Review and Measurements of Benchmark AHB2 Amp

[typericey]

They give them out with the ID cards at RMAF.

Full disclosure: I use Nordost Blue Heaven speaker cables that my audioph*** dad let me borrow indefinitely. I didn't bother replacing them coz they seem to get the job (of being a cable) done well. Why fix something that isn't broken? Speaking of isn't broken, the cables are at least 12 years old, and the plastic sleeve isn't yellow nor brittle, zero oxidation of the wires (still as silvery as new), no oxidation of the gold plated banana plugs and no failure whatsoever. Gotta give them credit for these.

About the vid: John mentioned that they went discrete instead of op amp chips (noteworthy).
And then a couple of considerations: the amp design seems so complex in order to measure well. This means too many components that can fail. It also implies that no one can repair these except maybe Benchmark. Is it designed to be repairable at all? Does Benchmark accept out of warranty repairs? I ask because if an amp performs this good then I suppose people would want to live with them longer than their wives.

 

[McFly]

So the amplifier doesn't limit transients, but the protection system decides what is 'tweeter damaging' and mutes the amplifier at high level, high frequencies. Right. And it's the same 'tweeter damaging' sine that triggered the protection at 133W@4ohms (well below spec) (both channels driven) in Amir's AP high to low sweep at approximately 45Hz after a period of less than 5 minutes duration? What evidence is there of range switching relay transients at that point?

View attachment 28382

Once again, I draw your attention to your continuous rated power specifications and their specified bandwidth of 20Hz-20Khz.

Simple question: Does your amplifier provide full power at any frequency from 20Hz-20KHz as per the rated specifications for a period of not less than 5 minutes*, both channels driven into 8/6/4/3 ohms without shutting down or muting? *as per FTC specs

Surely, this shouldn't need to be raised four times to get a straight answer.

So can we weld with this thing or NOT?! just kidding, I was the one who spotted this 'issue'. Benchmark has some explaining to do here.

 

[McFly]

There is a table in AHB2 manual stipulating different combination of lights:

View attachment 26586

The labels are a bit misleading in that I got the TEMP light to come on with a speaker terminal short as the table indicates.

Amir what "fault" was the amplifier showing when it went into protection after the AP sine sweep tripping protection at 133W into 4Ohms?

 

[amirm]

Amir what "fault" was the amplifier showing when it went into protection after the AP sine sweep tripping protection at 133W into 4Ohms?

My memory is only good for a day or two after reviews. I did not record what the lights did other than the unit shut down during that test.

 

[digitalfrost]

Anybody interested in getting one of these, here's a deal: https://www.ebay.de/itm/223565802321

 

[KSTR]

The tweeter protection, as implemented, would not be appropriate in a sound reinforcement environment, but the AHB2 is not intended to be used in that application. It is intended to be used in studio control rooms and in high-end hi-fi systems. In both of these intended applications, the amplifier will often be driving very expensive speakers that could easily be destroyed by the available power.

A very sane decision, AHB2 users likely will own expensive passive speakers which usually don't have any in-built tweeter protection, and the recording crowd also don't want tweeters to burn out from the occasional mixing desk feedback incident and such. Better be safe than sorry.
Only active speakers really have a chance to tailor maximum HF spot power vs tweeter protection. In the HEDD Type30 I'm using 300Watts amps on a 10W tweeter and a 60W midrange and a sophisticated protection was required, truly modelling thermals and dissipated power in the limiter side-chains.

 

[RichB]

So the amplifier doesn't limit transients, but the protection system decides what is 'tweeter damaging' and mutes the amplifier at high level, high frequencies. Right. And it's the same 'tweeter damaging' sine that triggered the protection at 133W@4ohms (well below spec) (both channels driven) in Amir's AP high to low sweep at approximately 45Hz after a period of less than 5 minutes duration? What evidence is there of range switching relay transients at that point?

View attachment 28382

Once again, I draw your attention to your continuous rated power specifications and their specified bandwidth of 20Hz-20Khz.

Simple question: Does your amplifier provide full power at any frequency from 20Hz-20KHz as per the rated specifications for a period of not less than 5 minutes*, both channels driven into 8/6/4/3 ohms without shutting down or muting? *as per FTC specs

Surely, this shouldn't need to be raised four times to get a straight answer.

Where these measurements taken in bridged mode?

First off, Benchmark has some of the most complete set of measurements published on their web site and manual.

From the AHB2 manual (rev G), the specification are at 1kHz:

Rated Output Power 1 kHz, < 0.00015% THD
- 100 W/channel into 8 Ohms, both channels driven
- 130 W/channel into 6 Ohms, both channels driven
- 190 W/channel into 4 Ohms, both channels driven
- 240 W/channel into 3 Ohms, both channels driven
- 200 W Bridged Mono into 16 Ohms
- 380 W Bridged Mono into 8 Ohms
- 480 W Bridged Mono into 6 Ohms

From the web site, 20Hz to 20kHz:

Continuous Average Output Power < 0.0003 % THD+N at full rated power, 20 Hz to 20 kHz

- 100 Watts per channel into 8 Ohms, both channels driven
- 130 Watts per channel into 6 Ohms, both channels driven
- 190 Watts per channel into 4 Ohms, both channels driven
- 240 Watts per channel into 3 Ohms, both channels driven
- 200 Watts into 16 Ohms, bridged mono
- 380 Watts into 8 Ohms, bridged mono
- 480 Watts into 6 Ohms, bridged mono

When shipped the AHB2, the Output Current was listed at 18 Amps peak.
The spec on the web site is now:

29 A peak into 1 Ohm, both channels driven

I would interpret the web site to mean that the continue power applies to all the items listed.
Continuous is listed but not "FTC".

I appreciate that the amp has extensive protection circuitry that is not in the signal path.
There are separate indicators for each channel, that's a first for me where most amps don't have clip indicators.

I have run this amp in stereo mode just below clipping driving the Salon2's and got them warm, not hot.
The only operational issue I had was the center channel amp was not turning on, which I am blaming on the cat loosening the trigger cable.
I plugged it in firmly and it works fine.

- Rich

 

[oivavoi]

I have run this amp in stereo mode just below clipping driving the Salon2's and got them warm, not hot.
The only operational issue I had was the center channel amp was not turning on, which I am blaming on the cat loosening the trigger cable.
I plugged it in firmly and it works fine.

- Rich

Curious what the clipping level is for the AHB2 with the Salons, Rich? How loud peaks are you able to get?

 

[RichB]

Curious what the clipping level is for the AHB2 with the Salons, Rich? How loud peaks are you able to get?

Louder than I care to listen.
I have used Lorde Royals as a test track with Oppo UPD-205 -> Benchmark LA-4 -> AHB2 (low gain) -> Salon2's.
The AHB2 base amp clip indicators blink at about +14 and the upper section amp clips at +15.
I would bridge these amps if I wanted more.

One thing I can do is to measure the volume using sine-waves and pink noise at my listening position at 2.83 volts, measured at the listening position. From there, compute the level attainable assuming the speaker is at 4 ohms.

- Rich

 

[oivavoi]

Louder than I care to listen.
I have used Lorde Royals as a test track with Oppo UPD-205 -> Benchmark LA-4 -> AHB2 (low gain) -> Salon2's.
The AHB2 base amp clip indicators blink at about +14 and the upper section amp clips at +15.
I would bridge these amps if I wanted more.

One thing I can do is to measure the volume using sine-waves and pink noise at my listening position at 2.83 volts, measured at the listening position. From there, compute the level attainable assuming the speaker is at 4 ohms.

- Rich

Cool, sounds good. I'm actually mostly interested in short dynamic peaks. I would like to get to 115 db peaks if possible, to be sure that the amp never clips. I never listen to more than 82-83 "baseline" listening, but I listen to lots of acoustic recordings with peaks that can get much much higher than the baseline.

 

[John_Siau]

So the amplifier doesn't limit transients, but the protection system decides what is 'tweeter damaging' and mutes the amplifier at high level, high frequencies.

Once again, I draw your attention to your continuous rated power specifications and their specified bandwidth of 20Hz-20Khz.

Simple question: Does your amplifier provide full power at any frequency from 20Hz-20KHz as per the rated specifications for a period of not less than 5 minutes*, both channels driven into 8/6/4/3 ohms without shutting down or muting? *as per FTC specs

Surely, this shouldn't need to be raised four times to get a straight answer.

The answer to your question is yes. The amplifier has been tested to FTC specs with the tweeter protection turned off.

The tweeter protection circuit limits the amount of time a full amplitude sine wave can be reproduced at full power. You will also find that you can output full power at high frequencies for several seconds before the protection kicks in (and that it cannot be triggered by music). The reason it kicked in at 133 Watts had to do with the long duration of the sweep which started at high frequencies. Relay switching transients in the AP test set may have also been a contributing factor.

The protection circuit has a frequency-sensitive integrator function that runs in the FPGA. The protection circuit does not interact with the output stage in any way other than turning on the mute when a potential for speaker damage is detected.

 

[etc6849]

If anything all these posts help keep this thread at the top which is where it should be. More people need to hear about Benchmark and especially the Benchmark AHB2 amps.

I've had several people tell me who is Benchmark... Of course, all are totally blown away when I play them some music on my actively tri-amped 5.1 setup. These amps really brought my setup to a new level (as in better sounding than any of the audiophool $300k+ setups I've heard in shops).

That's why I am letting it be. John did find his way through the mess to answer a few questions just yesterday so we are good there.

 

[Tks]

How people feel hour-to-hour and day-to-day (and so forth) is a huge player in what sounds good/different at the time IME/IMO. Properly done level-matched ABX/DBT testing with rapid switching takes that long-term physiological/emotional stability out of the equation. Certainly long-term trends can be valid, but what if the long-term trend is based upon e.g. two components that are not level-matched? What if you listen to the new amp a little louder just to hear how good it sounds, and then switch back to the old one at a little lower level? The lower level means the older one won't have as much bass, may not sound as "warm", etc. There are just way too many variables to screw up for me... I have to either do a good ABX test or pull out the measurement gear.

Should I throw subjectivists a bone since they aren't even capable of coming up with ridiculous refutations anymore?

Here's one for any on-lookers that want to have something they can take home back to base:

"The emotional stability of a person isn't in complete stasis, thus any claim of rapid ABX switching cannot be done fast enough with the speed in which atoms function. Since you objectivists can't quantify how much emotional/psychological stability exists for every metric we have for time-scale, you can't conclusively prove rapid switching is rapid enough to avoid the stability problem"

Give them some ammo to throw around and make them look more insane if they ever said something like this.

 

[RichB]

Should I throw subjectivists a bone since they aren't even capable of coming up with ridiculous refutations anymore?

Here's one for any on-lookers that want to have something they can take home back to base:

"The emotional stability of a person isn't in complete stasis, thus any claim of rapid ABX switching cannot be done fast enough with the speed in which atoms function. Since you objectivists can't quantify how much emotional/psychological stability exists for every metric we have for time-scale, you can't conclusively prove rapid switching is rapid enough to avoid the stability problem"

Give them some ammo to throw around and make them look more insane if they ever said something like this.

We were on a roll for a while....

- Rich

 

[restorer-john]

The answer to your question is yes. The amplifier has been tested to FTC specs with the tweeter protection turned off.

I finally appreciate an answer.

To me, it's a consumer electronics version of the MCAS software debacle in the 737MAX-8. As shipped from the factory, we have a system that takes over, can't be shut-off easily and prevents the product behaving like the tens of thousands of amplifiers tested before it, and worse still, not meeting its own advertised specifications which require 5 minutes at advertised power, not a mere and non-specific 'several seconds' before it crashes and shuts down.

Unless the user can easily defeat the time/frequency integrator part of the protection, you should either amend the specifications to reflect the true continuous ratings in accordance with the FTC's Amplifier Rule or remove the reference to continuous average power and call it 'several seconds' power.

Consider since 1974, manufacturers have played by those rules. Their amplifiers were big, hot, had tons of heatsink area, tons of dissipation in silicon, massive power supplies and little or no current limiting apart from short circuit protection, DC and sometimes over-voltage. Those amplifiers met or exceeded their ambitious specifications, year after year. They were built to do it. In short, consumers were getting what they paid for.

Continuous is listed but not "FTC".

Here is the relevant part of the FTC's Amplifier Rule:

The rule was retained after industry consultation:

This Benchmark power amplifier is no doubt a lovely product, built well and performs admirably in all areas except meeting its continuous power output ratings. Rather important wouldn't you say?

 

[Thomas savage]

Given the Market it's going into I think this amps dam near perfect and a big thanks to benchmark for helping those guys avoid blown drivers in their exotic speakers.

Trust me , it takes a age to get anything out of beautique speaker manufacturers so benchmark are actually helping folks here enormously.

It's not in the signal path and won't limit real world performance ( as far as I understand) .

Or you can remove these safeguards pass some test and blow up some drivers. For what ?

Oh and btw no one really deserves this amp to be this good for this price , it's basically a public service the benchmark folks are doing here so I'd just be grateful.

 

[restorer-john]

Or you can remove these safeguards pass some test and blow up some divers. For what ?

Same as advertising a vehicle with a top speed it cannot achieve due to a limiter they place on all the production versions sold...

5 minutes is 5 minutes. It is very clear.

The amplifier rule is there to stop deceptive advertising practices and ensure products meet their specs. There's way too much ducking, weaving and hiding behind 'these aren't the droids you are looking for' type hand waving.

 

[typericey]

Out of curiosity: is there still a critical mass of manufacturers who adhere to this FTC rule in 2019? I remember in the golden age of car audio there was CEA-2006...

Personally speaking, doesn't matter if the AHB2 doesn't meet FTC. It's current rating for real world musical transients (a few seconds?) is a whole lot already given it's size, price and measurements. I think doing 190W @4 ohms for 5 mins while staying clean will already defy physics given the size of this amp. The declared specs doesn't say "FTC" so IMO no deception here.

 

[blueone]

To me, it's a consumer electronics version of the MCAS software debacle in the 737MAX-8. As shipped from the factory, we have a system that takes over, can't be shut-off easily and prevents the product behaving like the tens of thousands of amplifiers tested before it, and worse still, not meeting its own advertised specifications which require 5 minutes at advertised power, not a mere and non-specific 'several seconds' before it crashes and shuts down.

John, I haven't been here very long, and typically I do enjoy your posts. But comparing AHB2's limiting logic to MCAS doesn't seem fair. MCAS is a make-shift software fix to potential aircraft handling problems that were caused by a risky design solution to fit larger engines on an airframe that wasn't originally designed for engines of that size. While my objective isn't to call out Boeing here, IMO Boeing really screwed up by trying to avoid a landing-gear redesign, which would seem to have resulted in significant airframe modifications, which would have caused Boeing to go through essentially a whole-plane redesign effort that would have delayed the MAX, and they'd have no product to compete with the A320neo. So Boeing took what could be construed as a corner-cutting path to market. Then Boeing compounded the problem by not educating and training pilots, having some control, ahem, design and implementation snafus, and apparently miscalculated how often MCAS would be activated, resulting in two crashes and a lot of lost lives.

Benchmark is probably guilty of a bit of lack of communication, but how would this limiter manifest itself in any real-world system playing music? Benchmark does not appear to be guilty of any corner-cutting a la Boeing. I think many of us are aware of the FTC rules, and Benchmark probably should have called this out knowing its target markets, but it seems to me that the only use-mode affected by this limiting logic is testing like Amir does. Or am I wrong?

I do see your position, but I'm thinking that giving Benchmark a bit of slack here is reasonable.

 

[restorer-john]

John, I haven't been here very long, and typically I do enjoy your posts. But comparing AHB2's limiting logic to MCAS doesn't seem fair.

Consider this, and tell me a comparison is not fair.

John_S is clearly aware and worried about tweeter lead-in wires getting vaporized when amplifiers burst into a few cycle oscillation on or about the onset of overload or when the amplifier is run at high power, high frequencies for an indeterminate period not disclosed to the owners.
He speaks of tweeter protection. He's also acutely aware of how little time a typical tweeter can survive a high power, HF burst.
He speaks of clipping causing tweeter damage. We know clipping itself doesn't damage tweeters, too much power damages tweeters. Where that power comes from, and what frequencies that power is manifested is the issue. Some amplifiers do it well, others, not so much.
He speaks of transient front end input overloads from AP range switching relays (shifting the blame), and yet no other amplifiers seem to be upset by it in any of Amir's tests. How his amplifier clips as it runs into its tightly regulated SMPS rails is not documented. How does it behave in an overload situation? What is its recovery like?

We don't know, because there is a system that prevents the amplifier reaching its rated power for more than a few seconds. The owners are not in control, the FPGA protection system is and apparently, it knows best...

...but how would this limiter manifest itself in any real-world system playing music...

You have to believe the manufacturer when they say it cannot be triggered by music and would only activate in an overload or series of events the owner was not made aware of. Sound rather similar to MCAS if you ask me...

(using MCAS in this discussion this in no way diminishes the terrible loss of life, it is merely illustrative and I certainly hope no-one reading this was touched by loss)