Class H power amps (moving air-speaker cone response time-amp transient response?)
 

I'm always trying to learn. At least one new thing each day (and that does not include what talking heads on various cable news channels spew). Here's one that grabbed my attention .... Class H power amps.

I've been told that I'm one of the misinformed that believes it takes a certain amount of time for a speaker cone to go from a resting position to full throw distance (or any amount of distance for that matter), and that time is many instances not the same as the frequency it is producing. Said another way, if a speaker is asked to produce a 100hz sound, it may take longer than 1/2 of that (a half cycle) to reach the excursion that is demanded of it. The first few cycles will be 100hz, but they may not be the entire amount of volume that is demanded of it right away. Mass of the moving parts, and other factors come into play in this ~misinformed~ idea.

(Then again, if the parts that move air don't get up to full distance at a given freq right away, why would they ever get up to full throw if the idea holds any water? I suppose it only holds any validity on the first half cycle .. and can we .. humans with ears and the limits those ears/brains have ... even perceive the lack of volume in a half cycle of frequencies above -- let's say -- 120 hz or so? If a ~ahem~ "slower" speaker can't get up to full-excursion at a given freq within the first half cycle .. would it EVER? It has to get all the way out to the demanded distance - stop -go the other direction twice as far as the first half cycle - then stop - go that whole distance again - stop - go that whole distance again - and so on. So if it can't get to just half the distance in the first half cycle, then what would make it possible to go the ENTIRE distance in a full cycle? Hmmmm .... I'm starting to see how that whole idea of "speaker speed" may be a bit bunk. Perhaps it IS misinformation after all. Then again, there are certain attack transients that exceed the average volume of a given note - the instant a string is slappa'd, it creates a sound that is louder than the rest of the note for a split second, so maybe THAT is what the "speaker speed" idea addresses. The speaker cone must travel further at the very instant it is asked to make sound than it does for the bulk of the sound, and it has to travel further in the same amount of time - a half cycle - than it does for the rest of the sound. At 120hz it has to travel - let's say - 2 inches in 1/60th of a second for the attack sound - but only maybe 1 inch in 1/60th of a second for the remainder of the sound. So perhaps that is what the speaker-speed thing is aimed at. Ok .. enough thinking out loud here with this off-topic babbling .... this is maddening ... my hair hurts).

Anyhow - returning to topic here .....

Well, I just read some stuff about a class of amplifier new to me (class H) that addresses some of the factors involved in getting air to move as closely to the same time as the signal is produced by the instrument. Said another way, from the instant you slapppa-da-bass until the air actually begins to move from the speaker cone's movement.

I like the idea presented, how a second rail is used to provide more power only when needed (part of the Class H ethos) .. but then again ... "paper notions" don't always align with "reality". I'm still not totally onboard with Class D amps (switching amps), but I'm getting there. For now, I'm a stubborn obsoletist that relies on older Class A/B amps (one advantage to that is that they are highly cost effective regarding initial purchase ... since they are "old tech" they can be obtained rather easily. That is to say ... they're less expensive cuz they ain't all trick). However as time passes I'm becoming more accepting towards switching amps. I will say that when they are done wrong, they can be real trainwrecks. Some Class D wall wart power supplies create problems that can be tough to troubleshoot ... ie; high pitched squeals and odd noise issues with stompboxes when one of the "modern type lightweight psus" are used. But I'm becoming less hard headed as time passes and the tech improves and better quality stuff is offered.

Here's the text that I found interesting on the Class H power amps... (quoting).

"....... Switching-mode (class D) power supplies work on demand. Instead of constantly working at full power and dissipating excess power as heat, they only ramp up the power output when needed—thousands of times per second. Considerably more efficient than traditional power supplies, switching-mode power supplies are small and light, yet deliver ample power.

Think of a Class H power amp as a car with two engines (in amps, you call them "rails"). One engine runs all the time. The other runs only when musical peaks demand extra power output. A Class H amp only generates a fraction more power than is immediately needed while the output stage operates at its maximum efficiency all the time. And just like in today's hybrid cars, the efficiency of this dual system is far greater than having one engine or rail that must operate all the time. Class H amps don't waste power, require much smaller heat sinks and much lighter power transformers.

It takes huge pulses of energy (current and voltage) to propel a woofer cone out fast enough to match a bass beat. That's called Transient Response and it's the holy grail of amp designers. By carefully selecting transistors with extremely high slew rates and optimizing other proprietary parts of the circuitry, the (*brand name*) amps are able to react instantly to even the most demanding electronic bass impulses. If the woofers in your PA system can keep up, your audience will hear a tighter, crisper, more natural sound.

Instead of operating relatively continuously like Class AB circuits, Class H amps, which were first described in NASA technology (U.S. patent 3,319,175), feature rail tracking for effectively modulating the power supply rails with only the peaks of the input signal. This technology has revolutionized pro audio amp designs with its outstanding performance and efficiency. When combined with switching-mode power supplies (class D) that replace heavy toroid transformers, Class H designs provide more dynamic punch and, because they are so much more efficient, run cooler and don't require huge, heavy heat sinks."
(end quotes)

Now I realize that information like this generates eye-rolling responses sometimes, simply because it's old news to the hip. But for people like myself that are deeply involved in many projects on a continuous basis, keeping up with the Jetsons on everything topical isn't high on the priority list. So if this is overdiscussed here, my apologies for kicking the dead horse. Heck, I have a hard enough time keeping up with my own ideas, let alone the technical progressions of amp makers.

My world is based on old analog, through-hole, flying lead, hand soldered constructs. Maybe one day I'll be more accepting of modern hi-freq switched, surface mount, pcb-mounted controls made-by-robots stuff.

Maybe .... one day. But by then, the world will be one giant integrated-biomass-circuit chip. Heheh ... The Matrix.

But I doubt it ... I'll probably always be that old freak that has the workshop in the desert with the 40 watt needle-point soldering pen, a loupe stuck in one eye, reading paper-bound schematics in my down time as relaxation.

Thanks for tolerating something I found as "new".

Flux.

Just FYI, class G and H are considered pretty much the same thing. The larger GK heads and the larger EBS heads use this type of power amp. I would not think there would be much difference between topologies (if any) regarding your premise (based on the many posts by EE's dealing with all these types of amps).


Edit: That info in bold seems quite wrong (i.e., they are calling SMPS 'class D', confusing power supply with amp topology'). Sounds like an amp company who uses Class G or H power amps trying to market that there is an advantage. You can run a class D amp with a traditional power supply (like the AG500) or a class A/B amp with an SMPS (Markbass LMIII). Seems like that is mostly marketing BS. Tha being said, back before class D became available, I do believe that class G/H runs a bit more efficiently versus traditional class A/B amps (kind of that car thing where you only use all cylinders when needed).

The short story is that as far as amps are concerned, ANY OF THE AVAILABLE TECHNOLOGIES WILL WORK FINE. Class-A, Class AB, class G, class H, class D, whatever. (The only exception is class-C, usable only for radio in general)

There is simply no truth to the idea that some styles of design are "slower" (whatever that is supposed to mean) in some way that a person can actually detect by ear. Nor that one is "more dynamic" (whatever that is supposed to mean) than another.

That sort of claim is either marketing BS, or the spew of an ignorant but talkative person.

Naturally, ANY technology can be so poorly used as to mess up the audio, so a poor design, if poor enough, could sound pretty bad; muffled, dull, non-dynamic, etc. But you knew that... right?

The choice of technology is based on other things..... Whether you will think it is cool enough to possibly buy, whether it can be done in a lightweight portable package, whether it is excessively expensive, etc.

Specifically, the multi-step power supply designs (class H for instance) use the same supply voltage as the non-stepped designs.

The notion that there is some sort of "added" supply to produce peaks is a bit backwards..... The difference is that they add INTERMEDIATE steps which lower the amount of power that is wasted as heat.

A conventional amp of the same power can produce the same results audibly, but will get warmer.

As for the speaker....

Basic physics says that ANY speaker whose cone has a non-zero mass must have some small amount of lag time. Good speakers have that small enough to not be a problem. The very same effect causes a limit to frequency response of that driver.

if you could directly move the air with NO mechanical cone etc, there would STILL be a lag time due to a finite force and the non-zero mass of air.

Luckily, the origin of the sounds (string movement, vibrating piece of wood, etc) also has frequency limits etc due to basic physics.....

Also, air can be compressed, which means the mic or ear drums won't be driven directly so, there's always loss of energy from amp output to mic/ears. Not that being mechanically linked from the driver to the mic/ears would be lossless, but the compression/rarefaction costs more energy.

+1 Again - Things that don't matter: Class of the amp, the topology of the amp, the topology of the power supply, color of the case, size of the logo

As long as they are well design they'll do the job, and sound the same.

the speaker still responds to the signal coming into it (assuming it can do so linearly) so the 100 Hz component starts a little late and ends a little late. If it started late and endedn "on time" it would not be 100 Hz anymore. I think of this as simply phase shift, and we know that real speakers have phase shift that varies with frequency.

similarly with amp rise times. It was actually Bob Lee who straightened me out on that. If the amp cannot reach the full voltage called for by the end of say the positive going portion of a signal, it doesn't reproduce a different frequency. It "turns around" when the input voltage reverses. So it simply doesn't reach full voltage at the relevant frequency. Which will look like a high frequency roll off at higher power levels.

I was shopping around for a power amp recently so I recognise the text you quoted as Behringer marketing blurb. It gets off to a bad start calling SMPS class D.

As others have mentioned, the advantage of classes G and H is in efficiency. Less wasted heat so smaller, lighter heatsinks are required, the average load on the power supply is less and the result is a saving in size and weight overall. There's no reason to expect any audible difference.

:p The Behringer source explains a lot to me.

Maybe so but I don't believe their marketing people are any more creative than most others. They've certainly had quality/reliability problems in the past but I suspect they don't deserve the level of derision they get courtesy of folk wisdom. To be fair, their conventional EP series power amps seem to have a good record for reliability and for delivering the claimed output.

Anyway, I couldn't resist a cheap experiment so I now have a 1U class H power amp that delivers (I think) 580W RMS into my 8 ohm cab. Haven't run it past apartment level yet. It could all go horribly wrong but we'll see.

I think most of the 'bad feeling' toward that company has to do with blatant copying of other companies' designs, etc. No big deal to me, since luckily I am in a financial position to not have to go to the complete low end of the category. However, that really bad description of Class H and the supposed advantages is pretty lame.

The point is, the 'interest' in that class of am shown by the OP is misplaced since he is basing it on that very inaccurate marketing blurb.

Regarding the OP, it's worth considering that an amp doesn't give a speaker a "position" to reach, it merely inputs an voltage/current that creates an acceleration in the driver cone. (This misconception causes no small grief in the 'underpowering' debate)

So it's not a question of whether a speaker would have trouble "reaching" the necessary point. Instead, it's a question of how fast the driver can be accelerated in opposite directions.