simple class D amp

[latala]

over the last 35 years i have built all kinds and classes of amplifier class a b ab but i have never tried a class d
do any you have a schematic hopefully with discrete components
that could be built any thing in the order of 20 watts or so
the few attempts i have made were very disapointing
regards trev

[bob123]

Heres one i made earlier, it has been built and tested and works with some tweaking and the correct components. I suggest using a voltage comp for the switching sections.

This circuit gives a power output of
(15 / sqrt 2)²/8 = 14W into 8ohms
:. 28W into 4 ohms.

[Gorilla]

Class D amplifers have two sub-catorgories; Carrier based and self oscillating. Carrier based usally use a triangle waveform to pulse width modulate the audio signal. Feedback can be added with degrees of success. Self-oscillating amps rely on some sort of positive feedback to achieve modulation. The UCD or SODFA amps are examples of this.

If you want something simple in concept try a PWM amp without feedback. Later if you wanted feedback could be added. More challenging would be a DIY UCD. The topology is almost unparalleled in terms of its component count simplicity but is more diffcult to understand.

Search the forum there are plenty of examples round.

[koolkid731]

Nah, All it takes is to have total phase shift of 360 degrees and unity gain at oscillation frequency.

[classd4sure]

Right on KoolKid.

Here's a DIY UCD that should get you up and running with some minor tweaking.

My DIY UCD

Regards,
Chris

[TerryG]

So would someone explain to me what I would hear differently if an amp had a 360 degree phase shift? I don't see how this would be a benefit or a downside. This keeps comming up and I don't get why.

TG

[bob123]

surly a 360 degree phase shift would mean the output would be almost exactly the same as the input!

[Gorilla]

surly a 360 degree phase shift would mean the output would be almost exactly the same as the input!

Yes, except there will be a system loop gain imposed. If the gain is greater than unity oscillation will occur (Look up your favourite text book or Wikipedia on Nyquist stability criteria).

The difficult bit, like TerryG has alluded, is how this oscillation modulates meaningful sound .

[classd4sure]

The filter coil integrates the square waveforme of the power stage, creating an error signal which includes all circuit delays. This passes through a phase lead network, combined with the delays produces the required unity gain at oscillation frequency condition. The audio signal is your setpoint, you compare the two and there's your PWM audio signal.

[latala]

thanks for the replies re the amp i have been sent a cicrcuit thanks bob
but as a bit of background when i posted the origional request i was thinking of something like the sinclair x20 or x10 has any one out there ever heard or got one of these as they were around in 1964/65 but then seamed ti disapear without trace any thougts or comments?
REGARDS TREV