New Coilgun Project / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

New Coilgun Project

1 2 3 4 next last

Move Thread

LAN_403

Saz43

Sun Nov 07 2010, 06:04AM

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

Hi everyone!

I'm designing a new coilgun! I'm trying quite a few tricks with this one, such as full/semi auto action and an IGBT half-bridge. However, I definitely need some experts to check my work and let me know if I've got a circuit that will do nothing or burn some very very expensive IGBTs. A schematic is attached, I did my best to sort it out and make it understandable.

I know that the 555 circuit, comparator circuit, and charging circuit work. I'm pretty iffy however on the half-bridge and IGBT driver madness. Here are some questions I have:

1. Is my half bridge circuit correct? Will my IR2110 function the way I have it wired up? I've done my best to understand the datasheet

and IR's application notes

but I have no experience with this IC and few operational circuits to reference.

2. Must D3 and D4 be fast recovery diodes? How can I predict the current through them?

3. I know that the low side IGBT 2 must be triggered in order to charge the boostrap capacitor. Is there a good way to make the capacitor charge as soon as the IR2110 is turned on? What if I put a resistor between the negative side of the cap and ground?

4. The 2N3904 transistor that grounds the gate of MOSFET1: This is meant to turn off the charging circuit when the coil is conducting. When this 2N3904 is conducting, will it mess at all with the HIN and LIN signals going to the IR2110 (shared node)?

5. I'm using MOSFET 2 to switch a solenoid that functions as the injector. Ideally, the injector will fire many times per second. Do I need an anti-parallel diode to protect the MOSFET, since the injector is an inductive load?

Please let me know if there is anything I can clarify. Your valuable help is much appreciated!

GhostNull

Sun Nov 07 2010, 07:06AM

Registered Member #2648 Joined: Sun Jan 24 2010, 12:45PM
Location: Australia
Posts: 291

Comments:
A closed loop optical control circuit would work better than a open loop system. Look here for some good info:

With your schematic I can't really understand the injector/control part. Is the inductor with a core is the solenoid and you are coupling it some how it the 555? Your IGBT symbols don't look right too =S

Answers:
1. The diagonal half bridge looks good but I have no idea about the driver
2. Not sure but recovery time shouldn't matter too much. If it is slow then it might put some stress on the IGBT and lengthen the pusle but if it is faster than the switching times of the IGBTs it won't really benifit.
4. How about putting a complementery pair or putting it in series with R1 and using a pull down resistor instead?
5. Yes, if you are hard switching an inductive load, you need to put a flyback diode in anti parallel with inductive load. Just like the main coil of a coil gun

That's about all I can say with my knowledge =S

Saz43

Sun Nov 07 2010, 02:21PM

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

GhostNull wrote ...

Comments:
A closed loop optical control circuit would work better than a open loop system. Look here for some good info:

Thanks Ghost. Actually, the 555 circuit is closed loop. The plunger of the injector has a 3V contact on it which triggers the 555 when the plunger is forward. An IR diode/photo transistor grounds the reset pin, stopping the pulse when the projectile is in the center of the coil. I know it looks a little confusing on the diagram but I tested this out in LTspice and it all works just fine.

Can you explain what you mean by complimentary pair?

GhostNull

Sun Nov 07 2010, 09:39PM

Registered Member #2648 Joined: Sun Jan 24 2010, 12:45PM
Location: Australia
Posts: 291

Sorry,
A bit hard to see in that schematic.

With the Complimentary pair, I might hace used the wrong word there =S look at this thread:

So you have a NPN on gate drain and it's complentary PNP on the gate source. Both the bases of the transistors are common and input. When positive is input the NPN turns on and charges up the MOSFET gate, when negative is applied the PNP turns on the discharges the MOSFET gate

Saz43

Wed Nov 17 2010, 03:09AM

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

Saz43 wrote ...

3. I know that the low side IGBT 2 must be triggered in order to charge the boostrap capacitor. Is there a good way to make the capacitor charge as soon as the IR2110 is turned on? What if I put a resistor between the negative side of the cap and ground?

I've stumbled upon an answer to my own question, which I thought I would share with you all. Here's a rather ingenious way to charge the boot strap capacitor automatically in between shots when the IGBTs aren't conducting. I found this on google weeks ago and just today realized what it does!

Can anyone comment on the purpose of D2?

Saz43

Fri Nov 26 2010, 09:08PM

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

Here's the latest schematic. I'd really appreciate any feedback and/or corrections.

Things I need help with (marked in red on the schematic) are:
1. I don't know how to chose gate resistors for the MOSFETs (the two 500 ohm ones in red)
2. I don't know how to chose the external capacitors for the IR2110 (except for the 100uF bootstrap cap)
3. Do the half-bridge diodes have to be the fast type? My coil current pulse is about 2-3ms long and rate of fire will be 15 RPS at the lowest energy

The diodes, coil, IGBTs, and other stuff without labels are components that I haven't picked out yet, or that I have lying around on junk boards.

Since there's a lot going on in this diagram for a coilgun, here's how it's supposed to work:

1. The AC inverter powers a voltage doubler which charges the 1510uF capacitor bank. Charging continues until the comparator (upper right) shuts off the MOSFET connecting the inverter to ground. The shutoff point can be changed via the 10k pot. Programmed range is 200-450V. At 200V, theoretical RPS is 15, at 450V, 4.

2. Close the safety and pull the trigger. If the capacitor bank is fully charged, the MOSFET connecting the injector to ground is conducting and the injector fires.

3. The injector has a 3 volt contact on it which, when it's far enough forward for the projectile to be in the optimum starting position, triggers a pulse from the NE555.

4. The pulse from the NE555 does the following things:
a. Triggers the IR2110 half bridge driver, which opens the IGBTs and fires the coil. The IR2110 low-side output (LO) also turns off the small NPN transistor that allows the bootstrap capacitor (100uF) to charge. This is necessary to prevent the firing capacitor from shorting through the high side IGBT.
b. Turns off the MOSFET powering the charging circuit, so it doesn't short through the coil.

5. As soon as the IGBTs are triggered, the firing capacitor discharges through the coil and the voltage drops, causing the comparator output to go high, turning off the MOSFET powering the injector. The injector (equipped with return spring) retracts, ending the trigger signal to the NE555 and mechanically preparing to load another round.

6. When the center of the projectile is at the center of the coil, the IR diode/photo-transistor path (located at coil entrance) is cleared, the photo-transistor conducts and grounds the reset pin of the NE555, ending the pulse and closing the IGBTs. The inductive kickback spike is piped back into the capacitors via the two half-bridge diodes.

7. Once the NE555 pulse is terminated, IGBTs are shut off and the charging circuit starts up again. The gun cycles.

8. Closing the switch that bypasses the injector MOSFET allows for semi-auto operation, where the gun fires as fast as you pull the trigger, independent of the capacitor voltage (injector not inhibited by the comparator via the MOSFET).

Also, a rough drawing of how I'd like to see it come together.

GhostNull

Sat Nov 27 2010, 11:34PM

Registered Member #2648 Joined: Sun Jan 24 2010, 12:45PM
Location: Australia
Posts: 291

1. MOSFET gate resistors might not be nessary. MOSFET gate resistors are used to prevent the parasitic oscillcations caused by stray inductance and the MOSFET gate capcitance when hard switching the MOSFET gate. If you are switching at low frequency and ringing is not a problem, or if you are soft switching, then a gate resistor will not be nessesary.

If gate resistors are nessesary you need them to be great enough to stop ringing but not so great that they slow turn on times to a noteable degree and/or cause the MOSFET to dissipate a large amount of power. Look at this

to see effect of a gate resistor.
500ohms would most likely be more than enough to stop ringing.

3. They need to be similar to the flyback diodes you would put on a normal coilgun solenoid. I remember seeing a thread where this was disscussed, try having a look around

Suggestion:
I suggest you should add some hysteris to the charger control so it isn't constatntly turning on and off.

That all I can help you with :D

Cheers
-Ken

Saz43

Sun Nov 28 2010, 06:01AM

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

Thanks again Ghost. I took what you said into consideration, including what you said earlier about the complimentary pair of transistors controlling my charger MOSFET. I designed a MOSFET triggering circuit that uses less power and switches the MOSFET two times faster than the initial design, thus less waste heat. Most importantly, with this circuit quickly turning off the charger, the IGBTs can switch much more quickly without the charger shorting through the coil.

Here's the latest schematic, IGBTs chosen and gate and base resistors now included. The MOSFET powering the injector is still wired to open more slowly (with IRFP450, 64nC/(14.8V/2000ohms) = 8.65us) because power is lower and switching speed isn't too important here since it's not racing the IGBTs.

ben5017

Fri Jan 06 2012, 07:40PM

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

Just came across this thread. I am working on a similar project except mine will be multistage. did you ever finish it or revise your latest schematic?

Saz43

Sat Jan 07 2012, 06:09AM

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

I actually scrapped this and did a total redesign. After completing the above design on paper and spending about $200 on parts, I found that a single stage, high voltage, capacitor driven coilgun just was not effective at reaching full-auto performance. My new design is an 8 stage, 50 volt LiPo battery driven coilgun. I'm also no longer using half-bridges in favor of single IGBT switching with series diode commutation, which gives most of the benefits of the half-bridge design but with far less complexity and cost.

Currently the project is maybe 15% through the testing and construction phase, with estimated completion by the end of 2012! In the process of creating the last design, I learned a ton about IGBTs and half bridges so if you have questions please ask.

1 2 3 4 next last

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