Class D amps?

[gtrmac]

I've been poking around on these websites that offer Class D amp modules and kits and I'm very interested to get a couple modules and try them out. I was considering a project using a discreet preamp and a Class D module to use as a bass amp for instance.

Has anybody else looked at this stuff?

http://www.hypex.nl/

http://www.41hz.com/

http://www.icepower.bang-olufsen.com/sw1067.asp

 

[bcarso]

Errr...how does the old saying go: I could tell you, but then I would have to kill you?

Seriously, I don't know much about any of them, but just a little about the B&O, which is a pretty decent hysteretic-converter analog-input class D machine. It doesn't have a clock oscillator per se and as a result will shift about under signal drive, which sort of cheats on the agency approval issues concerned with rf emission limits. Basically, it doesn't stay in one place long enough to be measured. People have reported that they sound good.

The B&O stuff are primarily the brainchildren of Karsten Nielsen, one of the spawn of the Tech Univ of Denmark, whose intrinsic modesty is on a par with a recently joined member of this forum (to quote the latter, "you know who you are.")

But his stuff is decent and a lot of effort has gone into making it work and be manufacturable, so they deserve to sell some amps.

The other guy ex of Philips sounds like a decent sort and probably quite capable, but he's in the honeymoon phase at this point so probably hasn't done much yet.

I don't know the guitar outfit at all---not to say they might not be fine.

 

[Svart]

Zetex makes Class D audio drivers, just hook up your signal and some mosfets and you are good to go. Digikey sells the ICs and a protoboard.


:thumb:

 

[rlantis]

I was considering a project using a discreet preamp and a Class D module to use as a bass amp for instance.

Has anybody else looked at this stuff?

I think there's a big future for Class D. Lot's of possibilities and advantages. Did my own bass rig, too, as you are considering. Good fun.

Here are a few other Class D amplifier links:
http://users.ece.gatech.edu/~mleach/ece4435/f01/ClassD2.pdf
http://www.classd.org/indexold.html

Thanks for the interesting links. Keep 'em coming. Enjoy!

 

[bcarso]

gtr, I opened the other site and realized they are selling Trip*th stuff. IMO those are not particularly good, although you have to appreciate the founder's persistence and marketing skills.

The Phil*ps integrated stuff is not bad btw, and is getting into some interesting power ranges now, especially into 4 ohm loads. The Zet*x imo is a bit dicey because of the way they drive the FETs: you have to be very attentive to the threshold voltages and control charge to avoid either excessive deadtimes or significant shoot-through currents. But they are easy to get hold of and play with and you will at the very least learn a lot.

Internationa R*ctifier showed me some very impressive hybrid designs with a new high-low side driver and their super low inductance packaged FETs. Noise and distortion were quite low for a less-than-perfect error amp in the loop. I need to follow up with them as they were close to release of the design, just a few issues to solve.

Cr*wn has a patent on opposing buck converters which when used with their patented pulse-width modulation scheme avoids shoot through intrinsically and done correctly delivers very nice performance. Unfortunately only H*rman companies can use the approach.

I would avoid the siren song of the "pure digital" approach for now, but that's another whole subject for another day.

 

[Svart]

I don't know much about class D stuff and haven't bothered to look yet, but is the output drive a totempole output centering around P/N channel FETs or do they just drive a lowside switch and let the switching idle somewhere mid-duty cycle(kinda like class A pwm..)? I assume they use totempole(or pushpull) but for high currents and faster switching times, wouldn't it be better to use an N-FET as the highside? I'll have to read the datasheets on the IRF floating highside drivers again..

I've worked with the HV floating drivers from IRF that will bootstrap a N-FET for use as the highside(which worked mighty well..). However, this might not be the best approach but i'm not a wizard at this by anymeans so it's all just ideas until i can work it out!

:thumb:

 

[rlantis]

Here (MicroSemi) is where I got mine (as an evaluation unit), and it works quite nicely:

http://www.microsemi.com/catalog/parmlist.asp?P0_CAT=PM&P1_TYPE=AMAX

 

[rlantis]

[quote author="Svart"]I don't know much about class D stuff and haven't bothered to look yet,:[/quote]

Here is a good primer-
http://www.classd.org/indexold.html

 

[Svart]

:thumb:
I'll be looking at that tonight.

 

[rlantis]

Here is a good primer-
http://www.classd.org/indexold.html

OOPS sorry...
I meant this link:
http://users.ece.gatech.edu/~mleach/ece4435/f01/ClassD2.pdf

 

[Svart]

ok thanks!

 

[bcarso]

Thanks for reminding me about Microsemi. I've used a lot of their very low power stuff and some higher power as well. They are very nice folks to deal with too.

 

[clintrubber]

from Brad:

The Phil*ps integrated stuff is not bad btw, and is getting into some interesting power ranges now, especially into 4 ohm loads.

You can switch back to writing Philips again, we've spotted this discussion already.

Br*n* - just met him a few times in real or mail, but he looked a very knowledgable person, to say the least. Didn't knew he was involved with Hypex.

Regards,

Peter

 

[bcarso]

"You can switch back to writing Philips again, we've spotted this discussion already. "

I don't know exactly what the thinly veiled names have to do with although I don't want to get into litigation with anyone. Reminds of what a patent attorney I used once said about Philips. Sorry I can't share it here ;-)

Also reminds me of the story I heard from a certain corporate VP at H*rman. He would get these cellphone calls from the Executive Chairman, who had become increasingly concerned about security. So the EC started using more and more code words etc. when calling. Problem was the VP had usually hardly any idea as to what the code was...

 

[clintrubber]

Reminds of what a patent attorney I used once said about Philips. Sorry I can't share it here ;-)

Making me curious ! :wink:

P. had the most patents again last year, but hey, that's probably not what that attorney said and it won't be a surprise - it's a small country here and it doesn't say much about their use and or quality of course. :wink:

 

[SSLtech]

I was going to give you all the 411, but the CEO from EV wouldn't give me the OK. The VP from CBS wants a PO ASAP... -WTF???

:wink:

Keef

 

[SonsOfThunder]

[quote author="SSLtech"]I was going to give you all the 411, but the CEO from EV wouldn't give me the OK. The VP from CBS wants a PO ASAP... -WTF???[/quote]

Gee Keef, I thought you'd keep that on the DL. :razz:

Unless things have changed in the last few years, T-P@th stuff was pretty fragile, you couldn't short it or it was instant toast. Not a robust design imho.

I can certainly see the use of class D in cell phones where they gotta get the MOST out of the battery and quality is not so big an item. And big class D car amps are certainly putting bread and butter on my table. But that doesn't mean I want to listen to them. In my opinion they are ok for bass, but the damping is usually quite poor and noticable due to the o/p filter. There are some newer designs that put the filter inside the FB path to try and compensate for this. I would not want to listen to "full range" class D for critical listening.

I think this does have some potential for improvement, but I don't think class D will ever replace a good old A/B for sheer sound quality.

Just my $0.02. Hope you don't ask for change back.

Peace!

 

[bcarso]

Roger K. The eagle has landed, toad is in the hole (and exploding). 10-4.

 

[bcarso]

quote: "I would not want to listen to "full range" class D for critical listening. "

I'm with you there, especially on the output filter issue. Once again Gerald Stanley at Crown has been his clever self though, and mentioned a recent patent where they get high damping factors with stability at high freqencies. He's also got one that addresses some of the shortcomings of "pure digital" amps, i.e. pulse-code-to-pulse-width amps. On the last one I said I was upset that he had stolen my ideas a few years before I thought of them. Funny haven't heard from him since---maybe he didn't appreciate the humor

I was at an AES presentation where this ****** alleged that digital amps sounded better as long as you got rid of the switching noise. He went on to say that various golden ears folk had though they sounded more tubelike. I couldn't resist barking from the back of the room "Well now don't you really think this is just because they necessarily have high output impedance, and people are hearing timbral shifts due to the interaction with real loudspeaker loads?"

Unfortunately marketing types have been getting the pure digital crap fed to them so much that they really believe it, which has forced engineers to figure out how to make it listenable. The requirements on the power supply are greatly exacerbated by the pure digital approach, resulting in some pretty grotesque stuff out there. No one argues the appeal of efficiency, but that's not what's driving the pure digital crusaders.

 

[Svart]

Hey Bcarso, I bet a class D power amp is the answer to my PWM motor driver problems!! oh wait.. it IS a class D power amp... nevermind..

"Well now don't you really think this is just because they necessarily have high output impedance, and people are hearing timbral shifts due to the interaction with real loudspeaker loads?"

:green:

that's awesome.. did they have much to say about that?

In reality though, if you are switching through a highly inductive load at a given frequency(whatever it is..) wouldn't you sustain an average voltage and current due to that very inductance and the associated capacitance?

:thumb: