Ultra Simple Class D

[portreathbeach]

That's a great help. Thanks

Has anyone built this and tested the quality of it?

[portreathbeach]

How big of a heat sink do these devices need. Do Q3, 4, 5 and 6 need to be sinked? And as for the MOSFETS, if I were to use IRF540 and IRF9540 as suggested earlier in the thread, is there a calculation to work out the heat sink size. I am used to building chip-amps, where there is a nice Excel spreadsheet to help you work out heat sink sizes.

Thanks again for all the advice

[portreathbeach]

Sorry to keep asking questions, but her goes....

I understand the principle of Class D amps now, but the output of the comparator will always be a a PWM square wave. It switches N and P MOSFETS when either positive or negative right? If the output of the comparator has to be positive or negative, that means 1 of the MOSFETS (in half bridge) or 2 (full bridge) would be on all the time? If there is no input at the front end of the amp, what is happening at the speaker?

[Simon B]

Originally Posted by portreathbeach

Sorry to keep asking questions, but her goes....

I understand the principle of Class D amps now, but the output of the comparator will always be a a PWM square wave. It switches N and P MOSFETS when either positive or negative right? If the output of the comparator has to be positive or negative, that means 1 of the MOSFETS (in half bridge) or 2 (full bridge) would be on all the time? If there is no input at the front end of the amp, what is happening at the speaker?

With zero input, the mark-space (here read up-down) ratio of the output devices is 50%. Half the time +ve, half the time -ve. Goes through the low-pass filter, averages to zero.

[portreathbeach]

Oh, I see. So if you PWM was say 200Khz, you have a 200Khz square wave at 50% duty and the low pass filter will not pass this so the speaker sees nothing. So even with no input signal, the MOSFETS are switching 1000s of times a second, but because of the filter, they are not passing any current, therefore not getting warm? Have I got that right?


Oh, just having another look at the circuit. Where is the actual filter?

[benb]

Originally Posted by portreathbeach

Oh, I see. So if you PWM was say 200Khz, you have a 200Khz square wave at 50% duty and the low pass filter will not pass this so the speaker sees nothing. So even with no input signal, the MOSFETS are switching 1000s of times a second, but because of the filter, they are not passing any current, therefore not getting warm? Have I got that right?

Yes, though of course there are losses due to switching times, the output driving stray capacitance and such (keeping the outputs shielded should help reduce radiated RF interference). It also helps a whole lot to make sure each MOSFET turns OFF before the other one turns on.

Originally Posted by portreathbeach

Oh, just having another look at the circuit. Where is the actual filter?

That's a good question. A speaker's voice coil itself is inductive at sufficiently high frequencies and will pass less current, but still, my guess is the voice coil will get excessively warm. I hope whoever makes this has fire insurance.

Oh, there's this from earlier in the thread (bolding mone):

Originally Posted by Simon B

Interesting thread. Just to let others see something of the schematics and pcb layout more easily, here are screenshots of them. For the original pdf file, see early part of thread.

Note that the output inductor is NOT shown either on the schematic or the pcb layout - whether air or ferrite cored it probably deserves to be tied down somehow.

Attachment 276326

Attachment 276327

Attachment 276328

[portreathbeach]

Cheers benb. The output part of this amp looks similar to most of the other half wave Class D amps I have seen, apart from the drawing not showing the inductor. The only problem I can see here is the dead time, there is no way of setting it, so you run into a problem of both MOSFETs being on at the same time for a very small amount of time!

benb, have you much experience with Class D amps, and if so, is this a good one to start with, or am I better off building one that uses specific ICs to drive the MOSFETs? If the later option is better, could you point me in the right direction to a simple class D schematic, preferably not using SMT ICs. Thanks again

[benb]

I have no actual experience with Class D for audio, I just know the theory. I've used Class D to drive small motors, which is much less demanding than (good) audio.

I've had good experience doing SMT, and my suggestion is don't be too afraid to try it. Get a good magnifier, a good small-tipped iron, keep everything clean and go for it. Googling for SMT tutorial brings up this:
Surface Mount Soldering 101 Video

[portreathbeach]

Hi benb. I'm not too worried about soldering SMTs as I have changed SMT LEDs in mobile phones before, it is more for the reason that standard ICs are a lot easier to use in breadboard prototyping. Thanks for the help, I'll keep looking around for a nice Class D to attemp

[DJNUBZ]

IMO Simple/Zen designs seem to always sound the best with class A and AB amps weither they are tube or SS. Does simple class D sound as good or is this more of a test of how simple we can make a class D amp and still have useable sound?