Review and Measurements of Benchmark AHB2 Amp

[tmtomh]

So what? the recording of both A and B track is using the same microphone and listening using the same headphone . Both A and B should sound the same. Are they same or they aren't - that answer is the key to decision making and drawing conclusion .

I am skeptical that any differences in amplifier sound will be large enough to survive the effects of the speakers and mic that sit between the original output and your headphone listening.

But I am happy to grant you that point for the sake of discussion: let's say that if the amps do sound different, you and your friends will be able to detect that from these recordings you make.

The problem I see is that if you all do hear differences, that's not the same as a preference. And if one amp sounds subjectively better to you than the other one on the recording, how do you know that amp will sound better to you in long-term, day-to-day live listening with the original speakers? For example, if the recordings reveal that one amp has slightly more bass slam, is that because that amp is more linear in the bass region than the other one, or is it because there's a small loss of energy at certain bass frequencies in the speaker-mic-headphone chain used to make and listen to the recordings, and the other, seemingly bass-deficient amp is actually the more linear one?

 

[Kal Rubinson]

Does anyone tried the benchmark speaker cables (speakon) ? Just want to know your findings about them.

I have but, for appearances sake, I use Canare 4S11 SpeakOn cables from BlueJeans. Either one is fine and there's no reason to change.

 

[Dgob]

No one asked - scroll up. I said what I have plans to do and asked one question - about the benchmark speaker cables and speakon. Not sure why people has to respond everything else but not to what was asked and consume ASR bandwidth , storage and lengthy threads

I loved the Speakon cables with the F208 and a pair of AHB2's. It was truly magical, but apart from the fact that this approach seems to be favoured amongst professional producers, I am not certain how important a role the cables played in that performance.

Good luck

 

[misterdog]

Does anyone tried the benchmark speaker cables (speakon) ? Just want to know your findings about them.

I'm now using AHB2 (stereo) with Benchmark/Speakon cables.

Previously I used Hypex NC400 mono's then moved to Neurochrome Mod. mono's.
Though I'm feeding tricked out Quad 989 ELS's so don't need the power of two AHB2's (so I tell myself )

So I will be interested in your findings, though note which post John Siau (director of engineering at Benchmark) 'liked' with regard your comparison regime (page 176).

 

[sandeep_s300]

I'm now using AHB2 (stereo) with Benchmark/Speakon cables.

Previously I used Hypex NC400 mono's then moved to Neurochrome Mod. mono's.
Though I'm feeding tricked out Quad 989 ELS's so don't need the power of two AHB2's (so I tell myself )

So I will be interested in your findings, though note which post John Siau (director of engineering at Benchmark) 'liked' with regard your comparison regime (page 176).

Send me PM.

 

[Platypus20]

I use the Benchmark wires with Speak-On/locking bananas on the 2 big stereos, work great

 

[dlaloum]

I'm now using AHB2 (stereo) with Benchmark/Speakon cables.

Previously I used Hypex NC400 mono's then moved to Neurochrome Mod. mono's.
Though I'm feeding tricked out Quad 989 ELS's so don't need the power of two AHB2's (so I tell myself )

So I will be interested in your findings, though note which post John Siau (director of engineering at Benchmark) 'liked' with regard your comparison regime (page 176).

Dual AHB2's won't do anything good for a pair of 989's... just increases the likelihood of ARC'ing damage (although the protection circuits are decent on the later generation Quads... but still!) - you will reach their max SPL's / compression point with a single stereo AHB2.

I still miss my 989's!

 

[John_Siau]

Dual AHB2's won't do anything good for a pair of 989's... just increases the likelihood of ARC'ing damage (although the protection circuits are decent on the later generation Quads... but still!) - you will reach their max SPL's / compression point with a single stereo AHB2.

I still miss my 989's!

The AHB2 in stereo mode is a perfect match for the 989's. The clip point of the AHB2 is exactly +/- 40 V peak (80 Vpp or 28.28 Vrms). This predictable clip point is due to the regulated supply rails and it makes the maximum output voltage predictable under all load impedances, audio wave shapes, and line voltage conditions. With a conventional amplifier, the peak output voltage will vary according to the AC line voltage, the audio wave shape, and the load impedance. The predictable peak output of the AHB2 completely eliminates the risk of arcing damage.

We have many customers that use the AHB2 (in stereo mode) with the Quad ESL-989's.

The Quads should not be driven by an AHB2 running in bridged mono mode! Bridged mono mode will deliver +/- 80 volt peaks and this will damage the Quads.

 

[John_Siau]

I use the Benchmark wires with Speak-On/locking bananas on the 2 big stereos, work great

We recommend the locking bananas on the speaker end and Speak-On connectors on the amplifier end. In our tests, spring loaded bananas and spade lugs often create more measurable distortion than the entire AHB2 amplifier. With most amplifiers these effects cannot be measured because the amplifier distortion is so much higher than that of the connectors. In contrast, the AHB2 is clean enough to expose the slight differences in connector performance. These slight measurable differences in connectors should be inaudible unless the connections are loose.

Spades can provide a good connection if they are clean and sufficiently tight. The main problem is that they tend to loosen as soon as you bump the cable. The locking banana plugs eliminate this problem. Ideally, I would like to see more speakers with Speak_On connectors.

The measurable differences in connectors tend to be much larger than the measurable differences in the actual speaker cables.

As for the speaker cables themselves, cable differences are barely measurable. Worst case, you may see a 0.1 dB variation in frequency response at 20 kHz between different cable types (assuming a 10-foot cable length).

Don't spend a fortune on cables. Cables should be selected on the basis of durability, quality of the connectors, and the desired look and feel. Don't expect to hear audible differences unless the cables are very long (over 100 feet long).

Buy acoustic treatments for your room, upgrade you speakers, amplifier, preamplifier, and DAC before considering a cable upgrade.

See my latest audio application note on this subject: "Audiophile Snake Oil"

 

[Reddoc]

We recommend the locking bananas on the speaker end and Speak-On connectors on the amplifier end. In our tests, spring loaded bananas and spade lugs often create more measurable distortion than the entire AHB2 amplifier. With most amplifiers these effects cannot be measured because the amplifier distortion is so much higher than that of the connectors. In contrast, the AHB2 is clean enough to expose the slight differences in connector performance. These slight measurable differences in connectors should be inaudible unless the connections are loose.

Spades can provide a good connection if they are clean and sufficiently tight. The main problem is that they tend to loosen as soon as you bump the cable. The locking banana plugs eliminate this problem. Ideally, I would like to see more speakers with Speak_On connectors.

The measurable differences in connectors tend to be much larger than the measurable differences in the actual speaker cables.

As for the speaker cables themselves, cable differences are barely measurable. Worst case, you may see a 0.1 dB variation in frequency response at 20 kHz between different cable types (assuming a 10-foot cable length).

Don't spend a fortune on cables. Cables should be selected on the basis of durability, quality of the connectors, and the desired look and feel. Don't expect to hear audible differences unless the cables are very long (over 100 feet long).

Buy acoustic treatments for your room, upgrade you speakers, amplifier, preamplifier, and DAC before considering a cable upgrade.

See my latest audio application note on this subject: "Audiophile Snake Oil"

Great post, John, terrific to have your contributions.
"spring loaded bananas and spade lugs often create more measurable distortion than the entire AHB2 amplifier"- makes one think, doesn't it

 

[radix]

Review and Measurements of Benchmark AHB2 Amp

The AHB2 has now been tested in the German trade magazine STEREO 8/2022 and has scored quite well. However, the performance data in bridge mode could not be reproduced. Unfortunately, there are no further details on this. Here is the measured DF at 4ohms (not bridged), which I have not seen...

audiosciencereview.com

JSmith

I realize @John_Siau is the authority on the AHB2, but I don't know where the "class G" came from. The Benchmark website says "A unique output stage incorporates class-H tracking power supply rails to improve efficiency." And again "The feed-forward error correction also allows the use of class-H tracking power supply rails without the usual distortion-performance penalty. These tracking rails significantly improve the overall efficiency of the amplifier while the error correction keeps distortion nearly nonexistent."

I'm pretty sure it's a class H tracking rail, not a class G switching rail. Otherwise, it would be the AGB2, no?

 

[Martini]

I realize @John_Siau is the authority on the AHB2, but I don't know where the "class G" came from. The Benchmark website says "A unique output stage incorporates class-H tracking power supply rails to improve efficiency." And again "The feed-forward error correction also allows the use of class-H tracking power supply rails without the usual distortion-performance penalty. These tracking rails significantly improve the overall efficiency of the amplifier while the error correction keeps distortion nearly nonexistent."

I'm pretty sure it's a class H tracking rail, not a class G switching rail. Otherwise, it would be the AGB2, no?

Named after Benchmark's founder, Allen H. Burdick.

"Our AHB2 power amplifier is a product that would have made Allen proud. It has the lowest distortion and lowest noise of any audio power amplifier at any price. Like many of our current products, the AHB2 was designed after Allen's retirement. Due to health problems, Allen was unable to participate in the design of any products released after 2007. Nevertheless, each new product closely follows Allen's vision for building transparent audio products.

Just hours before we announced our new power amplifier, we received the sad news that Allen had passed away. In honor of Allen H. Burdick, the AHB2 power amplifier bears his initials."

 

[radix]

Named after Benchmark's founder, Allen H. Burdick.

"Our AHB2 power amplifier is a product that would have made Allen proud. It has the lowest distortion and lowest noise of any audio power amplifier at any price. Like many of our current products, the AHB2 was designed after Allen's retirement. Due to health problems, Allen was unable to participate in the design of any products released after 2007. Nevertheless, each new product closely follows Allen's vision for building transparent audio products.

Just hours before we announced our new power amplifier, we received the sad news that Allen had passed away. In honor of Allen H. Burdick, the AHB2 power amplifier bears his initials."

That's interesting, didn't know that about Burdick. I always thought the H was from the class H rails.

 

[DonH56]

From the Benchmark website:

A.H.B.​

The letters A, H, and B allude to the unique topology of the amplifier's output stage. The AHB2 combines class-AB, and class-H topologies using a feed-forward error correction system.

But there is much more to the story -

The initials A.H.B. can be found on many of the circuit diagrams of Benchmark's early products dating back as far as 1983. These are the initials of Benchmark's founder, Allen H. Burdick. He personally designed many of Benchmark's early products.

Allen had a life-long passion for audio, and he became one of the leading innovators in the pro-audio industry. The AHB2 is named in his memory.

And this:

The AHB2 Name​

On September 27, 2013, everything was ready, but we still didn't have a name for the new amplifier. The faceplates were machined and finished but needed to be printed with the product name. I began thinking about the technology in the amplifier and how this could be incorporated into the name. The amplifier delivered class-A performance, had class-H tracking rails, and a class-AB output stage. It then struck me that the letters A, H, and B were Allen H. Burdick's initials. Instantly it was clear to me that the new amplifier should be named after the man who inspired it! At that moment, the PA2 became the AHB2. I announced the name to my staff, and we placed the order for the printing. Less than an hour later we got a phone call with the sad and shocking news that Allen had passed away. I am still stunned when I think about the coincidence of these two events.

Classes G and H are switched from USA/IEEE to some European designations. To me, class G uses power rails that are switched among two or more fixed voltages, while class H power rails track the input voltage so continuously vary. Benchmark states in several places that the AHB2 uses a tracking (regulated switch-mode) power supply.

 

[dlaloum]

I realize @John_Siau is the authority on the AHB2, but I don't know where the "class G" came from. The Benchmark website says "A unique output stage incorporates class-H tracking power supply rails to improve efficiency." And again "The feed-forward error correction also allows the use of class-H tracking power supply rails without the usual distortion-performance penalty. These tracking rails significantly improve the overall efficiency of the amplifier while the error correction keeps distortion nearly nonexistent."

I'm pretty sure it's a class H tracking rail, not a class G switching rail. Otherwise, it would be the AGB2, no?

I believe the definitions for class H and class G are effectively reversed depending on which side of the atlantic is using the terms.... which leads to a lot of confusion.

 

[tmtomh]

I believe the definitions for class H and class G are effectively reversed depending on which side of the atlantic is using the terms.... which leads to a lot of confusion.

Yes, I came here to say this as well. If Wikipedia is to be believed on this subject (and my reading around the web has turned up no evidence to suggest it isn't), the definitions of Class G and H are reversed in North America and Europe.

 

[Platypus20]

AHB2 is named after the founders initials - Allen H. Burdick

 

[DonH56]

AHB2 is named after the founders initials - Allen H. Burdick

Started as a nod to the operating class, then tweaked to honor Burdick.

Review and Measurements of Benchmark AHB2 Amp

So what? the recording of both A and B track is using the same microphone and listening using the same headphone . Both A and B should sound the same. Are they same or they aren't - that answer is the key to decision making and drawing conclusion . I am skeptical that any differences in...

www.audiosciencereview.com

 

[John_Siau]

From the Benchmark website:

And this:


Classes G and H are switched from USA/IEEE to some European designations. To me, class G uses power rails that are switched among two or more fixed voltages, while class H power rails track the input voltage so continuously vary. Benchmark states in several places that the AHB2 uses a tracking (regulated switch-mode) power supply.

The switched mode power supply is not tracking. It provides two sets of fixed voltage rails that are tightly regulated. The "tracking rails" are derived from the upper fixed rails using transistors that follow the analog waveform (with a few volts of headroom). This topology is cleaner than directly using the upper fixed rail. Efficiency is the same either way, but our approach does require more output transistors on the upper rails. Our analysis showed that the extra components reduce distortion (before applying feedback and feedforward correction).

In comparison to a class AB, overall efficiency is gained by transitioning between the lower and upper rails whenever required by the audio waveform. Furthermore, the switch-mode power supply is much more efficient than a linear supply (especially if you try to regulate the linear supply). In typical class AB designs, the output stage is the voltage regulator.

Obviously it would be possible to place a linear class AB inside tracking rails that are driven by a class D, but that was not our approach. We chose our topology because it makes it easier to obtain very high performance, and it avoids the delay mismatch between the class D and the class AB.

 

[DonH56]

The switched mode power supply is not tracking. It provides two sets of fixed voltage rails that are tightly regulated. The "tracking rails" are derived from the upper fixed rails using transistors that follow the analog waveform (with a few volts of headroom). This topology is cleaner than directly using the upper fixed rail. Efficiency is the same either way, but our approach does require more output transistors on the upper rails. Our analysis showed that the extra components reduce distortion (before applying feedback and feedforward correction).

In comparison to a class AB, overall efficiency is gained by transitioning between the lower and upper rails whenever required by the audio waveform. Furthermore, the switch-mode power supply is much more efficient than a linear supply (especially if you try to regulate the linear supply). In typical class AB designs, the output stage is the voltage regulator.

Obviously it would be possible to place a linear class AB inside tracking rails that are driven by a class D, but that was not our approach. We chose our topology because it makes it easier to obtain very high performance, and it avoids the delay mismatch between the class D and the class AB.

Thanks, I thought I had read that someplace here, but could not find the reference. I had in mind that it switched between two rails, but also that it was truly class H, but obviously did not remember the details.

A lot of designs (not necessarily audio amplifiers) use a linear tracking rail derived from a tightly-regulated SMPS rail to improve performance by controlling output stage transistors' Vce/Vds/etc. to improve linearity and reduce thermal effects. My experience was in essentially pulse systems where high bandwidth coupled with high short- and long-term stability was required to ensure pulse integrity. It was also flight hardware, where size, weight, and power were critical, so SMPS' were preferred where practical.

This was long ago, but at the time a tracking SMPS was extremely difficult to design due to numerous issues (loop stability and bandwidth with low enough lag for rapid tracking, charge pumping during transitions, etc.) The SMPS was a bigger design effort than the amplifier itself, and limited its performance. One thing we found, which may not apply to the AHB2, is that using feedforward with feedback was critical to achieving the lowest distortion, but the FF circuit was also more sensitive to power supply noise and artifacts. Given a tightly-regulated clean SMPS rail, designing an analog tracking rail was trivial in comparison. ("In comparison" is important; it was still not a trivial design, natch.)

Thanks John! - Don