DIY Good Class D amplifier

[Acca]

Hi,as I see,this is schematic with 3 fets?
They use single supply and output capacitors,so there can't be DC at speaker.If you don't wired speaker you can use 100R or similar to place on speaker connector and measure voltage on it.
Nice wiring

 

[PetruV]

Whoa, i thought this thread was dead,nice to see stuff added and new schematics,cant wait to try some of these.

 

[Acca]

It's not dead,I'll continue to post on this thread when I make something

 

[aman khalsa]

I don't know,any with small ripple.Because of it I use stabilised SMPS.
schematics:

sir i want to make this amps .. what is the wattage of this smps ?? and what core i can use ?? no or turns

 

[Acca]

Power is max. 150W because it's flyback.Core E42 (for this power) or E33 (60W),when dismantle (if use from TV or monitor) you should write all nuber of turns on paper.This will be used for calculation.If you can measure voltages on original SMPS and tie with number of turns this will be great.

 

[Acca]

Very simple self osc. half bridge (with fet's),about 300W or higher (with bigger core,I use cores from TV high voltage transformer,it can give about 600W) for powering amps,with short circuit protection.
Feedback turns on power transformer is calculated to give about 12V.
Original E33 core from AT/ATX will give +/-12V on original 5V pins and +/-28V on 12V pins.It can be usedfor split supply,only add more two diodes,inductor,capacitor and bleeder resistor.It's calculated to dissipate about 1W (simple,R=U*U).It not heat very much and it's better than IR2153 because isolated gate drive,and have short circuit protection (because if you short output gate voltage will be 0V and SMPS stops work).For GDT you can use E16 or E19 (without gap) and wind 22t. for primary and 2x28t. for sec.
If you reduce 56R in feedback freq. will decrease.I use E25 because it's very cheap to buy.
f meter shows 59.2kHz,but for power transformer calculating I use 58kHz,for safety margin.
Few pictures of SMPS and d3k amp (amp with 6N137):

 

[ionutzxpo]

hi acca, D6 is DB3?

 

[Acca]

D6 can be any diac,and D1-D3 is two zeners 16V 1,3W wired "back to back" or anode at anode or can be TVS diode.

 

[ionutzxpo]

great! thank you

 

[Acca]

If any RFI/EMI present,add 2n2 1-2kV capacitor from -310 to output GND.They will reduce it to minimal.

 

[ionutzxpo]

ok thank you il test this with an ETD49 core and some E19 recovered pc smps trafo's
the windings to primary and secondary is calculated as usual smps nothing special right?

 

[Acca]

Yes
For storage inductor you should use E33,I'll calculate no of turns and gap.

 

[Acca]

E19 recovered pc smps trafo's

They shouldn't have gap.Standby supply in PC PSU have gap,and base drive transformer not.

 

[ionutzxpo]

ohhh i gues you are right all of them have 0.5mm gap so they are a no-go for oscilating trafo
if we use ETD49 core and IRFP460 i gues we can make a 1KW shortcircuit protected very simple smps

 

[Acca]

Yes,because if you short output it will remove gate drive and smps will go in interminent mode.When fail is removed it continue to work normally.But it don't have overcurrent protection (like SMPS is 500W and you push it into 2kW)
You can use E16 from BDT from PC PSU,it don't have gap.For short circuit test use old or cheap fet's,if they blow,you will be happy

 

[ionutzxpo]

yep... irf8xx or 7xx they are very cheap here in romania even irfp460 are quite cheap and afordable

 

[Acca]

There isn't,IRF840 is about 1e and IRFP460 about 3-4 euros

 

[ionutzxpo]

its 3$ irfp460

 

[Acca]

OK.When you finish,share there
And test short ckt protect

 

[Mondan]

Hi,learn more about 2153.It can,but like a buck or boost converter,and not for classic SMPS
PWM on buck and classic SMPS is different.It has several ways to implement feedback.

It will be nice to poste some oscilloscope wave forms.
If in pin 3 (CT) inject a current you modify duty cicle. Is this correct?
In this way the capacitor will charge faster and discharge slowly.
But this procedure can modify the period?