Why am I burning my IGBTs ? / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

Why am I burning my IGBTs ?

previous 1 2

Move Thread

LAN_403

sofi

Sun May 18 2008, 09:16AM

Registered Member #1464 Joined: Sat May 03 2008, 10:56AM
Location:
Posts: 9

400A peak current? What IGBT are you using? (for min. 400V and 400A peak it must be a huge block and in addition extremly expensive - min $200). Are you sure that current through coil is 400A? Or at other side - are you sure that your IGBT can handle 400A? Is it realy IGBT?

Post every detail, for example coil resistance, inductance (or dimensions and wire gauge), IGBT number or type.

But I don't understand why you want to use IGBT, when you driving gate by mechanical switch? Device like IGBT (or MOSFET) allowing opening and closing at any moment (by superior switching logic), but in your design you only opening. SCR is best choice in that case.

Dago

Sun May 18 2008, 10:54AM

Registered Member #538 Joined: Sun Feb 18 2007, 08:33PM
Location: Finland
Posts: 181

If you are using a normal switch you really need to debounce it or you will switch the IGBT on and off during the switching state because of switch bounce and that may well be the reason why your IGBTs are burning up.

magictomcat

Sun May 25 2008, 06:22PM

Registered Member #1456 Joined: Tue Apr 29 2008, 09:02AM
Location:
Posts: 9

Hi,

(I apologize for replying so late, I wanted to make a few tests before.)
I think I am going to use an operational amplifier instead of a simple switch to control the gate of the IGBT. When it toggles from -12 V to +12 V, it doesn't deliver a perfect square, but rather an exponential signal lasting 0.2 Î¼s. But there is no bounce effect.

I will try my coil gun that way, and I will let you know how it works.

(I really have an IGBT supporting 400 A in peak current. It is indeed quite large, and expensive (but I bought a second-hand one))
Thanks anyway

teslacoolguy

Mon May 26 2008, 01:33AM

Registered Member #1107 Joined: Thu Nov 08 2007, 10:09PM
Location:
Posts: 792

For the money you would spend on a igbt like that why dont you spend the same amount on a hockey puck scr that is rated for 10+ ka?

Quantum Singularity

Mon May 26 2008, 01:42AM

Registered Member #158 Joined: Sun Feb 12 2006, 09:53PM
Location: Central Ohio
Posts: 282

sofi wrote ...

400A peak current? What IGBT are you using? (for min. 400V and 400A peak it must be a huge block and in addition extremly expensive - min $200). Are you sure that current through coil is 400A?

You can get the larger brick igbts on ebay 400A+ for a lot less than $200, heck I was bidding on a 1200V 600A one but it went for $38... to much for me lol!

edit-I see TCguy posted while I was typing, and I agree and thats one nice thing about SCR's, but for a reluctance type coilgun the ability to turn it off is very beneficial which you would need the igbt, of course if your not doing that though there is really no reason I can think of not to just use an SCR instead unless you already have the igbt and its big enough.

sofi

Mon May 26 2008, 12:21PM

Registered Member #1464 Joined: Sat May 03 2008, 10:56AM
Location:
Posts: 9

Sorry, my price estimation was created from e-shops price lists at small quantities (not ebay, or stock-market) - at 10 pieces is only $150/piece.

Anyway, I build much smaller handheld CG; that large IGBT need strong and complicated driving circuit (about 10A pulses to gate: +-15V, Rg=2ohm, to preserve microsecond switching times). I think that output from OPAMP is really unequal. OPAMP + 2 power transistors connected as push-pull = better solution.

I have optically triggered stages - detector is connected to fast OPAMP (comparator with hysteresis), output from comparator is connected to IGBT driver and driver is connected to halfbidge. This works well and safe.

magictomcat

Tue May 27 2008, 09:22PM

Registered Member #1456 Joined: Tue Apr 29 2008, 09:02AM
Location:
Posts: 9

i got the igbt for 30euros
Why should i use a resistor on the gate?do i really need a driver?can i avoid using one?
Thx

sofi

Thu May 29 2008, 08:28AM

Registered Member #1464 Joined: Sat May 03 2008, 10:56AM
Location:
Posts: 9

As I wrote, you don't need driver IC, you can use a simple circuit like this:

short description:
If you push the button, current start flow through C4 and R3, while C4 is fully charged (and it is charged really fast). This current will cause short voltage pulse on NE555 triger. NE555 is connected as monostabile (after it is triggered, generate one impulse). Impulse lenght T = 1,1*R1*C1 (see schematic).
This pulse opens main IGBT through NPN and PNP transistors (push-pull) and therefore define how long time will be projectile draged into coil (typically few miliseconds) - experiment with values for best results.

part of schematic in dash line maybe have not suitable values... is need to try

if I forgot something or I have some mistake, please anyone report

previous 1 2

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob