Running Benchmark AHB2 in bridged mode and 4 Ohm Speaker

Power specified at 6ohm is 130W. Theoretical ideal power bridged at 6ohm should be 520W and is 480W (7.7% less). Also current was changed to 29A per channel. Perhaps 18A specification was for the early versions.

The misinformation in this thread from an individual who should know better is a little disturbing. The motive escapes me, but it appears to be deliberately malevolent.

From John Atkinson’s measurement comments:
"Commendably, the frequency response was not affected by the setting of the sensitivity switch, or by using the amplifier in bridged-mono mode. However, the ultrasonic response in bridged-mono mode shelved down faster than in stereo mode, and as the two channels’ output stages are in series in this mode, the output impedance, again including the 6’ of speaker cable, was higher than in stereo, at 0.14 ohm at 1kHz."

And his final comment:
"Benchmark Media Systems’ AHB2 is an extraordinary amplifier. Not only does its performance lie at the limits of what is possible for me to reliably test, it packs high power into a very small package, especially when used in bridged-mono mode. It is truly a high-resolution amplifier."
Read more at https://www.stereophile.com/content/benchmark-media-systems-ahb2-power-amplifier-measurements#LdQ7lf...

I suggest people interested in the AHB2 seek out out tests/opinions from reputable sources that have actual hands on experience with the amp. Technical queries should then be directed to Benchmark themselves to avoid speculation and misinformation.

Perhaps 18A specification was for the early versions. R

Yes of course they were, just like when Sterophile does a mediocre review, the manufacturers comments always say they must have sent a faulty one in for the review.
Wouldn’t you think at they would send the best possible tweaked one they could!!!!
What a load of ****

Cheers George

I am amazed this has not been answered and laid to rest. What seems to be the problem? The 29 amps will only be supplied by the amplifier if the load resistance is low and the voltage is high enough to produce that current. Perhaps its the order of things. Voltage first, then resistance then figure the current. I think we have all come to agree that bridging does not allow the amplifier to provide more than 29 amps. Bridging just allows twice the voltage which increases the OPPORTUNITY for more current to flow, but its still limited to 29 amps. Which, BTW, is not enough current form some ESLs played loud.

I promise you that all amps that are claimed to do 100 watts into 8 ohms, 200 into 4 and 400 watts into 2 ohms have quite a bit more power at 8 ohms than stated. Thats how they do it, Its specmanship and nothing wrong with. They do it for the same reason that some here want to see those numbers keep doubling. If you truly have only 100 watts at 8 ohms you will not be able to double down. Not Possible.

Conventional amps have Emitter or Source resistors whose value becomes significant at low impedance loads, the resistance of wire and traces startes to come into play. Mainly the sag is in the power transformer. 

When you look in an amp and see nice heavy wire going to the speaker keep in mind that there are several hundred feet of rather thin wire in the power transformer. 

ramtubes319 posts12-07-2018 4:24am
"The 29 amps will only be supplied by the amplifier if the load resistance is low and the voltage is high enough to produce that current. Perhaps its the order of things. Voltage first, then resistance then figure the current. I think we have all come to agree that bridging does not allow the amplifier to provide more than 29 amps. Bridging just allows twice the voltage which increases the OPPORTUNITY for more current to flow, but its still limited to 29 amps."

Agreed, good post.

Ramtubes, AHB2 has load/line regulated SMPS, but as you mentioned, there are losses everywhere.  Exact doubling of power for half of the load means perfect regulation - an ideal voltage source, that is only possible with very deep negative feedback.  Such feedback might create instabilities or TIM distortions, but  feedback in AHB2 is quite different - it is non recursive, meaning that signal is not fed back to the input of the same amplifier but to another "error amplifier" that corrects output later.  Such arrangement eliminates instabilities (allowing for wide bandwidth of the amp) and TIM distortions.  I also noticed a lot of low bass added , in comparison to my previous class D amp (that had 5Hz low bandwidth limit).  My guess is that all DC servos have instability problems for very low frequencies, but AHB2, having no recursive feedback at all can go down to 0.4Hz.  Should 0.4Hz vs 5Hz difference be audible with 30Hz speakers? - I don't know, but it is in my setup (big surprise).  The same goes for the upper limit - more extension and airy treble.  Could this be phase delay at high frequencies with my class D amp?  Speaker itself has inductive character at higher frequencies accelerating phase.  In addition there are many amps, including one  of Lamm, that have lower (50kHz) bandwidth and are praised for wonderful airy highs.

I had mentioned in my review that the AHB2 is not able to drive the Kingsound King III electrostatic to an unlimited level of listening. However, there is no degradation of sound as the amp is pushed to its limit, and there is no degradation associated with mono use.

I had mentioned in my review that the AHB2 is not able to drive the Kingsound King III electrostatic to an unlimited level of listening. However, there is no degradation of sound as the amp is pushed to its limit, and there is no degradation associated with mono use.

How did one stereo amp sound to a level it was comfortable at, and then the bridged at the same level? To gauge which sounded better. Not which could go louder!

Cheers George

@kijanki 

Ramtubes, AHB2 has load/line regulated SMPS, but as you mentioned, there are losses everywhere. Exact doubling of power for half of the load means perfect regulation - an ideal voltage source, that is only possible with very deep negative feedback

Lets just take this first part. Feedback is not involved in doubling of power, deep or not. The losses I am speaking of are mostly in emitter resistors and turn on resistances of the output devices. Once an amp clips feedback cannot make up for these losses. 

I had mentioned in my review that the AHB2 is not able to drive the Kingsound King III electrostatic to an unlimited level of listening. However, there is no degradation of sound as the amp is pushed to its limit, and there is no degradation associated with mono use.

How did one stereo amp sound to a level it was comfortable at, and then the bridged at the same level? To gauge which sounded better. Not which could go louder!

George is correct and one amp might sound better. Im curious what the owner of these speakers considers unlimited level?  Lets watch the words.

BTW, ESLs exhibit clipping very similar to amplifier clipping. This occurrs when the air gap saturates, and I assure you it does. One can often smell the ozone close up to the speaker.... im not kidding.

Once an amp clips feedback cannot make up for these losses.

I agree, but amp that doubles power exactly has to have adequate power supply AND deep negative feedback.  Exact doubling implies perfect regulation (ideal voltage source), that is not possible without deep feedback.

@kijanki  I agree, but amp that doubles power exactly has to have adequate power supply AND deep negative feedback. Exact doubling implies perfect regulation (ideal voltage source), that is not possible without deep feedback.

What part do you agree with? If you do agree that feedback does not make up for these losses why do you keep bringing up deep feedback. (also a term I have never heard)

1. What about the emitter resistors? You cannot ignore them. 

2. How good is the supply regulation? 1% ?

3. What does the deep feedback have to do with it? If the damping is even 10 thats enough. Its only going to affect the shape of the clipping and make it worse not better.

If for some silly reason you want the power to exactly double then you make a power supply that puts out more voltage as more current is drawn. Kinda upside down idea. 

If you can site an amplifier that exactly doubles bring it on. Show me.

Ramtubes, I explained exactly what I agreed with by quoting your statement - read again.

Why are you bringing emitter resistors again, yes we know that they exist. 

I'm not sure why you're asking me what Benchmark voltage regulation is.   Perhaps you should direct this question to John Siau.

What does the deep feedback have to do with it?

Feedback lowers output impedance.  You cannot make exact doubling of the power if DF=10.

If you can site an amplifier that exactly doubles bring it on. Show me.

There are many of them (I'm surprised you don't know that).  For instance Rowland model 535 is specified for 250W@8ohm and 500W@4ohm.  Krell Evolution 302e specifies 300W@8ohm 600W@4ohm and 1200W@2ohm (both channels driven).  Pass Labs XS150 states 150W@8ohm 300W@4ohm  etc.

You keep asking "what feedback has to do with it" while I'm trying to show you, that exact doubling of power cannot be achieved without deep feedback.  Exact doubling of the power is treated on this forum as a virtue, while I don't think it is necessary a good thing and might be even harmful when deep feedback is used to spike-up amplifier's  specifications while creating TIM distortions at the same time.

@kijanki  Ramtubes, I explained exactly what I agreed with by quoting your statement - read again.

Why are you bringing emitter resistors again, yes we know that they exist.  

I'm not sure why you're asking me what Benchmark voltage regulation is.  Perhaps you should direct this question to John Siau.

What does the deep feedback have to do with it?

Feedback lowers output impedance. You cannot make exact doubling of the power if DF=10.

If you can site an amplifier that exactly doubles bring it on. Show me.

There are many of them (I'm surprised you don't know that). For instance Rowland model 535 is specified for 250W@8ohm and 500W@4ohm.  Krell Evolution 302e specifies 300W@8ohm 600W@4ohm and 1200W@2ohm (both channels driven). Pass Labs XS150 states 150W@8ohm 300W@4ohm  etc.

You keep asking "what feedback has to do with it" while I'm trying to show you, that exact doubling of power cannot be achieved without deep feedback. Exact doubling of the power is treated on this forum as a virtue, while I don't think it is necessary a good thing and might be even harmful when deep feedback is used to spike-up amplifier's specifications while creating TIM distortions at the same time.

I know of all those amps and those specs. But those are marketing numbers. Can you tell me the actual measured clipping power at those various loads. 

I keep asking about the emitter resistors because when you drop the load impedance and more current is drawn the drop across those resistors (and any other resistances) doubles and it appears you dont want to take that into account. 

I have not been asking or referring to the Benchmark in this discussion. 

Exact doubling of the power is treated on this forum as a virtue

,

I dont see why it is a virtue especially if it is just a spec and not what is really going on. Has Stereophile tested any of these amps. Please provide a link. 

Yes, drop across resistance will double, but reserves in power supply combined with the feedback will make up for that.

I dont see why it is a virtue especially if it is just a spec and not what is really going on. Has Stereophile tested any of these amps. Please provide a link.

I never said it is a virtue - just the opposite.

@kijanki 

Heres a Rowland that specs 60 into 8ohms,  120 into 4 ohms just the way you like it. The actual measured powers are "  Fig.7 indicates the THD vs power for 8, 4, and 2 ohms at 1kHz, one channel driven (bottom, middle, and top curves respectively; line voltage of 113V). The maximum output power at 1% THD was 84W (19.2dBW) into 8 ohms, 140W (18.5dBW) into 4 ohms"
Read more at https://www.stereophile.com/content/jeff-rowland-design-group-model-1-power-amplifier-measurements#D...

The spec is twice the power into 4/8 ohms is just as you like 120/60 watts. The measured is 140/84 watts, not double. Note as I predicted the 8 ohm power is understated as it must be. But 120/60 watts are just minimum specs, not what the amp CAN do. Good manufacturers always underspec to make sure they make spec. 

Here is a Krell speced very close to 2/1 power into 4/8.  However the actual ratio is 910/625 watts. hardly 2/1. 


 Although the Krell's specified maximum power into these loads is very high, at a respective 575W (27.6dBW) and 900W (26.5dBW), the amplifier exceeded those powers, clipping at 625W into 8 ohms (28dBW) and 910W into 4 ohms
Read more at https://www.stereophile.com/content/krell-solo-575-monoblock-power-amplifier-measurements#17OE2KQdwm...

Cant you see that the doubling down is just a spec and to make that spec they have to go way over on the 8 ohm to get the 4 ohm to work out. The Krell barely makes spec with only 10 watts to spare on 4 ohms but exceeds 8 ohm power by 50 watts. 

I have a friend that loves to see amps double down but none actually do. Its nice when they get close but its just impossible to be exact or just close. 

Where does this religion of doubling down get its followers?

I will clarify; the bridged AHB2 drives the Kingsound King III ESL much better than a stereo unit alone, or two stereo units in passive biamp mode. 

As the discussion of the thread centers around bridged operation, I don't believe I have stated the capacity of a stereo unit on this forum to drive the Kingsound King III.

When I refer to "level unlimited" I mean that the amp is at maximum output of the preamplifier turned full on and the speakers not hitting past perhaps 87-88 dB. I would call an amp "level unlimited" where in my use I would not turn the level up more, but the amp could certainly provide a higher listening level. It may not be a technically correct description, but it serves to indicate the situation. I do not attempt to listen to systems at "live", i.e. above 90 dB levels, and I do not recommend others to do so. 

I have said several times previously that there is NO sonic degradation in bridged mode, nor distortion associated with running the amp full out. The amp is FAR superior sounding in bridged mode regardless of the speakers used, efficient or less efficient. 

 The AHB2 was never even close to clipping. Relevant to that, summarizing the discussion from the Owner's Manual, "... the AHB2 has an Over-Temperature Fault circuit that monitors the heat sinks constantly. If either one reaches the max allowable temperature the channels are muted to protect the amp and speakers. All high-power systems are shut down to facilitate cooling... One or both heat sinks will be very hot.. power reset after cooling." (Again, summary) 

The amps were actually running cool with the Kingsound King ESL, I could keep my hand on them. The amp also has a "Clip" light indicator, which never, ever came on. The King III is nominal 6 Ohm and 83 dB sensitivity. 




 

@kijanki  Yes, drop across resistance will double, but reserves in power supply combined with the feedback will make up for that.

I dont see why it is a virtue especially if it is just a spec and not what is really going on. Has Stereophile tested any of these amps. Please provide a link.

I never said it is a virtue - just the opposite.

I, prehaps misinterpreted this statement as something you agreed with.. 

 

Exact doubling of the power is treated on this forum as a virtue.

 Under RMS testing (which I have responded to in your other thread) there is no power suppy reserve, its flat out. 

Please tell me how feedback is going to find power where there is no more. Feedback does not create power. Rather than keep telling me it does tell me HOW it does.

@douglas_schroeder    When I refer to "level unlimited" I mean that the amp is at maximum output of the preamplifier turned full on and the speakers not hitting past perhaps 87-88 dB. I would call an amp "level unlimited" where in my use I would not turn the level up more, but the amp could certainly provide a higher listening level. It may not be a technically correct description, but it serves to indicate the situation. I do not attempt to listen to systems at "live", i.e. above 90 dB levels, and I do not recommend others to do so

It appears you may be confusing gain with power as many do.

I mean that the amp is at maximum output of the preamplifier turned full on and the speakers not hitting past perhaps 87-88 dB.

This is the statement I do not understand. The amp is doing what?

The top setting on a preamp does not conform to any percentage of the power amplifier's potential output. Some systems max out on power at 9 on the clock. some at 12 some not even at full up, which is a good situation where the 9 on the clock is very bad, though I hear some like it as they think there headroom above that setting, where there is not. 

90 db is a nice listening level, however do you sometimes find a CD might not get to 90 at full preamp level? And of course some will be too loud.

ramtubes
It appears you may be confusing gain with power as many do.

He is, by me too.

douglas_schroeder

The amp is FAR superior sounding in bridged mode regardless of the speakers used, efficient or less efficient.

I ask again, then why sell it in stereo mode if it’s so hobbled in stereo mode???
You should then be suggesting that no one buy it for stereo mode listening there are far better offerings.

douglas_schroeder
I have said several times previously that there is NO sonic degradation in bridged mode, nor distortion associated with running the amp full out.

This is a furphy, I didn’t ask that half a page back, I asked a simple question of amp sound quality driving the Kingsound III it was ignored. I present it again.

"How did one stereo amp sound to a level it was totally comfortable at.
Then the bridged ones at the same level?"
"To gauge which sounded better."
"NOT WHICH COULD GO LOUDER!"

I think it all needs "paralleled interconnects attached to each channel", that should fix the problem/s.


Cheers George

@georgehifi 

I think Benchmark has done an excellent job of convincing everyone to buy two. Hail to marketing. All those who follow can do is quote the party line. We will never get them to understand. 

ramtubes
I think Benchmark has done an excellent job of convincing everyone to buy two.

Got a feeling there’s a couple of latent shills helping out too.

Cheers George

@georgehifi  Got a feeling there’s a couple of latent shills helping out too

Likely indeed. The marketing guy did step in at one point. We really gave it a good shot.

Check out this thread claiming there is no RMS power.

https://forum.audiogon.com/discussions/rms-power#1665330

 Read the banner at the top of the Wicki page and check out the references given below. especially #8. Then read all the compliants on the talk page. Im writing someting up.

https://en.wikipedia.org/wiki/Audio_power  

@kijanki 

to me

From Wikipedia "Audio Power"

As described above, the term average power refers to the average value of the instantaneous power waveform over time. As this is typically derived from the root mean square (RMS) of the sine wave voltage,[6] it is often referred to as "RMS power" or "watts RMS", but this is incorrect: it is not the RMS value of the power waveform (which would be a larger, but meaningless, number).[7][8][9][10] (The erroneous term "watts RMS" is actually used in CE regulations.[11]) This is also referred to as the nominal value, there being a regulatory requirement to use it.

Al, leave it (he is not going to get it). . 

@kijanki  Feedback lowers output impedance. You cannot make exact doubling of the power if DF=10.

Of course feedback lowers output impedance. However that has nothing do to with output power at clipping. Its not like there is some resistor in there.

I assume you are adjusting the input level as you explore these power output levels. 

What is this EXACT doubling of power all about.. Why does it have to be exact? 

I think, we are talking about two different things.  I have never mentioned actual amplifier's power at clipping but rather manufacturer's specifications (that are quite different), that calls for power doubling (or close to it). 

I have also noticed personal comments from you, placing my math ability in question and calling me to learn basic electronics.  I believe that I have some comprehension of basic electronics designing it for 40 years and therefore I view your comments as harmful, therefore refusing to comment on anything you have to say from this point on.  Perhaps one day, when you change your attitude we can start talking again.

@kijanki I think, we are talking about two different things. I have never mentioned actual amplifier’s power at clipping but rather manufacturer’s specifications (that are quite different), that calls for power doubling (or close to it). I have also noticed personal comments from you, placing my math ability in question and calling me to learn basic electronics. I believe that I have some comprehension of basic electronics designing it for 40 years and therefore I view your comments as harmful, therefore refusing to comment on anything you have to say from this point on. Perhaps one day, when you change your attitude we can start talking again

.
If we are talking about two different things what are they? Im just talking about simple RMS power. I try to make my comments as non personal as possible. I will go back and have a look at them and apologize appropriately. Im not here to pick a fight. Lets just get the information right. :)

I still dont see why a manfacturer calling for power doubing on a piece of paper has anything do to with what the amp does. What spec makes it different than clippiing. Most power amps have very low distortion till they clip.

Declining to answer simple questions or support statements does us no good in a forum. I am here trying to correct things. If I am incorrect then bring me the proof. This is really simple stuff.

Im sure you math skills are fine. This is about a concept. Starting a post with "There is no such thing as RMS power." Is rather bold. We measure it all the time. If one wants to be contrary to accepted terms he had better have some good material to back it up. This I have not seen.

I was fine having a discussion until you quoted a very poor Wickipedia article and.. did you have to add to AL ???

Al, leave it (he is not going to get it). .

I politely objected to that. Not going to get it? I also offered to apologize if I was incorrect.

Why not build a few amplifiers to see if your theories are correct and dont forget the emitter resistors..  :)

Ramtubes, PEACE!  I'm sorry for my comments.  It was uncalled for.  I have short fuse (electrical term :) ) recently, since I completely lost hearing in the right ear few months ago (replaced by loud white noise - so much for Benchmark's fantastic S/N ratio).  I will probably stay for a while, learning and possibly helping, but I slowly loose interest in this hobby.  Am I the perfect candidate for bridging AHB2 (only one channel needed)?  Again peace, man.

I don't understand the thinking that Benchmark is somehow duping people into buying two of theirs amps. I've never seen marketing for one amp (think it's obviously out there), let alone two. 

I purchased my second amp because of my, and others experience with moving from one amp to two. 

I never felt that the sound was in any way lacking when I was running one, though I wanted to see if I could get better sound by running two. And I did. 

@lordcloud  If you read the reviews of any Benchmark product in Stereophile, Absolute Sound, or elsewhere, and such reviews are quite frequent, you will find that Benchmark is always sending out *two* of its amps to the reviewer.

Kijanki  12-11-2018
I have short fuse (electrical term :) ) recently, since I completely lost hearing in the right ear few months ago (replaced by loud white noise - so much for Benchmark's fantastic S/N ratio). 

Ouch!  I'm of course very sorry to hear this news, Kijanki.  And of course I hope it will improve in time.  All the best,

-- Al 

Thank you Al, but I lost all hope - it is almost 6 months now and no improvement.  All audio tests show -95dB.  My doctor says it was most likely a virus from the shingles variety.  After 15 sessions in hiperbaric chamber and two direct injections of steroids thru the eardrum my eyes suddenly improved and  I'm typing this without glasses (+2 before).  Doctors have no clue why.

That’s so scary Kijanki, I too hope it comes back.

Love all your knowledge that you all have been speaking to. When building my system I considered more Benchmark amps and read a lot of discussion similar to your thread so I bought two and thought I could always get four if needed. I had completed the system in a tiny place in Cali and moved to a huge space now and got the system set up the way it should be. When showing off to friends and the occasional maintenance guy I would get my Maggie’s to clip and could hear it happening if I pushed them. So here I am thinking I may need more power. I would also get my Pass X-250 (going to my main subs) to bury the bias needle in large transients and movies especially. I have been so crazy with work I couldn’t get everything set up right. After my system was hit my lightening and bought a P15 to protect it the Pass would pull too much power and shut down the P15. I started plugging the amp into the wall directly at each listening session as everything was before the lightening and pulling it out afterwards. . .

So about out three weeks ago I was able to sit down and do room correction with REW and MSO adding in my SVS subs. Which by the way made the house and windows shake when I turned them on so bad that I literally jumped off the couch because they scared my soul, the water on my table had the ripples, and my wife jumped out of the shower upstairs yelling that lightening/thunder just shook the house and I needed to shut down the system and unplug it because lightening was in the area. Anyway I started with 4 and ended up getting to 12 biquad filters going in my Mini DSP and with some playing around got things to sound amazing. I noticed now with the room correction how much more accurate the music was and how little the subs had to work to make the transients that sent everything clipping before, now playing at -15 for movies and -20 for music.

Now I don’t come close to clipping playing louder than I would ever listen but do in demo mode on any amps + have everything running off of my P15. The power needle barely moves. . . With all the speakers running full range (obviously MSO attenuated the subs naturally). Who would have guessed it. FYI you barely hear the system upstairs now with all the subs working together even though everything is more impactful.

So what I’m saying is if your having clipping issues look at your room nodes, buy four quality subs, play with some room correction and placement of the subs and the power argument will become a moot point. I’m happy to answer any questions if people want to try this path.

Kijanki FWIW you are one of the no nonsense technical  guys on this site whose insights I value most.   Keep it up!

Thank you hifidream, mapman

There are many good comments in this thread but there is also much erroneous information. In most cases, this erroneous information is based on the assumption that the AHB2 behaves like most power amplifiers. The statements made would be true of most power amplifiers, but do not apply to the AHB2.

The AHB2 does not act like a traditional power amplifier because it has a very unique design. It has tightly regulated power supply rails, a patented feed-forward error correction system, and an offline protection system that is not in the signal path.

Unlike virtually all other power amplifiers, the THD produced by the AHB2 does not increase under load. The 8-ohm and 4-ohm THD vs. Power curves are absolutely identical. Even when loaded with 2 ohms, the THD is virtually unchanged. To the best of my knowledge, no other power amplifier can do this, or even come close to doing this. And, no amplifier delivers lower THD or lower noise at any price. The AHB2 holds the record for lowest THD and lowest noise. It is also important to understand that this performance is not just achieved at one particular load impedance, test frequency, or operating level. It is achieved into any rated load at any power up to full output power, and any test frequency between 20 Hz and 20 kHz.

The red clip lights on the AHB2 are driven by a circuit that measures the THD. If it exceeds 1% the lights will turn on. The AHB2 delivers an astonishing 0.00011% THD, 1 kHz at full rated power into any rated load.  The transition from 0.00011% to 1% is abrupt, but it occurs above the rated output power. This means that the AHB2 stays clean, even when you drive it almost to the rails. Ordinary amplifiers cannot do this. With most amplifiers, THD gradually increases with output level and with increasing current load. With the AHB2, THD is not a function of output current.

The impedance load ratings for the AHB2 are based upon long-term thermal limits (prolonged operation into the specified load impedance). This is not a limitation on the allowable minimum impedance of the loudspeaker. It is a rating for the nominal impedance of the loudspeaker. If the impedance drops to 1 or 2 ohms at certain frequencies or is presented with a difficult phase angle, this is not a problem for the AHB2. Feed-forward correction is stable into any load. Feedback is not stable into any load. The AHB2 uses a large amount of feed-forward correction and a minimal amount of feedback. This allows stability into virtually any load. Match your speakers to the AHB2 on the basis of nominal impedance not minimum impedance. The impedance minimums are not a concern.