Class D popping up everywhere. Innovation?

[Tomass]

Interesting subject, I had to study a bit about these in class for my course (electronic engineering). However my tutor was telling us that Class D is rarely used for audio amplification, guess she didn't look into bass amps

To me it doesn't really matter what class it is, if it sounds good, its good. Well it might bother me if someone tried to sell class C bass amps

 

[wneff]

Getting back to the original question: Class D popping up everywhere. Innovation?
I have wondered about this myself. The quick story: I was a Walter Woods owner from about 1996 until a couple years ago when it was stolen. I called WW for current price and delivery after the theft. At the time it was something like 2 years and $3000. I had paid around $1000 for my 450W green light WW back in 1996. The current high price on the Ultra is the reason why I didn't get another WW - and the lead time. Coincidentally, at this time, several mfrs were coming out with class D switch mode power supply amps, a term I had only heard from Walter himself - or after reading articles about his amps. My first purchase was the Acoustic Image Focus amp, which I had to send back twice for failures. Anyway, the point is that it was only $1000 vs. $3000 for a new WW. I now have the Carvin BX500 (it's ok), a GK MB2500 (sounds nice), and the new Genz 9.0. The nice thing is that these amps have a direct XLR outputs on them. And we all know that the Genz 9.0 is only around $800, which is a great price (of course this is true because they are not made here in the USA vs. Walter makes his amps here, but that's another topic).

So, the big question is: what took all these other companies so long to make an amp like WW had been making since the late 70s or early 80s? I used to think that he had a patent on the design or something - but who knows? Heck, even SWR is finally coming out with their new class D AmpLite (just saw it in the 1/10 issue of BP)

I believe Walter Woods had the patent....patents expire after 25 years.

 

[Bob Lee (QSC)]

Walter Woods may have had a patent on a particular design, but not on class D amplification itself.

 

[slyjoe]

My understanding is the class D amp appeared in the 50s. However, it wasn't until high powered, high speed FETs appeared (1970s?) that the topology made sense.

Bob or agedhorse will probably correct me if I'm wrong though.

 

[Volume4]

Interesting subject, I had to study a bit about these in class for my course (electronic engineering). However my tutor was telling us that Class D is rarely used for audio amplification, guess she didn't look into bass amps

It is actually showing up in quite a few consumer audio and home theater systems lately. I expect it to take over the inexpensive shelf system market very soon.

 

[Tomass]

Um, righto i'll eat my words there hahaha. I was not aware of that, because Class C is biased to only conduct 50% of the sine wave. Generally there used in RF applications, like CB radios and such. I wonder if they sound rubbish

 

[vin*tone]

Vol4.... (re: CD vs record) a digital audio file is delivered in 44.1khz chunks while the vinyl record isn't. My noobish understanding is that this can effect other harmonic content (you probably know more about that than I do). This is a different thing to a class D bass amp.

With the portaflex there is an "ext amp" plug which can be used as a line-in or line-out. You'd probably use the sansamp before the B15 though. Sounds great but not loud enough for live use.

 

[Tomass]

Vol4.... (re: CD vs record) a digital audio file is delivered in 44.1khz chunks while the vinyl record isn't. My noobish understanding is that this can effect other harmonic content (you probably know more about that than I do). This is a different thing to a class D bass amp.

With the portaflex there is an "ext amp" plug which can be used as a line-in or line-out. You'd probably use the sansamp before the B15 though. Sounds great but not loud enough for live use.

that 44.1khz frequency is the sampling rate, which is generally double of what is desired to be heard. When you use D/A and A/D converters, your taking little splices of a sine wave (audio signal) and converting it to a digital signal (using the Analogue to Digital Converter). Once that is processed and filtered etc, it is then re-assembled using the Digital to Analogue (D/A) converter. Because all the samples are taken as square waves, you loose a lot of nuances, harmonics. Because of this digital inperfection, you will get some instances where digital audio devices will sound a little colder. Hope I explained that alrite.

http://cnx.org/content/m11443/latest/analog_sampling.jpg

 

[Bob Lee (QSC)]

Um, righto i'll eat my words there hahaha. I was not aware of that, because Class C is biased to only conduct 50% of the sine wave. Generally there used in RF applications, like CB radios and such. I wonder if they sound rubbish

Class C is biased to conduct <<50% of the sine wave. The idea is to just get a pulse of energy to excite a resonant tank that fills out the sine wave.

They're not practical for general audio applications, and only slightly less impractical for single-frequency audio circuitry.

 

[Bob Lee (QSC)]

that 44.1khz frequency is the sampling rate, which is generally double of what is desired to be heard. When you use D/A and A/D converters, your taking little splices of a sine wave (audio signal) and converting it to a digital signal (using the Analogue to Digital Converter). Once that is processed and filtered etc, it is then re-assembled using the Digital to Analogue (D/A) converter. Because all the samples are taken as square waves, you loose a lot of nuances, harmonics. Because of this digital inperfection, you will get some instances where digital audio devices will sound a little colder. Hope I explained that alrite.

http://cnx.org/content/m11443/latest/analog_sampling.jpg

The samples aren't square waves. If the sample rate and bit depth are high enough, you won't lose nuances or harmonics within the audio spectrum.

 

[Tomass]

The samples aren't square waves. If the sample rate and bit depth are high enough, you won't lose nuances or harmonics within the audio spectrum.

Yer thats true, the technology is pretty decent now, although nothing is perfect in this world . Yer sorry my bad about the square wave, not all Digital systems use square wave.

By the way I have QSC PLX1104, love that power amp, so light and powerful

 

[Volume4]

Um, righto i'll eat my words there hahaha. I was not aware of that, because Class C is biased to only conduct 50% of the sine wave. Generally there used in RF applications, like CB radios and such. I wonder if they sound rubbish

Don't eat your words yet I meant Class D and it could just be that switching power supplies are what is actually taking over and I'm confusing the two.

Since we strayed onto the topic a little here, please let me recommend this book to any and all. It's a great book nd should be interesting to non-musicians and musicians alike.


http://www.amazon.com/Perfecting-Sound-Forever-History-Recorded/dp/0571211658

 

[Volume4]

that 44.1khz frequency is the sampling rate, which is generally double of what is desired to be heard. When you use D/A and A/D converters, your taking little splices of a sine wave (audio signal) and converting it to a digital signal (using the Analogue to Digital Converter). Once that is processed and filtered etc, it is then re-assembled using the Digital to Analogue (D/A) converter. Because all the samples are taken as square waves, you loose a lot of nuances, harmonics. Because of this digital inperfection, you will get some instances where digital audio devices will sound a little colder. Hope I explained that alrite.

http://cnx.org/content/m11443/latest/analog_sampling.jpg

That pretty much is what I was wondering might be going on in the newer Class D Bass amps. It seems I don't have to worry about it.

 

[Tomass]

Yer I didnt mean to sway it off topic so much, I was like you with this topic, really did not know anything about Class D amps. Good way to spend my morning

 

[Bob Lee (QSC)]

My understanding is the class D amp appeared in the 50s. However, it wasn't until high powered, high speed FETs appeared (1970s?) that the topology made sense.

Bob or agedhorse will probably correct me if I'm wrong though.

FET technology has been improving over the years. A general rule of thumb in class D design is that the switching frequency should be at least 10× the highest audio frequency that the amplifier will see.

However, higher switching frequencies complicate the amp designs. FET gates are capacitive, and yet their drive circuits have to be able to turn them on fast and off even faster. In addition, there has to be a short "dead time" between when the FET on one rail switches off and the opposite FET on the opposite rail switches on. This is to prevent a disastrous situation where their "on" periods overlap and cause shoot-through. That dead time, though, adds a constant time offset to each switching period that will cause distortion if it isn't corrected in the modulator. This is all easier to accomplish at, say, 50 kHz than maybe 300 kHz. That's a major reason why class D first gained widespread use in amplification of bass frequencies.

 

[J.D.B.]

The easiest way to spot a class D is, NO transformer. That's the REAL weight saving. As to their sound, well......Depends what yer going after.

Josh

 

[Bob Lee (QSC)]

The easiest way to spot a class D is, NO transformer. That's the REAL weight saving. As to their sound, well......Depends what yer going after.

Josh

No transformer? That's not an indicator of class D.

If you're talking about a power transformer, then any amp that doesn't run off of straight DC (i.e., batteries) has one.

If you mean an output transformer, then just about all solid-state amps are transformer free, whether they're class A, B, AB, D, G, or H.

 

[Tomass]

FET technology has been improving over the years. A general rule of thumb in class D design is that the switching frequency should be at least 10× the highest audio frequency that the amplifier will see.

However, higher switching frequencies complicate the amp designs. FET gates are capacitive, and yet their drive circuits have to be able to turn them on fast and off even faster. In addition, there has to be a short "dead time" between when the FET on one rail switches off and the opposite FET on the opposite rail switches on. This is to prevent a disastrous situation where their "on" periods overlap and cause shoot-through. That dead time, though, adds a constant time offset to each switching period that will cause distortion if it isn't corrected in the modulator. This is all easier to accomplish at, say, 50 kHz than maybe 300 kHz. That's a major reason why class D first gained widespread use in amplification of bass frequencies.

Wow, thats impressive engineering right there. Do these class D bass amps have a limit in the frequency spectrum though, because of the limits of the FET switching time.

 

[vin*tone]

OK so what about this......

[Invalid or Expired Link Removed]

The AG500 runs a Link Removed Class T amplifier and what look like a single pair of mosfets.

Tripath calls it a "digital audio amplifier".

 

[agedhorse]

FET technology has been improving over the years. A general rule of thumb in class D design is that the switching frequency should be at least 10× the highest audio frequency that the amplifier will see.

However, higher switching frequencies complicate the amp designs. FET gates are capacitive, and yet their drive circuits have to be able to turn them on fast and off even faster. In addition, there has to be a short "dead time" between when the FET on one rail switches off and the opposite FET on the opposite rail switches on. This is to prevent a disastrous situation where their "on" periods overlap and cause shoot-through. That dead time, though, adds a constant time offset to each switching period that will cause distortion if it isn't corrected in the modulator. This is all easier to accomplish at, say, 50 kHz than maybe 300 kHz. That's a major reason why class D first gained widespread use in amplification of bass frequencies.

As complicated as this appears to be in Bob's description, it's even HARDER in practice to design a good sounding, reliable class D amp that does not spew RF interference. Designing class D and SMPS products are related and a very highly specialized area of electrical engineering.

Tripath is a company that tried to promote a packaged solution, a one chip or module solution that could be used to develop amplifiers around. After losing many millions of dollars, they went bankrupt about 2 years ago.

Class D has been around since the late 50's and early 60's, Clive Sinclair (sp?) promoted a couple watt class D kit that worked sort of, was not reliable and very limiter in frequency response. There have been many attempts, Peavey introduced the DECA line about 20 years ago, they were ok when they worked but suffered from catastrophic failures when they didn't. Very sophisticated and they really did roll their own in a really unique (at the time) product.

There are a lot of class G/H power amps out there, Crest, Crown, Peavey all make them.