Inductive Energy Recovery for SCR-Capacitor discharge system / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Inductive Energy Recovery for SCR-Capacitor discharge system

Move Thread

LAN_403

Peter02

Sun Jul 05 2009, 03:22PM

Registered Member #1488 Joined: Sat May 17 2008, 10:41AM
Location: Germany
Posts: 18

Hi,
at the moment i am constructing a magnetic pulse device for launching disks of a coil. for me it is only to understand the mechanisms involved in the fields of switching high currents through inductive loads. Since the coil is not damped very much, i will surely get a conciderable amount of reverse charge on my capacitor, that is not a big problem because i use bipolar cap, but it sure would be nice to recover most of that energy.

After searching the net for a solution, i found the following paper showing a very simple method for doing that:

maybe someone else can use this information since i did not find that approach used anywhere in this forum.

Best regards

TheMerovingian

Tue Sept 22 2009, 10:43AM

Registered Member #14 Joined: Thu Feb 02 2006, 01:04PM
Location: Prato/italy
Posts: 383

Don't understand how the circuit work.

In the charging phase the capacitor is charged though the load.
When the SCR closesthe current starts to flow and the magnetic field is building up.
When the capacitor is discharged the scr cannot turn off because the magnet keeps the current but the diode kicks in shorting effectively the capacitor and the current dissipates and damps, no recovery phenomenon and horrible fall times...

what't the purpose ot this?

klugesmith

Tue Sept 22 2009, 05:17PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Makes sense to me. (Sorry this didn't get attention when Peter originally posted it.)

The main magnet gets the desired half cycle of current, as the capacitor discharges from +V to -V.
Second half cycle has current flowing in the "recovery choke", not the main magnet, and restores the capacitor to +V.

More precisely (without circuit simulation):
First quarter cycle:
C discharges from +V to 0, and magnet current increases from 0 to +I.
Second quarter cycle:
C "reverse charges" from 0 to -V and magnet current drops from +I to 0.
During this phase, current begins to flow in the recovery choke, which has 10 times as much inductance as the magnet.
At the end of this phase the thyristor turns off naturally (? though the paper suggests it could have gate-turn-off capability).
Third quarter cycle (about 3x slower than the first two, because 10x larger L)
Recovery choke current builds up to maximum as the capacitor voltage returns to 0.
Fourth quarter cycle (another slow one)
Recovery choke current drops to 0 as capacitor voltage recovers to +V (minus the losses).

[edit] The original application pulses a magnet to switch a particle accelerator beam. I guess it repeats frequently enough to make energy recovery worth the trouble -- for greatly reduced average load on HV power supply.
I challenge someone with spare time to simulate this circuit. Am not familiar with practical SCR models in free simulators. Could make an SCR by tying together standard library PNP and NPN transistors. If necessary, scale the currents down by using smaller V and C and larger L.

TheMerovingian

Tue Sept 22 2009, 11:15PM

Registered Member #14 Joined: Thu Feb 02 2006, 01:04PM
Location: Prato/italy
Posts: 383

This seems not too practical for coilguns, since the recovery inductor must be bigger and have strong core or it will saturate... The capacitor must be non polarized or it will blow because the recovery inductor current ramps op slower due to higher inductance causing much of the current to pass through the capacitor in the second phase. For high frequency switching it may be good, but for coilgun it is worthless

Doing some Pspice simulation with ideal inductors and real semiconductor models

klugesmith

Mon Oct 19 2009, 08:39PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

I think the circuit could be practical for coilguns using non-polarized capacitors, if rate of fire is important. Defibrillator caps roughly match the design in OP. I bet the recovery choke can be made with less wire than main coil, because a) it can have a more efficient core and air gap configuration and b) lower I, so can tolerate higher R without excessive damping.

Found some time to brush up on SwitcherCADIII (Linear Technology's free SPICE) and simulate the circuit. Topology is identical to that in OP. Found an SCR model based on PNP and NPN BJT's. Used actual L and C values; chose R values to get about 85% voltage reversal in main half-cycle and 80% voltage recovery in complete cycle. Set charging voltage to 100 V.

This waveform display shows the 3 branch currents (which add up to zero) and the voltage between the two main nodes. See an earlier post for explanation.

1255984275 2099 FT72496 Ier Ckt Wave Inv

When tools give a picture with dark background, I like to invert the colors. Then a hardcopy uses less ink and is easier to write on.

-Rich

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