Paralleling SCR question for coilgun / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Paralleling SCR question for coilgun

Move Thread

LAN_403

Quantum Singularity

Thu Mar 30 2006, 01:33AM

Registered Member #158 Joined: Sun Feb 12 2006, 09:53PM
Location: Central Ohio
Posts: 282

I have a question about arranging my scr's. I'll probably have a bank of them, and a bank of caps. Some have said for a bank of scr's to work properly you need to have a small amount of resistance in series with each one to ensure equal current division. Makes sense. But requires not so common high power low ohm resistors, extra space, and some extra circuit losses. I was wondering if this idea I have will work to avoid some of these.

Assume I have 4 caps and 4 scrs. Instead of paralleling all 4 caps and all 4 scrs, can I attach a scr individually to the output of each cap... and then tie the cathods together of all the scrs to form 'the bank'. That way each cap has its own scr and you dont have to worry about current dividing unequally as each scr only has 1 cap discharging through it?

rupidust

Thu Mar 30 2006, 04:23AM

Banned
Registered Member #110 Joined: Fri Feb 10 2006, 12:23AM
Location: Banned City
Posts: 85

One, have you tried just connecting the SCRs in parallel and testing it? High power SCRs are very robust. You start at low energy levels and build it up till it works or fails. If it were to fail, you take note at what condition it happened because you may in the furture want to do the same paralleling at lower energy levels; hence you will have the answer to this same question before the question even came to mind. The current may be the same for the single cap branches as that of a combined bank and one SCR. The Farads of the caps just dictates the length of time the pulse last, for the voltage determines the current.

But on the other hand, the greater the Farads the closer the Peak current will approach the DC current. An LC simulator will show probable differences in current.
http://mgc314.home.comcast.net/ckt_sim/ckt_sim_main.htm

In order to have an SCR to each cap, the SCR will be connect as a high side switch, meaning the common cathods are all ontop of the coil and the coil is connected to ground. Is this what you mean?

Quantum Singularity

Thu Mar 30 2006, 05:45PM

Registered Member #158 Joined: Sun Feb 12 2006, 09:53PM
Location: Central Ohio
Posts: 282

I am going to experiemtn with just one and start with lower voltages and work up just like you said. But I cannot find the specs on my scr's and I am pretty sure I will need multiple on my cpa bank. The caps will be 250V 9100uF, not sure how many I will use yet for my projects.

That schematic is essentially the same as what I drew on paper. The only difference I did (and I am not sure if it would function different or not) is that I tied all the cathodes to ground and put the load coil between the cap - and ground. Guess it would probably work either way the same - just might float the ground on the trigger or charger up depending on the config?

To keep the + cap sides seperate I would probably have to use a diode or something on a sperate line to each cap from the charging supply.

I dont know if this is worth the extra fuss than just tying a bank together but if a few extra leads and maybe diodes will help save an expensive scr or two than its worth it.

HV Enthusiast

Thu Mar 30 2006, 06:13PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Paralleling SCRs is not a very good practice. The problem is that the Vfwd on each one will be different and current will not be shared between them. Unless you come up with either a passive or active current sharing scheme, i would not recommend paralleling SCRs.

FastMHz

Thu Mar 30 2006, 06:34PM

Registered Member #179 Joined: Thu Feb 16 2006, 02:08AM
Location: Hagerstown, Maryland - Close to Prime Outlets
Posts: 287

I now have 5 in parallel. I'm 99.9% sure if I used just one it would die a horrible death on my launcher. I have a 10ohm resistor on each gate.

Marko

Thu Mar 30 2006, 07:08PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Paralleling SCRs is not a very good practice. The problem is that the Vfwd on each one will be different and current will not be shared between them. Unless you come up with either a passive or active current sharing scheme, i would not recommend paralleling SCRs.

In schematic SCRs are connected each with its own capacitor so I dont think current sharing would be a problem. If you tought just to paralell all and connect to whole bank then such problems would occur.

HV Enthusiast

Thu Mar 30 2006, 07:26PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

For the application, you really want to parallel the capacitors, and not discharge each one separately as shown in the schematic. You'll find your current pulse will get quite messed up if you do it this way and not have impedances matched on the connections, timing synched, etc...

If you are looking to maximize the peak pulse current, you want to parallel the capacitor bank.

rupidust

Thu Mar 30 2006, 09:41PM

Banned
Registered Member #110 Joined: Fri Feb 10 2006, 12:23AM
Location: Banned City
Posts: 85

The gates are tied to one trigger. Try first, then if fail, ask second. If you don't ty, failure is guaranteed.

Quantum Singularity

Fri Mar 31 2006, 02:51AM

Registered Member #158 Joined: Sun Feb 12 2006, 09:53PM
Location: Central Ohio
Posts: 282

FastMHZ - did you use any balancing resistors tied in series with the scr or just the ones on the gate?

EVR - I understand what you are saying. But do you have a better solution? The timing will be all tied together from the same trigger so that shouldnt be a problem. And with a fairly careful design using equal distances, buss bars, and solid connections, I am sure that I will be close enough with "the impedances matched on the connections". The caps (which you say to parallel) are probably going to be the biggest variable since most are like -20% to +50% and the ESR and ESL probably is off similarly between caps.

Has anyone else paralleled scr's succesfully and did you use some kind of R in series with the cathode or anode?

I have another question about how to hook one of these guys up since I dont have a specsheet I am unsure, I'll start a new thread for that.

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