university graduation class d project

[richie00boy]

Hi

For 250w into 4 ohmes, 100v ones are enough

Ahoj,

100v devices are on the absolute limit, as the DC rails needed are 49 volts. At idle the supply may be higher than 50 volts. Certainly, if you use a 40 volt transformer then you will need 100+ volt devices.

 

[umut1001]

for kikakak

hi..no problem..i worked a lot too..i am looking it for 2 months..to get lower distorstion i prefer you to use irf540 st mosfet instead of irf540 n..because it has lower input capacitance..best regards

 

[kikikaka]

Thanks a lot! Maybe it's a good idea. But I'm afraid it's impossible to purchase IRF540 here. 540N is sold everywhere...

 

[umut1001]

kiiiiikkaaaakkkkiiiikaaaaakikakia

thanks hahahaaa...do not help anyone.do not share anything..everybody helps you

 

[kikikaka]

No, no no!

I'm sorry. I think I was too raw few post earlier, and this caused a kind of misunderstanding...

I didn't say, that I won't help, just only that I don't want to share my whole project (including PCB, detaliled schematics, etc....)

I have a reason for that: I'm in the middle of my graduation process, and I don't want anyone to have suspicions about me of stoling my whole project from this forum.

Regards,
kikikaka

 

[luka]

Hi

That's ok, but still you can post some pictures of it. Always great too see new things

 

[kikikaka]

After taking this picture i had a smaller modification: decoupling protecting and signal ground.
Instead of shorting them together i put parallel R and antiparalel diodes to avoid hum and other noises came from earth-loop.

 

[luka]

Hi

Wow. Thats my response. Next thing would be to replace toroid with smps. How powerfull is it, the amp?

 

[kikikaka]

It's only 2*90W @ 4 Ohm @ about 1% THD. But it's far enough for my room...

But the problem: about at 25% power some distorsion can be seen on a scope. It' like on the attached image (that is from irf's class-d tutorial), but only on the lower halfwave. And it's seems to be independent from frequency.

I can hardly imagine how dead-time causes a distorsion like that....

As Umut said, i purchased four IRF540 (st) instead of IRF540N (irf), but they have a text 'CHINA' on them. Someone wrote that these are good only for the dustbin. What's the problem with them?

(sorry for the picture, but in the original document text and arrows are in the wrong place too)

 

[luka]

Hi

They are probably right. I have bought few 2n3055 from toshiba. I had 4 of them in parrallel for current sharing in supply. They are 15A each, but when I went with output current to 5A one of them got "smoked". So I had to replace them with the same from SGS. They worked well over 10A without problem.

So here is probably the same.But note that that "N" is there for a reason. N version has lower Rds-on, bigger current,...

Yea 2x100 is more then enough for normal room

 

[Whortless]

Hi Kikikaka,

As Umut said, i purchased four IRF540 (st) instead of IRF540N (irf), but they have a text 'CHINA' on them. Someone wrote that these are good only for the dustbin. What's the problem with them?



If you would like to damage your drivers, U can use "China" marked IRF540s. I agree with Luka, yes 540N has lower Rds on. But St's 540 has lower input capacitance and lower Qg. But, usually Rds on value has a reverse balance with Qg.
Hýýmm, China marked mosfets may be IRF510 or other lighter mosfet than 540. They could be celaning the original mark and then marking them as IRF540 or so. I tested them on my desings, they can only work for 5-10 seconds while original IRF540 can run for several hours without damage. I got china marked ones for 0.5$ while the original is on the same price and finnaly two fake 540 damaged my HIP4081's one channel.

Kind regards