Fixed "mistakes" on great sch :)

[ChocoHolic]

Zero DC and canceling of rail hum at 50% duty cycle...

 

[Mikelandjelo]

U r right.....Practicaly this is easy and works.....and it is killing high frequency from audio spectrum too.I tested this many times......And decided to kill hum on input side.....Now i have perfect high frequencies from audio,with very small bypass caps.....We agreed about one thing,hum is very high frequency.

 

[Mikelandjelo]

I found one more thing too.guy deleted ground on output stages 🙂 .... he did everything to show something and to prevent everything.This schematic works fantastic.

 

[nigelwright7557]

Are you going to build it ?
It will be interesting to see how good it sounds.

 

[Mikelandjelo]

I already made it , triangle wave is perfect.I made everything in modules to test it and it works perfect(very messy setup 🙂 ).Probably i will make proper layout in Protel and assemble it as one design.I am to bussy now but in next few weeks probably i will make it.

 

[Mikelandjelo]

Well,some more fixes

I defined few things here.Max modulation and highest efficiency is possible on frequencies at am of 65KHz couse comparator is slow.Also few caps r making performances to be lower than they can be good.

 

[nigelwright7557]

Well,some more fixes

I defined few things here.Max modulation and highest efficiency is possible on frequencies at am of 65KHz couse comparator is slow.Also few caps r making performances to be lower than they can be good.

I think those capacitors were to slow down switch on of mosfets.

Another method is to discharge mosfet gate through a diode but charge it up through say 10 ohms.

 

[ChocoHolic]

C5,C6,C11,C12 are for dead time adjustment.
The required values will depend on the chosen MosFets, chosen gate drive impedances and the desired trade off regarding idle losses vs dead time distortion.

 

[Mikelandjelo]

Also diodes for fast turn off r needed on gates.