Controlling SCR from Microcontroller? / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Controlling SCR from Microcontroller?

1 2 next

Move Thread

LAN_403

guitardude012

Tue Sept 09 2008, 04:11AM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

I have drawn this circuit minus the protection pieces, charger, etc. for simplicity. Also I would use a different diode for D1 but these are the components I found in eagle. Maybe a 12V battery would work better as well.

Would this work or am I completely off?
mod edit - oversized pictures

Barry

Tue Sept 09 2008, 04:44AM

Registered Member #90 Joined: Thu Feb 09 2006, 02:44PM
Location: Seattle, Washington
Posts: 301

Good thought Josh, now all you have to do is show the ground reference connection that allows IC1 to turn on transistor T1. The coilgun can't be floating, we need pin 12 to pull the base of T1 high and low.

As a result, for this to work, your circuit will connect the T1 emitter to the microprocessor ground. In turn, the SCR gate is then at ground. Do you really want to do that? It's a bit odd, since the microprocessor power can't share the same ground reference as the coilgun charger. This can make it hard to connect scope probes around there, too.

Also I'd be concerned about ringing on this ground plane during discharge causing problems to the micro. Not to mention, some component failures might smoke IC1. I haven't thought it through at all, and in theory your 1n4004 protection diode should prevent it, but I wouldn't bet my scope on it.

An optoisolator seems like a good idea. Anyone know if they're fast enough for this application?

Cheers, Barry
Bad grammar makes me (sic)

guitardude012

Tue Sept 09 2008, 04:50AM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

Thanks for the quick input Barry.

How about this schematic including your suggestions? Again drawn without charging circuit and most protection components.

edit: The optocoupler's input should also be connected to 5v via a pullup resistor and a toggle switch(normally disarmed).
edit again: Maybe it would be better placed on the 12V supply to the PNP

Steve Conner

Tue Sept 09 2008, 10:10AM

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

Cheap optoisolators run about 1 microsecond response time. The fastest one I can think of is the 74OL6000 that can achieve about 20 nanoseconds. But given that coilguns work in millisecond time scales, any old opto should do.

Don't forget to draw the ground reference for your 12V supply too! :P

guitardude012

Tue Sept 09 2008, 09:19PM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

Thank you Dr. Meh. I have cleaned the schematic up a bit and made that change:

guitardude012

Thu Sept 11 2008, 12:56PM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

I just found a cool optoisolator: HCPL-3120-000E by Avago It is an IGBT driver supporting 2.0A normally(2.5A peak). I see they even have a 5A version. It is rated to 500ns max switching time.

I am going to try it because it will allow me to remove R1,R2,and T1.

aonomus

Thu Sept 11 2008, 02:51PM

Registered Member #1497 Joined: Thu May 22 2008, 05:24AM
Location: Toronto, Ontario, Canada
Posts: 801

Correct me if I'm wrong, but since SCR's are avalanch devices, you pulse a quick voltage between gate and cathode and the SCR opens and stays open until voltage drops below a certain point for a required period of time. Because an SCR isn't like a MOSFET or IGBT having the large current capacity in the opto might not be so critical since you don't have to hold the gate open.

Also, on that thought you won't be able to control the length of each SCR's triggering, just the start of the pulse. The optos are still a good idea because of all the potential induced current into your low voltage/logic PSU.

guitardude012

Thu Sept 11 2008, 06:40PM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

I agree that this amount of current is excessive, but I am looking at a requirement of ~400mA pulse (100mA/SCR * 4 SCRs) and I needed an optoisolator anyways.

Also, because I am using a v-switch, I will be able to control SCR triggering right?

I must let the second capacitor drain very low but I can :
1)choose the on time of the first capacitor
2)increase max current for more power with the second capacitor(while turning off the first)
3)drain the second capacitor quickly, which becomes a very fast near constant amount of time(given the same initial voltage on the quenching capacitor). In my case I will aim for a turn off time of 50uS.

Ken M.

Thu Sept 11 2008, 11:48PM

Registered Member #618 Joined: Sat Mar 31 2007, 04:15AM
Location: Us-Great Lakes
Posts: 628

Why 4 100mA scrs and not a 1A or something, attempting to switch 4 scrs at the EXACT moment is said to be iffy.

guitardude012

Fri Sept 12 2008, 12:46AM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

I meant 4x100mA gate pulse. I am switching ~1500A peak and require the 35uS turn-off time you can't find in stud scr's.

edit: Besides I have a little bit of time to trigger all 4 and they should handle current trade-off nicely due to the positive thermal coeff.(One will take more current, heat up, and another steps in, etc)

Even if one is only taking 5% of the current it is still enough to keep the SCR triggered and holding the gate high shouldn't be required.

1 2 next

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