diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Class D (http://www.diyaudio.com/forums/class-d/)
-   -   Project : Practical Class-D PWM Subwoofer Amplifiers with optocoupled Gate Drivers (http://www.diyaudio.com/forums/class-d/25321-project-practical-class-d-pwm-subwoofer-amplifiers-optocoupled-gate-drivers.html)

Workhorse 4th January 2004 11:46 AM

Project : Practical Class-D PWM Subwoofer Amplifiers with optocoupled Gate Drivers
 
2 Attachment(s)
schematic of ramp and comparator

Workhorse 4th January 2004 11:49 AM

Power stage
 
2 Attachment(s)
Schematic

Workhorse 18th January 2004 10:43 AM

Reply
 
Is there any difference made by large propagation delays?:smash:

Cybergent 18th January 2004 07:02 PM

You mean propagation delay in the opto devices?

As long as the delay is the same for both devices and for rising and falling signals I donīt see any problem. It will just result in a total delay, but not affect the signal.

You just cannot add any NFB as it would be out of time.

Workhorse 20th January 2004 07:58 AM

Commentable Thoughts
 
2 Attachment(s)
THIS is RAMP GENERATOR SCHEMATIC

Switching Frequency=15KHz

phase_accurate 20th January 2004 08:40 AM

Quote:

As long as the delay is the same for both devices and for rising and falling signals I donīt see any problem. It will just result in a total delay, but not affect the signal.

You just cannot add any NFB as it would be out of time.
You may of course apply feedback as long as the delay is distinctively lower than the clock period.

Feedback is mainly affected by the switching frequency. The unity gain point shouldn't be higher than half the switching frequency.

Edit:
Forgot to mention that this statement is only valid for carrier-based class-d amps. For self oscillating designs the unity-gain point is higher, since we deliberately want instability in order to cause oscillation. This increases the amount of NFB that can be applied for a given switching frequency.

Regards

Charles

Workhorse 20th January 2004 09:31 AM

Commentable Thoughts
 
[B]Does self oscillating designs are more excellent and give better performance.

phase_accurate 20th January 2004 09:54 AM

I have never built a self-oscillating class-d amp (only carrier-based ones). The highly- praised Philips amps are self-oscillating so are the B&O ones (and also Lars' ZAP-pulse).
I only know the difference from simulations where the self-oscillating ones seem to be a little better as well (but simulations are simulations and real-world is real-world).

It is easier to take feedback from the output filter for self-oscillating topologies.

Apart from simplicity there are other advantages (one of them is good supply voltage suppression).

Some members here have built self-oscillating designs and they are convinced about their quality.

I by myself like the good carrier suppression one can achieve with carrier-based designs. Therefore I will build one in a feww weeks. It will also take feedback from the output filter (which is a little trickier than doing it with a self-oscillating design, but definitely not impossible).

Regards

Charles

rchua77 20th January 2004 10:12 AM

Hi guys,

this is really interesting, if anyone gets a working prototype working, do shout it out.

I'd like to try my hand building one too.

Regards

Workhorse 20th January 2004 12:21 PM

Reply
 
2 Attachment(s)
This is output schematic:


All times are GMT. The time now is 02:30 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright Đ1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2