V-Switch for Coilgun timing / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

V-Switch for Coilgun timing

1 2 next

Move Thread

LAN_403

guitardude012

Sun Aug 31 2008, 06:24PM

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

Hi everyone, I have a few questions about using SCRs for a V-Switch:

1. Could you put SCRs in parallel to improve current handling in the v switch, or would that prevent quenching? So using say, 2-3 SCRs for firing and 2-3 SCRs for quenching on a single coil?

2. It seems the to-220 SCRs such as those from littlefuse are better suited to this technique? SK055R is rated to .035ms max commutated turn-off time(t_q) where as Evgenij Vasiljev says we need .07ms for this rating. He mentioned .07ms as very fast for this rating and he seems correct, I can only find .1ms or slower when checking datasheets for stud SCR's. The to-220 SCRs on the other hand are much faster.

I am planning on using a microcontroller to control the gate of the "quench" scr.

How to choose the right capacitance for the quenching capacitor still seems unclear to me.

Thanks,

Josh

youngcoilgunner

Mon Sept 01 2008, 05:32AM

Registered Member #1633 Joined: Tue Aug 12 2008, 04:21AM
Location:
Posts: 45

yeah, you can parralell scr's, just like parralleling transistors. as for capacitance, I dont really know. I dont think there is a established method for this, try using camera flash caps..

guitardude012

Mon Sept 01 2008, 06:48AM

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

Thanks for the reply youngcoilgunner.

I guess I don't understand what D14 and R21 are for:

Also, I would not need D13 if I was using a microcontroller to control the second SCR(Q7) instead of the second inductor(L2) that he uses for sensing?

Edit: I figured out how to select the capacitance for the "quenching" capacitor(more or less). At the time of triggering the second quenching SCR, the main capacitor has discharged to some voltage. The quench capacitance should be chosen using an RLC sim such that it can maintain a voltage above the now partially discharged main capacitors voltage for a period of time slightly greater than the "max turnoff time" (t_q) of the first "firing" SCR.

youngcoilgunner

Tue Sept 02 2008, 04:31AM

Registered Member #1633 Joined: Tue Aug 12 2008, 04:21AM
Location:
Posts: 45

i guess that would do it. if your using something else to trigger the quench, yeah, you wouln't need the diode. though I would reccomend tying your ground from your sensor/timer circuit above the coil, otherwise there might be a problem.

rp181

Wed Sept 03 2008, 01:45AM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

D14 is a capacitor diode for reverse chjarge, in series with a bleeder.

guitardude012

Wed Sept 03 2008, 04:38AM

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

Do you need it though if you already have D12?

rp181

Wed Sept 03 2008, 05:44PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

no, the SCR is open during charging, this part wont be active. You dont have to have it, just make sure caps are hooked up properly/polarity is right, its a good precaution though.

Electroholic

Thu Sept 04 2008, 03:27AM

Registered Member #191 Joined: Fri Feb 17 2006, 02:01AM
Location: Esbjerg Denmark
Posts: 720

rp181 wrote ...

no, the SCR is open during charging, this part wont be active. You dont have to have it, just make sure caps are hooked up properly/polarity is right, its a good precaution though.

WRONG! D14 is a must, and will lead to exploding cap if left out.
Do you not see the voltage curve for the 47uf 800V bank?

In normal coil guns, BEMF is taken care of by a bypass diode across the coil. BEMF Current is not constrained, but merely bypassed so it doesn't over-discharge the capacitor.
In V-switch operation, the goal is to stop current from flowing, so ideally you shouldn't even have R20 and D12.
Now what will happen during switch-off is that the magnetic energy stored in the coil will be fed to the switch-off capacitor bank, backward! Which is not only bad for the capacitors, but it also means you will need SCRs with high enough reverse standoff voltage.
However, with careful balance of D12/R20 and D14/R21, reverse charge can be controlled.

Now what I don't get is why you are using V-switch with uC, with Diagonal half bridge you can actually recycle the magnetic energy instead of burning it with R20 and R21.

guitardude012

Thu Sept 04 2008, 04:18AM

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

If you time things correctly with the quench cap(s) you can actually achieve higher peak current than with the main capacitor(s) and turn it off much faster. I have not seen yet if diag. half bridges are faster and allow the same current peaks?

Also, I can use much larger currents because SCRs can be put in parallel much easier. I am looking at ~2000 amps peak per coil. A positive thermal coefficient is nice, plus I don't need the equalizing resistor between each IGBT.

I do appreciate your feedback electroholic and will try to make a more careful comparison of v-switches to diagonal half bridges.

Thank you for the explanation and warning about D14.

edit: If my bank is charged to half the voltage of his, how can I estimate his -1200V proportional to my values?

edit: is that negative voltage spike caused by L1 or L2? I guess I am still unsure of why D14 is necessary if I don't have L2. Does D12 not protect the Quenching capacitor the same as the capacitor bank in a basic coilgun?

Electroholic

Thu Sept 04 2008, 09:06AM

Registered Member #191 Joined: Fri Feb 17 2006, 02:01AM
Location: Esbjerg Denmark
Posts: 720

guitardude012 wrote ...

edit: If my bank is charged to half the voltage of his, how can I estimate his -1200V proportional to my values?

Depends on a lot of things, the size of your two cap banks plays a huge role, peak current will determine the size of the BEMF, and the values of your diode resistor combos. namely D12/R20, D14/R21.

guitardude012 wrote ...

edit: is that negative voltage spike caused by L1 or L2? I guess I am still unsure of why D14 is necessary if I don't have L2. Does D12 not protect the Quenching capacitor the same as the capacitor bank in a basic coilgun?

Basically, the BEMF here is from discharging both capacitor banks thru L1. And because the main capacitor bank is taking out of the current path, all the energy goes to the smaller quenching bank. And because the bank is a lot smaller, voltage goes way up.
In a normal coil gun you just slab a huge diode across the coil(With no resistance other than the coil's own) and all is good. But as I said, your goal here is to stop the current as fast as possible. So you put R20 there to limit the amount of current going thru D12 and R20. And have D14 and R21 take care of the now negatively charge quench capacitor.

Without D12 and R20, V-switch will have total control over coil current, but then you need high voltage SCRs and quench bank will have to take a lot of abuse.

With just D12 and no series resistance, it will just be a normal coil gun and V-switch will not be able to stop the current.

1 2 next

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