I was testing my 50V -> 5-15V synchronous buck regulator, when the 0.01R current sense resistor, series with the 50V DC bus exploded, and the PCB was burning in flames 
The output current was about 60A. I think the mosfets cross-conducted, and they become a "short-circuit", and the 0.01R burned.
The MOSFETs were FDP3632, paralelled 2 on the upper side, and the lower side.
btw, the 1.5mm^2 wire in a bucket of water was a good dummy load
Here was the resistor....
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/dscn0466.jpg
Wear eye-protecting glasses, when you experiment with high power stuffs!!!!
Your safeti is the most important!
(I was wearing also)
The output current was about 60A. I think the mosfets cross-conducted, and they become a "short-circuit", and the 0.01R burned.
The MOSFETs were FDP3632, paralelled 2 on the upper side, and the lower side.
btw, the 1.5mm^2 wire in a bucket of water was a good dummy load
Here was the resistor....
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/dscn0466.jpg
Wear eye-protecting glasses, when you experiment with high power stuffs!!!!
Your safeti is the most important!
(I was wearing also)
Danko said:I was testing my 50V -> 5-15V synchronous buck regulator, when the 0.01R current sense resistor, series with the 50V DC bus exploded, and the PCB was burning in flames
The output current was about 60A. I think the mosfets cross-conducted, and they become a "short-circuit", and the 0.01R burned.
The MOSFETs were FDP3632, paralelled 2 on the upper side, and the lower side.
btw, the 1.5mm^2 wire in a bucket of water was a good dummy load
Here was the resistor....
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/dscn0466.jpg
Wear eye-protecting glasses, when you experiment with high power stuffs!!!!
Your safeti is the most important!
(I was wearing also)
Hi,
Cool pic
You might suffer from uneven power loss in the fets as you have two in parallel......
I always use one big fet rather than two..... much better solution IMHO.
\Jens
Hi!
Finally, back from school....
Here are 2 pictures from the prototype, as Eva asked
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/dscn0465.jpg
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/dscn0453.jpg
The actual state is not exactly that, on the picture. I removed the 0.01 Ohm SMD resistors from the PCB, and moved to a separate small PCB. (not that, wich exploded)
This is on picture 0465. There is some calculation, on the desk
poobah! I gonna lower the inductor value. I haven't thought the inductor....
I have a small problem with dead time. If there is no dead time, that is a problem. Cross conduction, and smoke, and other bad things happen... If there IS dead time, then there is a short time, where none of the MOSFETs are conducting, so the inductor's stored energy can cause a voltage spike on the MOSFETs. The inductor is about 9-10nH. Can this inductor cause such high voltage spike, that can kill a 100V MOSFET? How can I avoid this spike?
JensRasmussen!
I had to use 2 paralelled MOSFETs, becouse I only had "just" 80A types.
Finally, back from school....
Here are 2 pictures from the prototype, as Eva asked
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/dscn0465.jpg
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/dscn0453.jpg
The actual state is not exactly that, on the picture. I removed the 0.01 Ohm SMD resistors from the PCB, and moved to a separate small PCB. (not that, wich exploded)
This is on picture 0465. There is some calculation, on the desk
poobah! I gonna lower the inductor value. I haven't thought the inductor....
I have a small problem with dead time. If there is no dead time, that is a problem. Cross conduction, and smoke, and other bad things happen... If there IS dead time, then there is a short time, where none of the MOSFETs are conducting, so the inductor's stored energy can cause a voltage spike on the MOSFETs. The inductor is about 9-10nH. Can this inductor cause such high voltage spike, that can kill a 100V MOSFET? How can I avoid this spike?
JensRasmussen!
I had to use 2 paralelled MOSFETs, becouse I only had "just" 80A types.
Danko,
If you are sure you have no deadtime, that is the first place to work...
During your dead time, the inductor will draw its current from the body diode in the lower FET, this should not cause a large spike. But, the body diode of a Low-voltage FET is not the best thing either.
Check your calculations for peak inductor current. If your inductor saturates; it becomes a short circuit.
P.S. 9nH does not sound right. What is your Fs, Vi, Vo, and Po?
If you are sure you have no deadtime, that is the first place to work...
During your dead time, the inductor will draw its current from the body diode in the lower FET, this should not cause a large spike. But, the body diode of a Low-voltage FET is not the best thing either.
Check your calculations for peak inductor current. If your inductor saturates; it becomes a short circuit.
P.S. 9nH does not sound right. What is your Fs, Vi, Vo, and Po?
Yes, schematic is coming soon. I just have to make it "human readable" format. It's quite messy....
poobah:
Fs=222kHz
Vi= 50V DC , from a 1.5kVA toroid
Vo=2.5V-15V
Io= /max/ 70A
7 turns on the inductor, and 70A /neglecting the dI/ through the inductor is 490 Amper*turns. The core, I'm using, the max. Amp*turns is about 7-800.
It should not saturate....
About dead-time.... I placed on the PCB R-C 's, but I think they caused too small dead-time.
luka: I have only ~20A diodes, and I had to paralelled many of them, to achieve safe operation. Oh, and the power loss would be almost double.
poobah:
Fs=222kHz
Vi= 50V DC , from a 1.5kVA toroid
Vo=2.5V-15V
Io= /max/ 70A
7 turns on the inductor, and 70A /neglecting the dI/ through the inductor is 490 Amper*turns. The core, I'm using, the max. Amp*turns is about 7-800.
It should not saturate....
About dead-time.... I placed on the PCB R-C 's, but I think they caused too small dead-time.
luka: I have only ~20A diodes, and I had to paralelled many of them, to achieve safe operation. Oh, and the power loss would be almost double.
The schematic:
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/gkrellShoot_09-21-06_201024.png
some notes:
The "voltage" line somes from the output. On the output there's a pot.meter, the "voltage" line is connected to it's wiper.
The "current" comes from a rail-to-rail ( Texas TLC2272 ) op.amp. The current shunt's voltage is amplied by the op.amp.
the circled R-C parts are makeing the dead-time. But... There's is a mistake, I think. One R-C slows down the turn-ON, the other slows down the turn-OFF.
The feedback around the TL494 isn't final.
http://sziget.mine.nu/~danko/aramkor/car-smps/side-project/gkrellShoot_09-21-06_201024.png
some notes:
The "voltage" line somes from the output. On the output there's a pot.meter, the "voltage" line is connected to it's wiper.
The "current" comes from a rail-to-rail ( Texas TLC2272 ) op.amp. The current shunt's voltage is amplied by the op.amp.
the circled R-C parts are makeing the dead-time. But... There's is a mistake, I think. One R-C slows down the turn-ON, the other slows down the turn-OFF.
The feedback around the TL494 isn't final.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.