Class D audio driver with external mosfets

[[email protected]]

Any competitors to this product:
http://www.irf.com/product-info/datasheets/data/irs2092.pdf

?

Any comments on the above?

Michael

 

[[email protected]]

I don't know if this competes but it's something else to look at..http://www.st.com/stonline/books/pdf/docs/5200.pdf

D from BC
[email protected]
British Columbia
Canada

THD = 10%? At 25 watts? You can't be serious...?

The idea was to have external mosfets, with a cheap-ish driver. That
way, if the mosfets blow, or need extra cooling, action could be
taken.

I kind of liked this:
http://www.nxp.com/acrobat_download/usermanuals/UM10155_2.pdf

but then I noticed the Patent they have on it. d'OH...

Michael

 

[Eeyore]

Any competitors to this product:
http://www.irf.com/product-info/datasheets/data/irs2092.pdf

Check out Hypex.

Graham

 

[Eeyore]

I kind of liked this:
http://www.nxp.com/acrobat_download/usermanuals/UM10155_2.pdf

but then I noticed the Patent they have on it. d'OH...

Is that the UcD design ?

Graham

 

[Vladimir Vassilevsky]

Any competitors to this product:
http://www.irf.com/product-info/datasheets/data/irs2092.pdf

There used to be the analog class D ICs from Zetex and few other
vendors, however it seems like everybody went digital.

Any comments on the above?

0. It works.
1. It is specifically made to force you to use DirectFETs.
2. It is quite tricky to make it work as good as advertized, however it
is possible.


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

 

[[email protected]]

There used to be the analog class D ICs from Zetex and few other
vendors, however it seems like everybody went digital.


0. It works.
1. It is specifically made to force you to use DirectFETs.
2. It is quite tricky to make it work as good as advertized, however it
is possible.

Vladimir Vassilevsky
DSP and Mixed Signal Design Consultanthttp://www.abvolt.com

DirectFETs, eh? So no IRF530s, even if paralleled?

I don't suppose it's possible to use something like this to drive
external mosfets (instead of speakers directly)...?
http://www.ti.com/lit/gpn/tpa3123d2

Thanks,

Michael

 

[Vladimir Vassilevsky]

DirectFETs, eh? So no IRF530s, even if paralleled?

Possible. However the THD and the power efficiency will suffer
significantly (compared to the DirectFETs).

I don't suppose it's possible to use something like this to drive
external mosfets (instead of speakers directly)...?
http://www.ti.com/lit/gpn/tpa3123d2

This one is closer to the boom-box class, whereas IRS2092 pretends to be
high fidelity. What are you trying to build?


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com

 

[[email protected]]

Possible. However the THD and the power efficiency will suffer
significantly (compared to the DirectFETs).


This one is closer to the boom-box class, whereas IRS2092 pretends to be
high fidelity. What are you trying to build?

60Wpc audio amplifier for home party use, maybe adapt it to a homemade
car stereo amplifier once I get a good-enough boost converter built.

Class D would allow me to use a smaller (cheaper) transformer... just
so long as it sounds reasonable: <1% THD+N...

Thanks,

Michael

 

[Eeyore]

60Wpc audio amplifier for home party use, maybe adapt it to a homemade
car stereo amplifier once I get a good-enough boost converter built.

Class D would allow me to use a smaller (cheaper) transformer... just
so long as it sounds reasonable: <1% THD+N...

THD+N alone tells you quite little about the sound. The harmonic content
distribution is vastly more important.

Typical solid state amps with 1% THD sound disgusting. You only need a tiny
transformer by my standards (I've done up to 2kW for professional SR use),
build a gainclone (google it). You may want a slightly larger transformer (abd
heatsink) for 'party' use since the continuous power demand will be a LOT
greater than normal hi-fi use.

Graham

 

[[email protected]]

THD+N alone tells you quite little about the sound. The harmonic content
distribution is vastly more important.

Typical solid state amps with 1% THD sound disgusting. You only need a tiny
transformer by my standards (I've done up to 2kW for professional SR use),
build a gainclone (google it).

What's the fun in that?

Besides, from experience (party at my in-laws' house) chipamps burn
out easily from too much partying, and they're expensive to replace.
Brother-in-law's JVC unit had an STK 412-090 "chip amp" - both
channels on one chip, IIRC. That one blew out one night. Heat
transfer isn't all that great either. With individual transistors you
have a better chance at cooling... plus if they blow out, they're
cheap to replace.

I was going to build this (second schematic):
http://sound.westhost.com/project03.htm

Cheapish output transistors (cheaper than the 3A project transistors),
and reasonable THD.

Then I thought, why not go Class D... smaller transformer...

You may want a slightly larger transformer (abd
heatsink) for 'party' use since the continuous power demand will be a LOT
greater than normal hi-fi use.

Yep, just so long as the transformer can handle full continuous power
+ heat generation for all channels...

Graham

Michael

 

[Eeyore]

What's the fun in that?

Besides, from experience (party at my in-laws' house) chipamps burn
out easily from too much partying, and they're expensive to replace.
Brother-in-law's JVC unit had an STK 412-090 "chip amp" - both
channels on one chip, IIRC. That one blew out one night. Heat
transfer isn't all that great either.

Why do you think I said it needed a good heatsink ?

I 'torture' my designs into 4 and 2 ohms loads with ultra-compressed low dynamic
range 'DJ@ music. Why ? Because we know people do that. Ditto for speakers.

You're talking to an expert in the field of audio amplication and reproduction. DO
NOT try and make out you're cleverer.

If you want to make a piece of junk, fine but don't bother listening to me will
you.

You need to learn about output stage efficiency, Rthetaj-c, Rthetaj-a and
Rthetac-a. Motorola used to have some ANCIENT app notes on it. It's kinda assumed
these days that DESIGNERS (and this is a design group) might understand these
thing.

You also need to know about Iav vs Irms in the power supply, ripple current, ripple
voltage etc etc.

It's not FUN (esp getting it wrong) it's WORK ! And I see no sign of you having any
aptitiude for learning.

Graham