Class-D Amp with Discrete Gate Drivers

[phase_accurate]

And therefore to get even low figure i have used the discrete modulator and comparator scheme using opamps to get 0.01% THD

Congratualtions for that choice.

Regards

Charles

 

[kASD]

Reply

This is the discrete version of Frontend circuit
ITS switching frequency is 250KHz
THD=0.01%
Frequency response=10Hz TO 25KHz

 

[Circlotron]

Quasi-complementary fanatic? Have you done any mosfet amps like this using lateral fets (Hexfets)?

 

[phase_accurate]

Graham, an IRF 640 IS a hexfet AFAIK.

Regards

Charles

 

[kASD]

Reply

I have designed power amplifiers using N-CHANNEL IRFP250N MOSFETS in quazicomplementary manner

 

[Circlotron]

Yes, that's what I meant - a conventional amp with qcs. I've got lots of N-channel fets from SMPS's, even a number of 500v 32 amp ones and basically no P-channel ones. Gotta go now. Midnight again!

 

[kASD]

Reply

WHAT is meant by QCS and can i help u plz

 

[phase_accurate]

Did you ever have a look at this recent thread ?

http://www.diyaudio.com/forums/showthread.php?s=&threadid=25790

Regards

Charles

P.S. It isn't using the same type of FETs but should be reworkable.

 

[kASD]

Reply

i have designed quasicomplementary amps using n-channel mosfets upto 2KW in CLASS-AB mode
NOW what ur plan CIRCLOTRON

 

[IVX]

Dear Amp Man,
i'm sorry , but only one thing of the your circuit can be working - triangle generator, however at 60khz(isn't 250khz) and it have almost sinusoidal form because opamps too slow. Your feedback is positive (only regarding upper fet, lower fet is always ON and with 60khz frequency it's OFF for 300nS ), input R=1K & C=4.7U gives -11db at 10hz, however your circuit with lm3524 has much more errors. Why you write 0.01%, but not 0.0001%? Why are you stopped?
Man, just tell me - "smile, you in candid camera"
Best regards Ivan.

 

[kASD]

Reply

Hey IVAN
I have build the prototype in reality and the opamps are operated
@250KHz and the global feedback is negative.
The distortion figures i stated are absolutly wise and not a humorous detail.

 

[kASD]

Reply

What will be the best ?
A Triangular Wave or SAWtooth wave to be used as modulating carrier?

 

[Cybergent]

I would go for the triangular wave, as I think PWM conversion will be more linear due to less dU/dt in the reference signal.

 

[km]

hi.

>Dear Amp Man,
i'm sorry , but only one thing of the your circuit can be working - triangle generator, however at 60khz(isn't 250khz) .........Man, just tell me - "smile, you in candid camera">

hehe , i have suggested before to amp_man (or should i say amp_men....?) to build at least one of the amps before talking so much about how good it is

btw. some time ago you said you never heard about factor and his projects but the other day you said in emails it is actually your (late) brother , right?

are we all on candid camera here?

rgds karsten madsen - cadaudio.dk

 

[dmfraser]

Driver IC

Look up the IR2110 series of driver ICs. They can handle up to 600V supply rails.

 

[kASD]

Commentable Thoughts

Can this proposed schematic will give out of phase drive signal correctly??. I want to implement it because LM319 is dual comparator and therefore I want to use the full IC. and also i dont want to use CMOS ICs.

Comments,Suggestions,Insults,Remarks are most welcomed.

Regards
AmPmAn

 

[subwo1]

I thought that if the analog signal is triggered on two comparators that the distortion would be higher since they may not trigger at the same time.

 

[kASD]

Commentable Thoughts

I thought that if the analog signal is triggered on two comparators that the distortion would be higher since they may not trigger at the same time.

But Sir, I have seen this type of implementation in several High-end Class-D amps.

Ampman requires more suggestions indeed.

Regards
AmPmAn

 

[phase_accurate]

I thought that if the analog signal is triggered on two comparators that the distortion would be higher since they may not trigger at the same time.

One can deliberately provoke this in order to achieve adjustable dead-time

Regards

Charles

 

[kASD]

One can deliberately provoke this in order to achieve adjustable dead-time

Regards

Charles

Ampman ThanX SIR, PHASE_ACCURATE for his INVALUABLE suggestion . So, can i implement this circuit in my New CLass-D project.

With Sincere Regards
AmpMAN