New Coilgun Project / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

New Coilgun Project

first previous 13 14 15 16 next

Move Thread

LAN_403

Saz43

Fri Jan 18 2013, 05:53PM

Registered Member #1525 Joined: Mon Jun 09 2008, 12:16AM
Location: America
Posts: 294

o yea

DerAlbi

Sat Jan 19 2013, 02:47AM

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

Your simulation with this diode is ridicules^^ Did you ever had a look at the model parameters? 0.56Ohms junction resistance??? Never ever!! this simulation result is anywhere near reality

Please try again with 4x "MBR745" and report back

It has 20mOhm junction resistance which is very much closer to the powerdiode specs of your real world design.

The parasitics: its just a good gues on what short cables and solder connections can do to your design. Its nothing specific, however since the parasitics are there is much better to include them as not to include them - although they are not correct, the real result will benefit if you design your precautions with larger parasitics

Please consider your circuit again with simple math:
1N914: 0.56Ohms*4*240A = 540V + 50V Suply = 600V.
Again with 20mOhm: 0.02*4*240A = 20V -> 70V at your "IGBT". Much better isnt it? Of cource this is only true if you DO NOT use 200mA rated small signal diodes

And you will not use them, right?

Saz43

Sat Jan 19 2013, 04:40AM

Registered Member #1525 Joined: Mon Jun 09 2008, 12:16AM
Location: America
Posts: 294

How do you tell the junction resistance of a diode?

MrFlatox

Sat Jan 19 2013, 09:31AM

Registered Member #9349 Joined: Mon Jan 07 2013, 08:50AM
Location: France
Posts: 102

I am not sure, but your resistance junction can be calculated from the Uforward-Iforward chart from your diodes datasheet. On this graph, R is the "slope rate" (not sure about the word though, I only know few words about mathematical english), that you can obtain by this derivative formula : R = dU/dI.

Hope it's true, and hope it cans help.

Saz43

Sat Jan 19 2013, 09:56AM

Registered Member #1525 Joined: Mon Jun 09 2008, 12:16AM
Location: America
Posts: 294

MrFlatox wrote ...

I am not sure, but your resistance junction can be calculated from the Uforward-Iforward chart from your diodes datasheet. On this graph, R is the "slope rate" (not sure about the word though, I only know few words about mathematical english), that you can obtain by this derivative formula : R = dU/dI.

Hope it's true, and hope it cans help.

Oooh, I see exactly what you mean. Wow, the diodes I'm using are definitely in the 10's of mOhm range. It was just an unlucky chance that when I made my simulation I happened to randomly pick diodes with ridiculously high resistance.

Fortunately I think it should be easy to change this part of my design. Thanks DerAlbi and Flatox, hopefully this will help improve my coilgun.

MrFlatox

Sat Jan 19 2013, 10:23AM

Registered Member #9349 Joined: Mon Jan 07 2013, 08:50AM
Location: France
Posts: 102

So, with the corrected value of diode junction rsistance, what is the voltage spike your IGBTs are seeing ?

Saz43

Sat Jan 19 2013, 10:52AM

Registered Member #1525 Joined: Mon Jun 09 2008, 12:16AM
Location: America
Posts: 294

DerAlbi was right, it was 74V according to the simulation.

What I'm going to do instead is pull out three of my diodes and replace them with a 3 Ohm power resistor. So I'll have one diode and 3 ohms in series. This raises the voltage right to 600V and in theory the current decays in 0.22ms.

The 4 diodes in series I *was* using looks like it's taking more like 1.63ms to decay the current.

DerAlbi

Sat Jan 19 2013, 11:32AM

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

Yes using a resistor is a good thing to speed up the currend decay. However your design becomes verry unflexible with this in terms of maximum current and pulse duration. But if designed correctly... cool

You know what? I thought abougt this resistor aided thing a while ago and came up with the idea that one could use thin copper wire to achieve this resistance. This thin wire can then be wound in a bifilar way to the accelerator coil. This would cancle out immmediately any inductance with absolutely no suckback.. and if the wire is thin enough it should fit into the unused space between the windings..
... just a thought^^

MrFlatox

Sat Jan 19 2013, 02:30PM

Registered Member #9349 Joined: Mon Jan 07 2013, 08:50AM
Location: France
Posts: 102

Interesting idea.

If I understand you well, your idea is to put a wire wound resistor in series with the anti parrallel diode, and wind this resistor around the main coil, but in a reverse way ? Is that it ?

Ash Small

Sat Jan 19 2013, 03:33PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

MrFlatox wrote ...

Interesting idea.

If I understand you well, your idea is to put a wire wound resistor in series with the anti parrallel diode, and wind this resistor around the main coil, but in a reverse way ? Is that it ?

I think he means that you require a certain length of wire to achieve the required resistance. If you 'bend the wire in half', then twist it together, you cancel out any stray inductance, at least for the length that is twisted together, you will still have the two 'ends' that connect between diode and coil, but stray inductance will be considerably reduced.

At least, that's what I think he means.

first previous 13 14 15 16 next

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