OT: Class D Amplifier Design

Some people seem to actually prefer the 2nd harmonic distortion that toobs produce.

Inferior only in terms of efficiency I would venture to say. I don't think you can beat it for fidelity.

Reply to
Cursitor Doom
Loading thread data ...

r all.

D amp ? You have no emitter resistors and foldback in the power supply coul d not react fast enough so what do you do ?

r
n

_original.pdf

really?

here's the patent

formatting link

Reply to
Lasse Langwadt Christensen

:O (gob smacked!)

Reply to
David Eather

Yes but I figured out what happened. You need to enclose URLs with less-than and greater than ("") symbols, particularly if there are spaces in it.

Reply to
krw

Yet in the most common push pull tube-transformer applications it still produces odd harmonics.

What is of note, is tubes give softer limiting than their solid state cousins, so less of the higher harmonics.

A Class A stage, depending on feedback, is no more linear than others. What it does though, in the same way as AB classes do, it has inherently very good linearity at low volumes.

You're probably right about fidelity re class D, I suspect most error and noise comes from controlling jitter.

--
Mike Perkins 
Video Solutions Ltd 
 Click to see the full signature
Reply to
Mike Perkins

And the PWM encoder (triangle linearity, etc.).

Reply to
krw

Re class D, I'm not so much worried about distortion or freq response. What I worry about is intermod with the switching frequency. Is this a problem, if not why not? Mark

Reply to
mkolber1

The sum is going to be well out of the audio band. The switching frequency is at least 10x the frequency of interest.

Reply to
krw

** Fraid that drivel makes no sense at all.

Since you are more than little peculiar and slow to comprehend, I will spell it out in baby talk.

Internally, there are four (4) class D amplifiers wired in " bridge mode" pairs - one pair for each stereo channel. At the speaker output terminals, both signal lines are hot. There is one SMPS feeding the lot.

** They add up just fine to anyone with normal intelligence and some familiarity with audio electronics terminology.
** You really are colossally dim witted. Go f*ck yourself.

.... Phil

Reply to
Phil Allison

And then if you make the output filter more complex, you run into the problem that to get really low harmonic distortion numbers you need to take feedback post-filter in lieu of/in addition to feedback from the switching node, so you have to figure out how to make that loop stable in light of the additional phase shifts.

Reply to
bitrex

Some class-D amplifiers have excellent distortion performance:

formatting link
formatting link

Some people think that the PWM should be generated in the digital domain and then put to the power stage. Of course the gain of the power stage is proportional to the power supply rails, and any ripple will cause distortion, so then you turn the problem of designing a power amplifier into a problem of designing an ideal power supply regulator, which is pretty much just as hard. Also, variation in the rise and fall times of the power stage would affect the distortion. Also, at high power levels the output filter inductor is likely to produce significant distortion (or at least more than a good amplifier produces). For all of those reasons, it is better (cheaper, smaller, more efficient, better performance) to take feedback from the output terminals so that the modulator corrects for all of the non-ideal things, rather than generating an ideal bitstream and trying to build an ideal power stage, and failing because of the imperfections of real components.

Reply to
Chris Jones

Well, quite. I suspected maybe some shyster is putting fancy connectors on cheap TV coax and hoping to lure in some of the more gullible "audiophiles."

Reply to
Cursitor Doom

I would not try to put tens of real watts through them.

By the way - RCA connectors are no too good either, gold plated crap stays crap.

There are good gonnectors, e.g. Lemo, and they should match the pricing profile of audiophiles. It's no wonder that they cost like a Swiss watch, they are made in the same way.

--

-TV
Reply to
Tauno Voipio

Is there any kind of clock synchronisation needed in the class D modulation process?

Reply to
Chris

Nope. Serious audiophiles demand gold *N-types* for the superior frequency response. ;-)

Reply to
Cursitor Doom

bs produce. "

Yup. Those studies date back to the 1070s even with reguawlr listeners, not audiophiles. It was found that they prefer the even order harmonic distort ion to the odd order distortion.

I believe (and it seems some experts agree) that is because even order harm onic distortion is actually harmonic. For example if you play a middle A on a piano it is 440 Hz. The second harmonic is at 880 Hz. You can play the n ext A up the scale and it harmonizes perfectly. Now if you go to third harm onic I am not sure, maybe you get a D sharp or some shit (could look it up but ain't) and that doesn't really "harmonize". Actualkly the A and D sharp do make a chord but it is used in music sparingly. you have to have a pla ce in the music for it to fit. Like minors and sevenths and fifths, they ha ve their place but are not used all that much for obvious reasons. They sim ply don't sound all that great. They are like, in speech, a preposition or something, I forgot the word but it means to connect two things together. ( damn am I old)

Another thing, which is actually put forth by me is that higher order disto rtion is more irritating. Like if you have some third harmonic distortion, it is not as annoying as fifth harmonic distortion, or even high like eleve nth harmonic distortion. this is due to the same effect, provided you have good ears, that makes a single piccolo discernible in a full blown orchestr a. It is because the frequency is so far removed from the bulk of the mater ial.

nk you can beat it for fidelity. "

That is probably why single ended tube amps sell for so much. Now that is a udiophoolery. You got a $20,000 amp thaqt puts out five watts a channel and then have to buy a phono preamp for another two grand. But it is a hobby f or them.

Like why does Jay Leno need all those cars ? We can call it audiophoolery a ll we want but I will be happy to build you a simple amp, no tone controls, switches maybe three inputs and no phono stage using one 6BQ5 per channel for a few grand. I'll even make it look nice. And you should see some of th is stuff. A plate chassis resting on a piece of finely finished mahogany or whatever. And every capacitor replaced last year with twenty buck audiophi le caps. And rolling tubes, they just buy tubes and replace them and see if some sound better than others.

I am almost thinking of getting into the business, I can certainly handle i t but I am tired. I am tired of the whole field. And then to deal with thos e people, well, a twenty grand amp ? I kinda think I don't even want it in my house, I don't want to be responsible for it. Not unless I get a grand t o change one capacitor. Some would probably actually pay that...

Like my buddy's lawyer friend, has the limo that was behind the one JFK was riding in when he got shot. Why ? He is more narcissistic than Donald Trum p. (I am NOT saying to vote for Clinton, not at all, but facts are facts, t he guy is full of himself, and this of course does not mean the same is not true of Hillary Clinton) Bottom line is that the really rich want to show off. I've known lots of people like that and lately I find them intolerable , and one of them owes me money and calls to brag about his new speakers he is buying. Good bye asshole, I was born without you.

Anyway that may explain why I am not already in the audiophoolery game. Bec ause of the type of people I would have to deal with.

Like I said before, 12-2 Romex probably makes the best speaker wire. In fac t I think I am going to go on Audioasylum and tell them to do a side by sid e comparison with their $300 speaker wires. I think some of the will be sur prised that twenty bucks worth of Romex sounds better. Lower resistance mea ns better damping. Unless they are using welding cables which won't fit the connectors. Those are stranded and woven because they are expected to be m ove almost constantly. Usually, you don't move speakers around all that muc h, especially when they weight 300 pound and the olady is always bitching a bout it.

Anyway, I am not so sure single ended class A always sounds better. Push pu ll class A can damp in a way nothing else can. Can you hear it ? How much m oney does anyone want to spend to find out ?

Reply to
jurb6006

That appears to show the Crown configuration, which they tout as even more efficient than regular class D. Has some sort of coils in it.

Crown amps are usually two ohm stable and if they maintain that design phil osophy then the current is detected by coils, which is what I would do. The frequency it set, and you don't even need a coil in series with anything. Just wrap a wire around the collector (or drain) feeds to the outputs. The problem is calibrating it. But if you want it to work fast enough it defini tely has to come after the main filters. That's because idiots go out with the amp on and set at a high volume and THEN connect the speakers, and then they short the wires and wonder why the amp is smoking. and I know one guy , was a bar owner, hooked up a QSC in bridged mode which delivers 1,300 wat ts into eight ohms to ONE speaker and the speaker catches on fire.

Reply to
jurb6006

No, not really but it is a good idea to have both channels running at the s ame speed. Say if they are off by 100 Hz, you might just hear that via grou nd currents and imperfect filtering in the power supply. So i a way it woul d be best to have them running in synch, but not critical. If you get the r est of the design right it doesn't matter, but when shit gets into mass pro duction out come the bean counters and it could fall to pieces.

So the answer is yes and no. If the two channels share the same power suppl y it is really good idea to have them synched unless the designer quit NASA the other say. In real life just either synch them or separate them.

Reply to
jurb6006

I do not understand, why we still are debating _analog_ audio connectors.

Why not go completely digital ?

While in the tube era, it was necessary to separate the turntable from the amplifier and the amplifier from the speaker, there are no such requirements today.

Just use active speakers and all the issues with speaker cables goes away. The cross-over filtering can be done digitally in the speaker box and the (Class-D) amplifier(s) move the voice coil(s).

Thus, each speaker box is supplied with a full bandwidth digital signal and some DC power (or mains AC-power).

To transfer signals to the speaker boxes, even such simple connection as 24 bit linear SPDIF would be sufficient to carry the attenuated (by any volume control) signal to the speaker boxes.

Reply to
upsidedown

You jaw jaw jaw far to much, buster. any chance you cut the wordcuont of your stupid ignoramus postings?

Reply to
Gunther Heiko Hagen

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.