Mosfets for coilgun switch?
Reaching, Thu Nov 02 2006, 07:55PMokay. i got a bunch of ixfx27n80Q mosfets very cheap and thought to use them for a coilgun switch instead of an igbt. my thought was to parallel 5 of these mosfets to get around 500A peak. the on resistance with 5 in parallel is 60milli ohms. i want to use them up to 400volts. is it worth a try or would the mosfets blow away immediatly? yeah i know the losses are I²*r but the result would be 15000watts ,thats nonsense, time is a another factor or not? a pulsewidth of 500ys or so cannot result in dissiplating 15000watts in the mosfets,.. maybe someone with more knowledge can explain if it would work or not. i got them for a few bucks and i have 45 of them
Re: Mosfets for coilgun switch?
Marko, Thu Nov 02 2006, 08:21PM
Yup, it's not nonsense, it's 15000 watts. Not all that much considering your coilgun operates at 200 000W peak for given data. Voltage drop across mosfets in that case is going to be 30 volts compared to 'few' volts on typical IGBT in same application, ant it's now all about if you want to trade this for mosfet's speed and hope for higher efficiency, although mosfets seem to out-cost IGBT quite a bit, but since you got them free you can try..
Marko, Thu Nov 02 2006, 08:21PM
Yup, it's not nonsense, it's 15000 watts. Not all that much considering your coilgun operates at 200 000W peak for given data. Voltage drop across mosfets in that case is going to be 30 volts compared to 'few' volts on typical IGBT in same application, ant it's now all about if you want to trade this for mosfet's speed and hope for higher efficiency, although mosfets seem to out-cost IGBT quite a bit, but since you got them free you can try..
Re: Mosfets for coilgun switch?
Reaching, Thu Nov 02 2006, 09:07PM
okay i tried it with a simple test setup and the results seems promising. i used a 1000yf 250volt capacitor, full charged to 240volts. 4 of the ixfx27n80 mosfets in parallel gate source resistor of 470 ohms (to discharge the gate in a time X) on the gate of the mosfet a 1,5yf cap in parallel with a 1k resistor (to charge the gate up in a time X) no pulse width modulated ic or so and it works quite well. dont know the L of the coil but it has a resistance of about 0,7 ohms. 0,7 ohms+0,075ohms on resistance=0,775ohms. 240/0,775=309,67A peak. dont know the exact on time of the mosfets yet but after one shot the cap is discharged to about 13 volts. calculated losses in the mosfets are 309,67A²+0,075 ohms on resistance, should be 7192watts dissiplation, so why wont the mosfets die, they even dont get warm.. i dont believe in this silly formula any more
. tommorow i´ll test it up to 400volts. i even dont use a fast recovery diode so the internal body diodes cant be so bad
Reaching, Thu Nov 02 2006, 09:07PM
okay i tried it with a simple test setup and the results seems promising. i used a 1000yf 250volt capacitor, full charged to 240volts. 4 of the ixfx27n80 mosfets in parallel gate source resistor of 470 ohms (to discharge the gate in a time X) on the gate of the mosfet a 1,5yf cap in parallel with a 1k resistor (to charge the gate up in a time X) no pulse width modulated ic or so and it works quite well. dont know the L of the coil but it has a resistance of about 0,7 ohms. 0,7 ohms+0,075ohms on resistance=0,775ohms. 240/0,775=309,67A peak. dont know the exact on time of the mosfets yet but after one shot the cap is discharged to about 13 volts. calculated losses in the mosfets are 309,67A²+0,075 ohms on resistance, should be 7192watts dissiplation, so why wont the mosfets die, they even dont get warm.. i dont believe in this silly formula any more
. tommorow i´ll test it up to 400volts. i even dont use a fast recovery diode so the internal body diodes cant be so badRe: Mosfets for coilgun switch?
Marko, Thu Nov 02 2006, 10:13PM
'Few kilowatts of pulse dissipation trough few us is OK, and that's why pulse current is defined.
During that brief period only couple of joules of heat wll be transferred to die, and mosfet will be OK until you make it dissipate enough energy to push it's 'surge' junction temperature (Tmj as in datasheet) over 150C where it dies.
So 'silli formula' is completely true.
If you want to exploit benefits of this mosfet, I think you should make proper high-current gate drive to turn it on and off in it's rated delays, or use halfbridge switching as an advance.
(I don't have expeirence in that part so I leave it to you).
Body diodes are generally very slow, and in comparison to their switching times some fast diodes would be of great help there.
Marko, Thu Nov 02 2006, 10:13PM
'Few kilowatts of pulse dissipation trough few us is OK, and that's why pulse current is defined.
During that brief period only couple of joules of heat wll be transferred to die, and mosfet will be OK until you make it dissipate enough energy to push it's 'surge' junction temperature (Tmj as in datasheet) over 150C where it dies.
So 'silli formula' is completely true.
If you want to exploit benefits of this mosfet, I think you should make proper high-current gate drive to turn it on and off in it's rated delays, or use halfbridge switching as an advance.
(I don't have expeirence in that part so I leave it to you).
i even dont use a fast recovery diode so the internal body diodes cant be so bad
Body diodes are generally very slow, and in comparison to their switching times some fast diodes would be of great help there.
Re: Mosfets for coilgun switch?
Reaching, Thu Nov 02 2006, 10:34PM
yeah sure, its just a test setup at the moment, only useable for the trigger stage or injector or so, . i want to build a multistage coilgun with around 10 stages or so so timing and proper fotogates are a must. for the moment its just a test if it works or not and it seems to work very good if they can switch 1000yf 250volt.
Reaching, Thu Nov 02 2006, 10:34PM
Firkragg wrote ...
If you want to exploit benefits of this mosfet, I think you should make proper high-current gate drive to turn it on and off in it's rated delays, or use halfbridge switching as an advance.
If you want to exploit benefits of this mosfet, I think you should make proper high-current gate drive to turn it on and off in it's rated delays, or use halfbridge switching as an advance.
yeah sure, its just a test setup at the moment, only useable for the trigger stage or injector or so, . i want to build a multistage coilgun with around 10 stages or so so timing and proper fotogates are a must. for the moment its just a test if it works or not and it seems to work very good if they can switch 1000yf 250volt.
Print this page