Going to build a Coilgun
Sonny, Sun Mar 12 2006, 08:11PMHello, i'm from Sweden and i'm a new user to this forum and site. Great one =)
I got a projectwork to do so i decide to go for a coilgun. I'm awaiting 10, 400V 3900uF capacitors which i think will arrive next week. I decide for using five coils in this one (bought 10 capacitor for future plans =)). The biggest problem i got is that i don't realy know how large current i can discharge from my capacitors yet, any ideas? I was thinking about 100-150A for 5ms at most.
Ordered some "IRG4PH50U" IGBTs from Digikey earlyer this week but i dunno if they are on vacation now. When i was using the coil calculator on the "magnetic gun club" i notic a performance increas with the AWG22 wire instead of using autocalculation (AWG24 wire). The other specs was: 390V, 120A, 11mm diameter, 25mm lenght and .1 ohm external resistance.
I'll post some pics when i star working on this project.
Re: Going to build a Coilgun
..., Sun Mar 12 2006, 09:49PM
the max current you can discharge from a capacitor is governed by the circuit's resistance/inductance drop. If you are calculating it stick just with the resistance of the coil to determine the amount of current your switch needs to be rated for.
The absolute max you can draw is governed by the caps internal resistance/voltage it is charged to. I would say yours are about .005ohms, so you at 400v could get (voltage[volts]=current[amps]*resistance[ohms]) 80ka out of them, but that would probably damage the cap. I would suggest you keep it under 1ka per cap.
How long you could draw that current for is limited by the amount of energy stored it the cap, generally expressed in Joules (J=.5*capacitance[farads]*voltage[volts]^2). So each cap has about 300J. Determining how long a pulse you can get is a little trickier, because as you use energy from the cap the voltage drops, so the current drops, etc. For a given resistance you would take
seconds[milliseconds]=resistance[ohms] *capacitance[millifarads]*ln(initial voltage[volts]/final voltage[volts]). So if you wanted to keep the cap voltage above 200v, with a 3.9mf cap, a 1ohm coil and an initial voltage of 400v you would get about 2.7ms. Don't take that number too seriously because it does not take the highly inductive nature of the coil into account, but it will give you an idea of how long (the actual result will be longer)
Using larger wire increases the performance because it lowers the resistance of the coil, which wastes less energy so you get higher efficiencies.
good luck!
..., Sun Mar 12 2006, 09:49PM
the max current you can discharge from a capacitor is governed by the circuit's resistance/inductance drop. If you are calculating it stick just with the resistance of the coil to determine the amount of current your switch needs to be rated for.
The absolute max you can draw is governed by the caps internal resistance/voltage it is charged to. I would say yours are about .005ohms, so you at 400v could get (voltage[volts]=current[amps]*resistance[ohms]) 80ka out of them, but that would probably damage the cap. I would suggest you keep it under 1ka per cap.
How long you could draw that current for is limited by the amount of energy stored it the cap, generally expressed in Joules (J=.5*capacitance[farads]*voltage[volts]^2). So each cap has about 300J. Determining how long a pulse you can get is a little trickier, because as you use energy from the cap the voltage drops, so the current drops, etc. For a given resistance you would take
seconds[milliseconds]=resistance[ohms] *capacitance[millifarads]*ln(initial voltage[volts]/final voltage[volts]). So if you wanted to keep the cap voltage above 200v, with a 3.9mf cap, a 1ohm coil and an initial voltage of 400v you would get about 2.7ms. Don't take that number too seriously because it does not take the highly inductive nature of the coil into account, but it will give you an idea of how long (the actual result will be longer)
Using larger wire increases the performance because it lowers the resistance of the coil, which wastes less energy so you get higher efficiencies.
good luck!
Re: Going to build a Coilgun
Sonny, Mon Mar 13 2006, 02:30AM
Okej, thanks for the help. The IGBTs can only hadle 180A peek current so maybe i should get some other or go for two or three IGBTs per coil to get the maximum out of the capacitors. Stupid to wast enegy left in the capacitors =P
Wonder if my MOSFET/IGBT-drivers can handle three transistors (it's an "IR2121"
)
Sonny, Mon Mar 13 2006, 02:30AM
... wrote ...
the max current you can discharge from a capacitor is governed by the circuit's resistance/inductance drop. If you are calculating it stick just with the resistance of the coil to determine the amount of current your switch needs to be rated for.
The absolute max you can draw is governed by the caps internal resistance/voltage it is charged to. I would say yours are about .005ohms, so you at 400v could get (voltage[volts]=current[amps]*resistance[ohms]) 80ka out of them, but that would probably damage the cap. I would suggest you keep it under 1ka per cap.
How long you could draw that current for is limited by the amount of energy stored it the cap, generally expressed in Joules (J=.5*capacitance[farads]*voltage[volts]^2). So each cap has about 300J. Determining how long a pulse you can get is a little trickier, because as you use energy from the cap the voltage drops, so the current drops, etc. For a given resistance you would take
seconds[milliseconds]=resistance[ohms] *capacitance[millifarads]*ln(initial voltage[volts]/final voltage[volts]). So if you wanted to keep the cap voltage above 200v, with a 3.9mf cap, a 1ohm coil and an initial voltage of 400v you would get about 2.7ms. Don't take that number too seriously because it does not take the highly inductive nature of the coil into account, but it will give you an idea of how long (the actual result will be longer)
Using larger wire increases the performance because it lowers the resistance of the coil, which wastes less energy so you get higher efficiencies.
good luck!
the max current you can discharge from a capacitor is governed by the circuit's resistance/inductance drop. If you are calculating it stick just with the resistance of the coil to determine the amount of current your switch needs to be rated for.
The absolute max you can draw is governed by the caps internal resistance/voltage it is charged to. I would say yours are about .005ohms, so you at 400v could get (voltage[volts]=current[amps]*resistance[ohms]) 80ka out of them, but that would probably damage the cap. I would suggest you keep it under 1ka per cap.
How long you could draw that current for is limited by the amount of energy stored it the cap, generally expressed in Joules (J=.5*capacitance[farads]*voltage[volts]^2). So each cap has about 300J. Determining how long a pulse you can get is a little trickier, because as you use energy from the cap the voltage drops, so the current drops, etc. For a given resistance you would take
seconds[milliseconds]=resistance[ohms] *capacitance[millifarads]*ln(initial voltage[volts]/final voltage[volts]). So if you wanted to keep the cap voltage above 200v, with a 3.9mf cap, a 1ohm coil and an initial voltage of 400v you would get about 2.7ms. Don't take that number too seriously because it does not take the highly inductive nature of the coil into account, but it will give you an idea of how long (the actual result will be longer)
Using larger wire increases the performance because it lowers the resistance of the coil, which wastes less energy so you get higher efficiencies.
good luck!
Okej, thanks for the help. The IGBTs can only hadle 180A peek current so maybe i should get some other or go for two or three IGBTs per coil to get the maximum out of the capacitors. Stupid to wast enegy left in the capacitors =P
Wonder if my MOSFET/IGBT-drivers can handle three transistors (it's an "IR2121"
)Re: Going to build a Coilgun
Yohan, Mon Mar 13 2006, 08:58PM
Yohan, Mon Mar 13 2006, 08:58PM
or go for two or three IGBTs per coil to get the maximumCareful. IGBT's in parallel will not share the current load equally. I'd be willing to bet that you'll end up with two or three paper weights after your first trial. However, MOSFETS in parallel will divide the current equally! But, finding sufficient current ratings among MOSFETS can get very expensive very quick.
Re: Going to build a Coilgun
Sonny, Tue Apr 04 2006, 11:42AM
Okey. Have been testing the IGBT switches on a 2.5 ohms coil now and it doesn't look to well. 4 of my IGBTs ( ) have stoped working though they have not exploded. And this was only at ~300V. Those IGBT are missing an internal diode so i thought to solder a diode ( ) over the drain-source would do the trick but it didn't. Maybe that diod is to slow? I also got 3 of that type over the coil.
I'm now considering to buy some FETs instead but it's a quite expensive way. Anyone of you coilgun-builders have any good FETs to recommend? Think i have foun two that could be a better idea then my IGBTs.
or
Sonny, Tue Apr 04 2006, 11:42AM
Okey. Have been testing the IGBT switches on a 2.5 ohms coil now and it doesn't look to well. 4 of my IGBTs (
) have stoped working though they have not exploded. And this was only at ~300V. Those IGBT are missing an internal diode so i thought to solder a diode ( ) over the drain-source would do the trick but it didn't. Maybe that diod is to slow? I also got 3 of that type over the coil.I'm now considering to buy some FETs instead but it's a quite expensive way. Anyone of you coilgun-builders have any good FETs to recommend? Think i have foun two that could be a better idea then my IGBTs.
orRe: Going to build a Coilgun
EN, Tue Apr 04 2006, 01:52PM
Actually, you can parallel IGBTs quite successfully, provided that you have IGBTs with a positive temperature coefficent, which i believe the 3rd generation or NPT or whatever IGBTs have.
Mosfets are quite expensive for high current and voltage and have a rather high resistance. So they wouldnt be my first choice.
The diodes across the coil being too slow isnt the reason why the igbts died, as the turn on speed of diodes is extremely fast, even on slow diodes.
Did you have zener diodes from the gate to the source? I think there is a strange effect that when switching high current in IGBTs, the gate raises its voltage above what is applied, and without zener diodes the gate may be overvolted, resulting in fuxxoring the igbt.
EN, Tue Apr 04 2006, 01:52PM
Actually, you can parallel IGBTs quite successfully, provided that you have IGBTs with a positive temperature coefficent, which i believe the 3rd generation or NPT or whatever IGBTs have.
Mosfets are quite expensive for high current and voltage and have a rather high resistance. So they wouldnt be my first choice.
The diodes across the coil being too slow isnt the reason why the igbts died, as the turn on speed of diodes is extremely fast, even on slow diodes.
Did you have zener diodes from the gate to the source? I think there is a strange effect that when switching high current in IGBTs, the gate raises its voltage above what is applied, and without zener diodes the gate may be overvolted, resulting in fuxxoring the igbt.
Re: Going to build a Coilgun
Sonny, Tue Apr 04 2006, 02:32PM
No, i don't have any zener diodes from gate to source. It's the 4th generations IGBTs so i'll try paralell some later. What voltage should i get on my zeners? I'm triggering them with 9V from a 555-timer.
Sonny, Tue Apr 04 2006, 02:32PM
No, i don't have any zener diodes from gate to source. It's the 4th generations IGBTs so i'll try paralell some later. What voltage should i get on my zeners? I'm triggering them with 9V from a 555-timer.
Re: Going to build a Coilgun
EN, Tue Apr 04 2006, 09:10PM
I would use a zener diode with a voltage higher than your gate driving voltage but lower than the voltage at which the gate breaks down. so anything between 9V and 30V (perhaps more).
Just by the way, 9V is on the low side when switching high current. you want to make sure that the igbt stays in stauration so I would recommend using higher voltage (like 15V or even more, but not much more than 30V)
Next step would be to think about switching the gate to -15V for turn off so that it can switch off with less losses.
EN, Tue Apr 04 2006, 09:10PM
I would use a zener diode with a voltage higher than your gate driving voltage but lower than the voltage at which the gate breaks down. so anything between 9V and 30V (perhaps more).
Just by the way, 9V is on the low side when switching high current. you want to make sure that the igbt stays in stauration so I would recommend using higher voltage (like 15V or even more, but not much more than 30V)
Next step would be to think about switching the gate to -15V for turn off so that it can switch off with less losses.
Re: Going to build a Coilgun
Sonny, Tue Apr 04 2006, 09:44PM
Okey, thanks for the info. The -15v part can be a little hard i guess. I got the 555-timer and a gate-driver to play with for each coil. I'm aiming for a battery driven coilgun with a Li-Polymer battery pack. 4 of those cells will give me 14.8V to play with.
I have been testing my 2.5 ohms coil with two FETs rated at 300V 160A peak current each with no problems at all (70+ shots) even tried to charge the capacitors to 315V with no FET break down. Think i'll order some of one of the FETs i meanshend (spelling?) in one of the earlier posts.
EDIT: Can i use those ? Was thinking that the zerner diodes between gate and source maybe will blow when this large current rushing through the FET?
Sonny, Tue Apr 04 2006, 09:44PM
Okey, thanks for the info. The -15v part can be a little hard i guess. I got the 555-timer and a gate-driver
to play with for each coil. I'm aiming for a battery driven coilgun with a Li-Polymer battery pack. 4 of those cells will give me 14.8V to play with. I have been testing my 2.5 ohms coil with two FETs rated at 300V 160A peak current each with no problems at all (70+ shots) even tried to charge the capacitors to 315V with no FET break down. Think i'll order some of one of the FETs i meanshend (spelling?) in one of the earlier posts.
EDIT: Can i use those
? Was thinking that the zerner diodes between gate and source maybe will blow when this large current rushing through the FET?Print this page