pfc900w

[mirbagheri1122]

hello
i`ve builded a power factor correction 900w 385v based on der274 from power integration design example.
my problem is blowing pfs729 as i increasing load from 200w to 400w.
pls guide

[ali_asadzadeh]

Can you explain more what's going on wrong?

[mirbagheri1122]

Hello
Thanks for your time,
I read your reply.maybe my problem is carbon contaminants under pfs that cause  arcing,although I have  not connect the new replaced PFC to power yet.
I have attached correct schematics PDF for your reference, please note that in der274 there is no snuber cap,But in PI expert software it showed a capacitor should be placed across boost diode in schematics, and I have added a 100pf 1K cap,is that a problem?

[bktemp]

The inductor looks quite small. Are you sure it is not going into saturation at higher current levels and overloading the mosfet?

[mirbagheri1122]

Thanks,Actually my heat-sink is quite huge,It has these dimensions(15cm X 5cm X 10cm) and my inductor core is sendust 125u from Magnetics inc the part number is 0077254A7, I have attached the datasheet for your reference

[bktemp]

Something is odd:
The core has 168nH/T², therefore it needs 109 turns for 2mH. But I can count only around 34 in your picture. This would give around 200uH.
109 turns at around 5A peak current (rough estimation for 900W) gives a NI value of 545AT. The graph in the datasheet ends at 400AT. At 545AT the core will be totally saturated.
Either way, the core is too small.

[mirbagheri1122]

Thanks for your reply, Actually if you see the inductor picture again you will see some gray tapes on inner layers, there is 3 layers of wire and the total count is 109 turns, I have added the tape for increasing the insulation between layers

[mirbagheri1122]

I have cut down the inductor wires to be much lower, But I have not tested the new chip,I'm worried about blowing a new chip again,do you have any Idea?

[ali_asadzadeh]

Connect your PFC device directly on your Heatsink

[johansen]

several things can be happening. first is that the inductor is probably saturating.
second is that the inductance of the output capacitor and the boost diode is too high and the fet is blown by overvoltage.
third is that the fet fails due to over current.. probably caused by a saturated inductor.

[Circlotron]

How long are the earth tracks on the pcb, particularly on the mosfet source lead? Could you post a pic of the pcb layout? At these power levels layout is everything. I spend several months working on a 3.6 kW pfc about 10 years ago (someone else's design, I was just the lab hack) and at full load particularly when you wind the AC input down toward the minimum they can be scary!

[mirbagheri1122]

hi thanks for your reply
because of emi i added another trace between source of ic and small low esr caps and connected floated heatsink to source.now my circuit is giving 700 w without overheating but more than 700 is overheating.my final layout is added to this post.

[mirbagheri1122]

hi could you send me your 3.6kw design

[mirbagheri1122]

hi thanks for your time
my reverse recovery charge is 50 nq and in attention to frequency and voltage give me about 1 w dissipation.
cmos of ic is aout 200pf that is nonsignificant.and my heatsink resistant is about 0.8.
and my silpad is gray.