rails and there is no current limiting circuit in series with the output stage. The output current of the AHB2 is only limited by an overload protection system that measures the output current and then shuts the amplifier down is an unsafe load (or short circuit) is causing excessive output current. Until this shutdown threshold is reached, there is no limit to the output current delivered by the AHB2. Before this threshold is reached, the high-current warning lights will flash, but this is not an indication that current is being limited, it is just a warning that you are near the shut-down threshold."

Thank you for your response John, I genuinely appreciate the time you took, and the insight gained whenever you do so. And for the record, I am 99% satisfied with this ground breaking amp, just curious about something I don't understand (doesn't make sense).


An old Accuphase amp was chosen because it was old, it was available, and it was similar in power but different in observed performance by a magnitude greater than the difference in specs - other amps were also tried but being more powerful they didn't really provide a representative comparison (purely subjective of course).

"With the Accuphase, the combination of late delivery and less damping will emphasize the bass on some recordings. You seem to have found one recording where this added emphasis was desirable in your room with your speakers, but it is quite likely that the bass in a different recording would sound uncontrolled.

You can always find one or two recordings that will be enhanced by a defect in the reproduction chain. The problem is that this same defect will have a detrimental effect on far more recordings than the few that it may "improve". It is also not possible to rule out a difference in playback level, given the fact that the levels were not accurately measured when making these comparisons. I like to play a test tone at low levels while connecting an AC volt meter across the speaker terminals."

I agree with your explanation that I could simply be focused on one anomaly rather than a regular occurence - one way to find out - I am going to repeat the test in a week or two, with similar tracks that feature big transients. See if I can replicate the behaviour on more than one track/ audio sample.

"The red clip lights on the AHB2 are triggered by a difference between the input signal at the XLR inputs and the output signal at the speaker terminals. If the difference is more than 1% the light will flash. In other words if the output sagged by 1 % or if the THD exceeded 1% this light would light. The fact that the lights are not coming on is iron-clad proof that the AHB2 is having no difficulty delivering the voltage and current necessary to move and damp your drivers. The light is triggered by a full input-to-output comparison, not simply by output-stage clipping.

If you are not lighting any warning lights on the AHB2, the voltage at the speaker terminals is precisely tracking the voltage at the input terminals (adjusted for the gain of the amplifier). The AHB2 is having no difficulty driving your speakers. The same may not be true of the other amplifier.

This is puzzling actually because at no point did a single light flicker on the AHB2. Not once. Is it possible the amp considers the short sharp transient signals as noise and ignores them? Or is it simply a question of available headroom? Or limits to rate of change rate? I think I understand the inbuilt protections in the AHB2 but they don't seem to explain the behaviour observed.


''That said, there is one aspect of the design that must be emphasized: The power supply is very strictly regulated, the 100 watts of continuous power generating only about 110 watts on instantaneous peaks. This is salutary because it keeps the amp within its noise and distortion specifications vis-à-vis its power output. But there are other amplifiers with less tightly regulated power supplies that allow for much greater power output on instantaneous peaks (this is a design feature on the NAD amps). The price there is increased distortion, but the argument goes that because the music is so loud, and the duration so brief, the extra distortion goes essentially unnoticed. Both approaches are valid, so you pays your money and takes your pick according to your needs, desires, and wallet. ''

This quote was taken from an Absolute Sound review of the AHB2. https://www.theabsolutesound.com/articles/benchmark-ahb2-amplifier/?page=2

I think some of the subsequent comments on the post were kind of going towards this idea. And it seems supported by what you were saying John (if my understanding is correct) - that the transients are tightly regulated in the AHB2.

As was pointed out to me by @GXAlan push-pull means 380W into 8 ohms with 29A in Mono, compared to 100W into 8 ohms and 18A in stereo. This lines up with my observations when running the amp in mono through the one speaker. [EDIT: meaning when running one speaker with the AHB2 in mono the driver movement was able to match the other Accuphase amp] So could it simply be a case of not enough headroom (1x AHB2 in stereo) for dynamics in a poor recording which are otherwise uncommon (the dynamics not the poor recording), or perhaps these dynamics normally meet other limits before coming up against the amps design itself? [EDIT: so the transient demand could never exceed the 110W stereo limit of the AHB2 yet for the mono test to make a difference it must?]

I have to admit some ignorance, but wouldn't those instantaneous peaks actually be instantaneously multiples higher than the power levels either side. Wouldn't running the AHB2 in mono reduce this - I think that could be what I observed.


Definitely very interested in analysing that particular section of the music (signal), and then maybe understanding what demand is being placed on the amp/ speakers dynamically.

[EDIT: link to audio levels of track portion in question added - https://drive.google.com/file/d/1b0rzJo2Sf2L_1j-jKSNOpcqmOJxbcY9X/view?usp=sharing

FYI the level in the audio goes from -16dB to -2 or -1 dB almost instantly]

Cheers.

 

[gvl]

If the clip LEDs didn't light up you were not hitting the headroom limit and the amp could deliver all the output that the input signal called for, that's what John said I think.

 

[anmpr1]

To my old experienced eyes they look too small to be real and expensive. It will take another generation of "philes" to know and appreciate the small form factor that these Class D type amps are capable of and give back stellar performance to boot.
Give me a big honkin pair of Krell monoblocks to ease my worried mine.. @RayDunzl has to invite a couple of big strong friends over any time Mrs. Dunzl wants to do the good housecleaning.

No doubt that Benchmark's form factor is the future, today. The old-style 'tons of iron' design will always be around, for nostalgia mostly. Growing up in the large box 'metered' and 'wood encased' era it's difficult to accept from a cosmetic standpoint. I look at the AHB2/DAC3 and am frankly amazed--due to its lack of size. The only thing I can relate it to is during the late '70s and early '80s some of the large Japanese manufacturers offered 'mini' components, which were essentially full featured amps/preamps about the size of current Benchmark kit.

That said, the fit 'n finish of Benchmark is first rate. The etched logo on the thick aluminum panel exudes its own kind of lux. And, let's face it, the gear is pro-oriented. With that in mind it looks a thousand percent better than something from (modern day) Crown, or Behringer--both looking quite agricultural (farm equipment) in appearance. And I don't mean in the Lamborghini tractor sense!

Obviously with my 'facepalm' I was being a bit tongue-in-cheek about wanting Benchmark to look like Accuphase.

 

[scott wurcer]

[EDIT: meaning when running one speaker with the AHB2 in mono the driver movement was able to match the other Accuphase amp] .

This would seem to indicate the observed driver motion is the correct behavior and not an issue with simple damping. I'm still curious about your observation of "no driver movement at all" with no indication of a fault condition.

 

[John_Siau]

Thank you for your response John, I genuinely appreciate the time you took, and the insight gained whenever you do so. And for the record, I am 99% satisfied with this ground breaking amp, just curious about something I don't understand (doesn't make sense).


An old Accuphase amp was chosen because it was old, it was available, and it was similar in power but different in observed performance by a magnitude greater than the difference in specs - other amps were also tried but being more powerful they didn't really provide a representative comparison (purely subjective of course).



I agree with your explanation that I could simply be focused on one anomaly rather than a regular occurence - one way to find out - I am going to repeat the test in a week or two, with similar tracks that feature big transients. See if I can replicate the behaviour on more than one track/ audio sample.



This is puzzling actually because at no point did a single light flicker on the AHB2. Not once. Is it possible the amp considers the short sharp transient signals as noise and ignores them? Or is it simply a question of available headroom? Or limits to rate of change rate? I think I understand the inbuilt protections in the AHB2 but they don't seem to explain the behaviour observed.


''That said, there is one aspect of the design that must be emphasized: The power supply is very strictly regulated, the 100 watts of continuous power generating only about 110 watts on instantaneous peaks. This is salutary because it keeps the amp within its noise and distortion specifications vis-à-vis its power output. But there are other amplifiers with less tightly regulated power supplies that allow for much greater power output on instantaneous peaks (this is a design feature on the NAD amps). The price there is increased distortion, but the argument goes that because the music is so loud, and the duration so brief, the extra distortion goes essentially unnoticed. Both approaches are valid, so you pays your money and takes your pick according to your needs, desires, and wallet. ''

This quote was taken from an Absolute Sound review of the AHB2. https://www.theabsolutesound.com/articles/benchmark-ahb2-amplifier/?page=2

I think some of the subsequent comments on the post were kind of going towards this idea. And it seems supported by what you were saying John (if my understanding is correct) - that the transients are tightly regulated in the AHB2.

As was pointed out to me by @GXAlan push-pull means 380W into 8 ohms with 29A in Mono, compared to 100W into 8 ohms and 18A in stereo. This lines up with my observations when running the amp in mono through the one speaker. [EDIT: meaning when running one speaker with the AHB2 in mono the driver movement was able to match the other Accuphase amp] So could it simply be a case of not enough headroom (1x AHB2 in stereo) for dynamics in a poor recording which are otherwise uncommon (the dynamics not the poor recording), or perhaps these dynamics normally meet other limits before coming up against the amps design itself? [EDIT: so the transient demand could never exceed the 110W stereo limit of the AHB2 yet for the mono test to make a difference it must?]

I have to admit some ignorance, but wouldn't those instantaneous peaks actually be instantaneously multiples higher than the power levels either side. Wouldn't running the AHB2 in mono reduce this - I think that could be what I observed.


Definitely very interested in analysing that particular section of the music (signal), and then maybe understanding what demand is being placed on the amp/ speakers dynamically.

[EDIT: link to audio levels of track portion in question added - https://drive.google.com/file/d/1b0rzJo2Sf2L_1j-jKSNOpcqmOJxbcY9X/view?usp=sharing

FYI the level in the audio goes from -16dB to -2 or -1 dB almost instantly]

Cheers.

Your assumptions are well thought out but incorrect. The AHB2 power supply is tightly regulated. The transients are not regulated.

If the clip lights on the AHB2 are not turning on, the transients are not being clipped, attenuated, or distorted. The clip lights have a timer that keeps them on for at least 1/4 second so that every clip event is always visible. If the lights are not turning on, the output is following the input exactly. Remember that the AHB2 also has a 200 kHz bandwidth, so it also follows high-frequency transients with precision and with phase accuracy.

What I failed to point out in my prior post is that a poor damping factor can have and audible impact on the frequency response of the amplifier. The Accuphase has a relatively low damping factor (60) and the variations in speaker impedance at low frequencies can have a big impact on the frequency response of the Accuphase amplifier. A damping factor of 60 means that that output impedance of the Accuphase is 8Ohms/60=0.13 Ohms. If your speakers dip to 1 Ohm at low frequencies (not uncommon), this would create a 1 dB error in the frequency response of the Accuphase and this would be accompanied by a significant change in the phase response. Difficult speaker phase angles will have a similar effect on the response.

The published frequency response of the Accuphase is flat but this measurement was made using an ideal resistive load. Speakers are not an ideal load.

Regarding the cone excursion: A non-linear phase response at low frequencies can have a significant impact on the excursion of the speaker cone. This is essentially a low-frequency ringing caused by a non-linear phase response. The cone is overshooting beyond the desired movement. This will cause an emphasis of resonant frequencies and a suppression of other frequencies.

 

[Soniclife]

This would seem to indicate the observed driver motion is the correct behavior and not an issue with simple damping. I'm still curious about your observation of "no driver movement at all" with no indication of a fault condition.

So am I if the levels were closely matched. I wonder what happens if the volume is turned up on the stereo version until the cone starts moving, the clip light comes on, or fear of something breaking kicks in.

 

[GXAlan]

@John_Siau
Could the gain difference between the P-300 and the AHB2 also be an issue in terms of having the P-300 combination outputting at higher volumes without careful level matching?

P-300, input sensitivity is 1V, 150W into 8
AHB2, input sensitivity is 2V, 100W into 8

Any comments on the Kenwood Sigma Drive philosophy to get ultra-high damping factors?

----
@Coach_Kaarlo
What did you have set up for the gain?

16 dB swings doesn't seem like much. How loud were you listening?