Fullbridge Class D PA ultra high power

[kees52]

Some playing I have done, but now I did receive the line transformer for the allfet, so get wind it, need 2000 windings each side.

I did use a changed bessel low pass, I did see after tests that when set -3db point higher, lowpass put more power out, did get 0.023 % distortion with 2 ohms load but with a multipole filter, who needs much work to get hd low. shoot through mosfets was 18 amps.

Did also play with the chip IR2011 these did go nice, has low hd also, but with multipole filter use, and not yet a dead time setup. The model is not oke or I do something wrong but the negative side mosfet do not get full on, it go to half the top fet does, so 80 volts and lower side 40 volts, I do not now why.

Last photo with the blue carrier you see that the negative side is strange, needs the same as the positive side,

regards

 

[kartino]

I dont understand why people always do hijacking other people thread

Guys, make your own thread.
Even I am not active doen't mean that you can use my thread for your irrelevant subject create by me. The intention to create thread is to easy find and discuss specific thread. In this case is my design. You can use this thread ONLY to discuss all related to my design as per subject.

Regards,
Kartino

 

[kartino]

This is update for IR based fullbridge.
It is working as it is and tested by many users in Indonesia.

 

[kartino]

And this is UcD version discrete based fullbridge class D.
Also it is working as it is and tested by many users in Indonesia

 

[kartino]

If you asking why I update the schematic after bad things in this forum just because I don't want somebody blame me, say that my schematic is not working.

 

[kartino]

To see my amps, check my youtube:

https://www.youtube.com/watch?v=ueUOhhtKqc8&list=UUN_8t1dl1M1ljVAIP6VYPMA

 

[kartino]

No PCB layout available.

 

[stewin]

hi katrino have you ever tried to remove the speaker on d2kfb when it is playing . and how did you manage to deal with noise issues in ucd xlite2fb???

 

[kartino]

hi katrino have you ever tried to remove the speaker on d2kfb when it is playing . and how did you manage to deal with noise issues in ucd xlite2fb???

The amp just fine when disconnected during playing music.

Noise in UcD can be eliminated with:
1. Fully synchronized
2. In vice versa dont make two channel with same speed. Make other channel a bit lower or higher speed. 10% difference just fine.

 

[kartino]

I mean same speed but.not synched. The same speed but not synched tend to have resonance and possible noise.

Another option is by using ferrite core because ferrite core is usually E or EI closed magnetic field. That is why hypex use ferrite core.

 

[ravpreet420]

thanks kartino for share ��

 

[redmx]

Wrong diodes

And this is UcD version discrete based fullbridge class D.
Also it is working as it is and tested by many users in Indonesia

Hello kartino, in your diagram ucd the diodes D4 and D5, do not work there, should with a lower vf.

 

[redmx]

SMD

My modulator / demodulator, driver, 450KHZ, zero noise, instead of ir2110, use fan7392 with 4 output amps, two ground planes, referenced to gnd and, - B, MORE than 300 amplifiers in operation.

 

[kees52]

My modulator / demodulator, driver, 450KHZ, zero noise, instead of ir2110, use fan7392 with 4 output amps, two ground planes, referenced to gnd and, - B, MORE than 300 amplifiers in operation.

Looks nice driver, I have looked at it, the 600 volts bootstrap give it more immunity for defects and it is fast, however there is not a dead time circuit in it so need a extern schitttrigger delay.

the blown drivers I read much is because possible to high bootstrap voltages with high power amps, these ic,s can prevent, but the use of transient diodes is always a safe idea.

Is there a simulation model for this FAN7292?

regards

 

[kartino]

Hello kartino, in your diagram ucd the diodes D4 and D5, do not work there, should with a lower vf.

It is just fine.
It is not a concept but tested working schematic

Why do you think it is not working?

Maybe you mention about D4, D6, D28, D30
Yes for that should be schottky diode

 

[kartino]

My modulator / demodulator, driver, 450KHZ, zero noise, instead of ir2110, use fan7392 with 4 output amps, two ground planes, referenced to gnd and, - B, MORE than 300 amplifiers in operation.

400kHz will only work.well with small and fast mosfet. Less than 200W output. Yes, it sounds but not have good clarity. My two cents advise.

 

[kees52]

400kHz will only work.well with small and fast mosfet. Less than 200W output. Yes, it sounds but not have good clarity. My two cents advise.

Or very beefy driver, like a totem pole, maybe a extra totem pole outside the driver to push the mosfets to fast switching, most mosfets do switch in the ten's of Nseconds but need to be driven properly.

 

[kartino]

Or very beefy driver, like a totem pole, maybe a extra totem pole outside the driver to push the mosfets to fast switching, most mosfets do switch in the ten's of Nseconds but need to be driven properly.

It not too good enough for big mosfet AND higher voltage as you will need higher voltage for high output. In short only small amp sounds good.

 

[kees52]

It not too good enough for big mosfet AND higher voltage as you will need higher voltage for high output. In short only small amp sounds good.

Good respons, and I think you are right, high speed big mosfets are not yet available, class D is just in child shoes, even when people think it is not, it is.

 

[kartino]

Good respons, and I think you are right, high speed big mosfets are not yet available, class D is just in child shoes, even when people think it is not, it is.

I mean good sounds is for hi fi with 400kHz or higher PWM only small class D.

For pro Audio a 250kHz to 300kHz is good enough. It can be built up to 3000W easily. My amps work at 270-300kHz for half bridge. 200-250kHz for fullbridge. The amps sound superior compared to other class both clarity and power handling