IGBT ratings/magnetic projectiles / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

IGBT ratings/magnetic projectiles

previous 1 2 3 next

Move Thread

LAN_403

Yandersen

Thu Dec 20 2012, 09:52PM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

OMG...
RTFM for any of the SCR you have. Please, at least once. Or check the Wiki on those devices.
SCR can be considered as two diodes: common cathode goes toward the ground, power anode - to the positive voltage level of the supply, gate is the supplementary anode. When you let current go from gate (+) to cathode (-) then power diode turns into conductive state and latches untill the current stops running from anode (+) to cathode (-).
If you check your SCR with ohmmeter, you will discover short circuit between gate and cathode. It is a protection resistor built inside the device in parallel with imaginary gate-cathode diode which ensures no unwanted current goes through that diode so device will not turn on accidentially. So the safest way to trigger SCR is to have separate battery connected minus to the cathode and plus to the gate via switch and resistor. If you want to use the main power supply to trigger the gate through a resistor and a push button, then make sure cathode is connected directly to ground. If you put a little cap in series with resistor, then you can make a single pulse to the thyristor instead of flushing unnecessary current through the gate all time push-button is held.

If you put a coil between cathode and ground, you still will be able to switch the SCR on, but right after that cathode will fly up to the HV+ resulting in negative voltage bias between gate and cathode which supplementary diode can not sustain (only up to -10V is allowed).

Maxwell

Fri Dec 21 2012, 12:44AM

Registered Member #8497 Joined: Tue Dec 04 2012, 06:24PM
Location:
Posts: 74

So basically you're agreeing with everything I've said... Because you have just described my above circuit with the exception of a the switch - which I simplified as a +5V pulse.

Jack A

Fri Dec 21 2012, 02:13AM

Registered Member #2975 Joined: Wed Jul 07 2010, 12:19AM
Location:
Posts: 28

Okay, I tried to work it out myself, and was told by other threads that the only limiting factor is the heat dissipated.
Since the total energy passing is relatively low, and this IGBT has low voltage drop and high max heat dissipation, I see no reason why this IGBT (http://australia.rs-online.com/web/p/igbt-transistors/0192613/ or IXGP16N60B2) is not suitable (with a big heatsink).
<ended up buying this igbt

>

Back to the issue of driving the IGBT; can I run it by delivering 14.6V from a battery through a MOSFET controlled by an Arduino?
<you can't, the igbt needs to be driven back to zero to switch off>
<no longer relevant> The gate is just a capacitor right?

Cheers,

Yandersen

Fri Dec 21 2012, 03:21AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Jhackulon, forget about heatsinks in coilguns. Pulse time is so small that heat does not reaches the heatsink anyway - either overheated semiconductor crystal blows, or everything works cool. Charger may have heatsinks, but not shooting circuit.

Maxwell, buy a glasses and reread. You have cathode (C) and anode (A) messed up, and 100K resistor is not necessary. And 200Ohm may not give enough current to trigger power SCR like 70TPS12. Put 51Ohm instead. After that I will agree. And I would recommend to put paralleled 1MOhm resistor and 1uF cap (plus to the +5V) in series with the 51Ohm resistor - this will save a lot of energy for the Universe.

Jack A

Thu Jan 17 2013, 06:36AM

Registered Member #2975 Joined: Wed Jul 07 2010, 12:19AM
Location:
Posts: 28

Ok, so on stage 1 we will have a maximum of about 700A, this IGBT (

has a pulsed rating of 160A, two in parallel would have a pulse rating of 320A, which should be sufficient due to the very short pulse. As far as triggers go, how does this optocoupler (

look for being controlled by my arduino and switching 14.7V from my battery (at this voltage, the collector-emitter saturation voltage is only 1.7V)?

<I have given up, rightly so apparently, on overamping the switch, as described below>

Cheers

Ternal

Sat Feb 02 2013, 02:54PM

Registered Member #9893 Joined: Tue Jan 29 2013, 04:04PM
Location: Belgium
Posts: 1

Half bridge igbt up to 300A peak /1200V.
This is an old one though,I got it from a frequency modulator for elevator motors.
Pulsed 500A through it from 10 000ÂµF cap charged at 330V.
Igbt's dont like higher pulses, SCR's are quite abuseable though at least from my expierence.

DerAlbi

Sun Feb 03 2013, 03:20PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Jhackulon: paralleling IGBTs is a verry hard problem to solve by its own. In no way you can expect that the current you can handle rises linear with device-count.
At least your IGBT has a positive temp-coeff - however this only solves a small part of the problems.

Personally i would recommend not to exceed twice the rated continous current per IGBT and stay far far away from its pulse-ratings in a parallel config.

The real problem is the switch-off... timing differences will draw all the inductive current from the last active device -> internal SCR will latch up.. and BOOM

To share the switch off current properly you might consider:
1) ferrite bead on collector-leg of every IGBT.
2) active clamping
3) lowest possible gate voltage. Even 2) may not cause an excess of this voltage. This prevents SCR-latchup. (Thats what caused Ternals release of the magic smoke)

Your devices are verry fast and switching differences arent mentioned in the datasheet :-/ Maybe for a small setup you get away with the ferrite beads... i would be interested in the results and how much you are able to switch in the end. But looking at the Vg_th voltage.... it strays like hell.. good luck^^

Remember that every IGBT needs its own gate resistor.

Jack A

Tue Feb 05 2013, 10:48AM

Registered Member #2975 Joined: Wed Jul 07 2010, 12:19AM
Location:
Posts: 28

Okay, I've come up with a compromise. For my preliminary circuit testing, I'm going to add resistance to the circuit, up to about 1-2 ohms. This will keep the max current at about 200-300A (my IGBT has a pulsed rating of 280A) for a lower voltage of 310V.
In terms of drivers, is a simple IC easy to set up?
I found this driver IC (

), will this accept a PWM signal from my arduino? How do I wire the IC to allow for quick charge and discharge of the gate capacitor? Or is the example circuit on the last page of the datasheet acceptable?
I need some help with this switching stuff, it's very unfamiliar to me.
Cheers

DerAlbi

Tue Feb 05 2013, 08:36PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

This whole issue is annoying as, while I have the rest of the circuit ready to go, the issue of switching isn't going away by itself.

You build everything except the only thing that is tricky?

Well^^ maybe you should think about yourself. Pushing away the hard problems will never work in the bigger picture. You gonna mess up every project with this attidute, since your design ist not problem-oriented - how should it if you dont think about it in the first time. Dont get me wrong.. noone can know everything... but you knew you want to switch, and even you knew you dont know how, you decided to put work in everything else - why? There is everytime the big chance you fail^^. hmmh.
The whole thing sounds like a bigeyed kid saying "i want to build something, but i dont kow how, so please build it for me, so i can show it to my firends and claim its mine"

So. Back to [annoying] Topic^^

I'm going to add resistance to the circuit, up to about 1-2 ohms.

Remember to put the Resistor in series with the coil directly! Dont put it in the Collecot-Path only, since small power-resistances are often wound wire - so its inductive - this inductance needs to be dampened as well.
Btw: its verry well possible, that the resistor blows immediately. Hotspots in the material can lead to some ugly effects. I _could_ be an other design-flaw.

In terms of drivers, is a simple IC easy to set up?

Yes of course, if you know what you are doing^^ So.. No

How do I wire the IC to allow for quick charge and discharge of the gate capacitor?

Didnt you just linked in a Datasheet?

Did you read it? No? Why not? GUY... YOU DIDNT EVEN LOOK AT THE PICTURES

Quick tip: page 15.

will this accept a signal from my arduino?

Well... i recommend !!NEVER EVER!! look on page 6 of the Datasheet. Seriously. Dont do it.
It will accept it, but it might not react to it. Depends on VCC.

but this switching stuff seems a little unfamiliar to me

So.. just start with a coilgun...
High current, big explosions; high woltage, big shocks! Perfect to get a *feeling* how switching works.. i would wish i learned the basic the same way

Sup guys, I am building a proper coilgun

Yes, you do.

Sry i am a little pissed -.- your questions show that you put only minimal effort into the problems - barely enough to come up with these questions.. You are able to build things that are already designed - thats ok. But you fail misserably when its comes to show willingness for _understanding_ what you do. This is disrespectful to everybody helping.

Yandersen

Wed Feb 06 2013, 12:14AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Propeller coilgun maybe?..

previous 1 2 3 next

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