V3 Coil Carbine / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

V3 Coil Carbine

first previous 12 13 14 15 next last

Move Thread

LAN_403

ben5017

Wed Oct 09 2013, 12:30AM

Registered Member #3315 Joined: Thu Oct 14 2010, 04:23PM
Location:
Posts: 156

12 Kohm between collector and emitter when gate voltage applied.

gate voltages used. 6 & 12 & 15 (when 6 didnt work)
do I need a resistor and/or diode between gate and emmitor?

IGBT: IXYS IGBT IXGN200N60A

(I wish they would tell you what the suffixes on the datasheet stand for)

By "on" I was looking for nearly no resistance (mohm) between collector and emitter. I had no trouble when driving the 422 off of a 555...

Saz: it does light up LED, I also tried my injector and nothing happened but a blown fuse after a second or so. there is to much resistance to get a magnetic field strong enough to move my bolt inside the injector.

DerAlbi

Wed Oct 09 2013, 01:35AM

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

ben... you are using IGBTs. not Mosfets.
why do you expect anything even close resistve behavior!?

...well small signal behavior suggests 17.8mOhm with 10V @ Gate. but YOU cant measure that.
What sufix do you mean? The "A" is descibed in the datasheet first page..................

Yandersen

Wed Oct 09 2013, 10:38AM

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

Funny, ben. The device you use rated 600V, 300A peak. The voltage drop is 2.5V. On digikey it costs $38.
Let's calculate the equivalent resistance of it when 300A conducted:
2.5V/300A = 8.3mOhm.
Now, the voltage rating of the switch must exceed the supply voltage 2 times. It means, that 100V rated device is good enough for you, ben.
So let's see what would be the MOSFET options. Here is the 100V rated MOSFET equivalent for $1:

Here is the 2 times better option for $4:

And 3 times better option for $7.4:

The MOSFETs linked are much smaller in size and have also lower driver's requirements then your IGBT. And unlike with the IGBT you can overload those higher then the pulse rating specifies given the pulse is shorter as MOSFETs do not latch. From the other hand, on low current the MOSFETS will show lower voltage drop while with IGBT you will always have 2.5V.

Steve Conner

Wed Oct 09 2013, 10:53AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

An IGBT has a junction, so it doesn't have a Rds(on). It has an exponential voltage drop like an ordinary transistor, starting out about 0.5V for small currents and heading up logarithmically from there.

The lowest voltage rating of IGBT you can buy is something like 250V. Below 250V, MOSFETs work better.

Modern IGBTs are latchup free. One popular device is rated at 600V, 110A continuous, 300A pulsed at 25'c case temperature. In Tesla coil work we run it off a 320V DC bus and overload it to 450A or even 600 if you are feeling lucky. You would need a huge array of 600V MOSFETs to match this.

ben5017

Wed Oct 09 2013, 03:16PM

Registered Member #3315 Joined: Thu Oct 14 2010, 04:23PM
Location:
Posts: 156

YAN: Don't use digi-key unless you have too. Kind of a rip off. FYI I got 20 of the IGBT's for 200$ on ebay.

from your math, I concluded my 12 Kohm is reasonable when only conducting the current of a single LED

2.5V/xA=12000ohm
x= .2 mA (for single LED)

(I could have sworn on my test setup when driving it off 555 & 422 that resistance was in mohm region with no load besides the multimeter??)

The reason I got higher voltage rated IGBT's (600v even though im powering my coils off of 24V) was to be able to withstand a larger voltage spike, allowing me to use more TVS diodes in series to kill the current as quickly as possible minimizing suck back.

STEVE: Can you explain further the significance of â€œRds(on)â€ for a newbie and is my above reasoning correct?

DerAlbi

Wed Oct 09 2013, 05:10PM

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

ben, you are completely off. sorry for that. your 200$ are spent compeltely wrong.
this is a nightmare in a 24V-System to have IGBTs.
You obviously are not interested in how things work. This is why you kind of deserve that desaster. Just using google and reading about IGBTs would have helped. Yes even for a starter - Yes its work, Yes it hard. Yes is important.
You seem to look at a datasheet and every charactierized value you dont understand seems not existant for you. so you ignore e.g. Vcesat instead of googling it and this brings you here.
You hold your multimeter somewhere in the circuits and getting some values. what for?

You would be much better with simple IRF1404 using active clamping (or let the avalance in the body didoe take care of). Did you calculate wow fast the current decays if you allow 80V? Was the value really that bad?
"the higher the better", right? Sounds cooler at least -.-

And PLEASE stop measuring resistance over a transistor. its a completely stupid value. it tells you exactly nothing.
what bring you to this idea? can you explain how an IGBT work, and why it would be usefull to measure resstance between C and E ??

When you finally show your ultra hyper cool thing on youtube or be the biggest guy in town for your friends, will you mention that you dont even know how the thing works? -.-
"somethihg switching in there somehow maybe a bit for some time"
I will laugh if you loose all your coolnes when finally someone asks why you didnt use a potentiometer with a verry fast motor if you just put in variable resistances in your design...

Sorry for that. And: Rsd(on) is verry well documented by google.
look what i have found on the first few links after googling for "rsd(on) mosfet"

really cool, issnt it? Takes much less time than waiting for a forum answer too.

heeey and look what i found when googling for "mosfet versus igbt".
First link:

And you know... if sometimes the first link issnt satisfiying, maybe reading it will help too. The things you dont understand will cross you at your search many times and then - when reading enough - you will get a picture of how things work. there are also good youtube videos.
But yes, all this takes time. Time you will not be admired by your buddys for beeing the cooles guy with the hyper ultra electrical cannon that spits out a iron projectile.

if i tell you now that Rsd(on) is the resistive value of a mosfet between Soruce and Drain when its fully on, so that adding extra gatevoltage will not dreacease Vds anymore, you will get the meaning of Vds, but you miss the bigger picture that would help you really to understand things.

Vcesat is the same thing for Transistors. Its the Vce-Voltage that is reached as soon as the basis-current is high enough so that Vce will not decrease anymore. Your IGBT has a 2.5V Vcesat. So it doesnt mater much, how much current you apply you will always measure the voltage around 2.5V. Sometimes less, (low current) somtimes more (high current)

End this is why measruring resistance is useless. you always have something around 2.5V. So you apply 15V at the gate, and measure resistance with 10mA. you get 250Ohm, you measure it with 10A, you get 0.25Ohm. you use 100A... and so on.
Its always the same Transsitor. so what physical meaning has this resistnace??? NO MEANING AT ALL.

The onyl thing is certain: by using IGBTs you waste 2-3V of your Battery-Voltage in your IGBT. This is a expected loss >10%. This will lower your effective Voltage at the coils -> you need lower inductance -> less turns -> less force -> less speed.

This is why IGBTs are high voltage devices. 2.5V taken from 400V dont matter much.

READ. start at wikipedia, then search for Application notes. Then start with wikipedia again when you encoutner something you dont understand. Forums arent that effective to read, because the content is often shitty.

ben5017

Wed Oct 09 2013, 06:01PM

Registered Member #3315 Joined: Thu Oct 14 2010, 04:23PM
Location:
Posts: 156

Mostly fair enough. I am learning as I go, by the end of the project I will have a better grasp on how everything interrelates.

However I think "nightmare" is a little strong. I used the same IGBT's in my test rig with satisfactory results.

I will adjust my coil geometry to take into account the 2.5 voltage drop and still end up maxing out the pulse discharge of my lipo.

Yandersen

Wed Oct 09 2013, 07:53PM

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

Why don't you simply buy some MOSFETs instead? Can't spend extra $20 for more suitable devices? If your current is limited by the coils' resistance, then optimizing the speed of current decay by 600V is so nonsense...

DerAlbi

Wed Oct 09 2013, 08:32PM

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

Well learning by making mistakes is ok so far. Even yan makes mistkes

*troll*

If the results are ok for you then its fine. However keep in mind that there are other posibilities that dont require that much amount of money to achieve the same result OR spend the same amount and get even more "satisfactory" results. .... i must confess that even my charger costs more.
Have fun so far

Just a tip: if you spend your LiPo a big cap, maybe the voltage will not break down so much due to internal resistance

so you can compensate some effects...

ben5017

Wed Oct 09 2013, 10:24PM

Registered Member #3315 Joined: Thu Oct 14 2010, 04:23PM
Location:
Posts: 156

YAN: I probably will. Coil Inductance plays a larger role than resistance in limiting the current true?

any tips/ things to look out for when paralleling mosfets?

ALBI: Its all good. I just dont make the same mistake twice. + its not like i wont use the igbts for something ;) ,especially for the good deal I got. Mistakes are inevitable in my first serious electronics project. just hoping this will be the biggest one of the project.. (fingers crossed)

first previous 12 13 14 15 next last

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