Capacitor bank help / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Capacitor bank help

Move Thread

LAN_403

onidiablo

Sat Apr 03 2010, 11:00AM

Registered Member #2779 Joined: Sat Apr 03 2010, 10:51AM
Location:
Posts: 7

Hi all ive been watching this site for ages and have finaly joined up! but anyways ill go straight to the point. when i make my capacitor bank can i put the capacitors on a prototype board or is there any downside to that? if you know an better altenitive to this please share im open to ideas. thanks in advance. btw im using at least 10 photoflash capacitors also would using 12 add alot of diffrence rather than 10?

Bonehead

Sat Apr 03 2010, 12:02PM

Registered Member #2046 Joined: Sun Mar 22 2009, 01:12PM
Location: Stockholm, Sweden
Posts: 23

Most photoflash caps I've looked at are around 330V 80ÂµF. So that'd make a difference of around 160ÂµF.

800ÂµF/960ÂµF. I doubt it'll make a huge difference but it depends on projectile weight and the coil.

For such a lower power bank I personally don't think a prototype board will reduce the efficiency very much but you could just aswell solder them all to a thicker wire.

Like this:

But obviously photocaps are smaller and you wouldn't need as thick wire.
Sidenote: Caps in photo are 4x2(parallelxseries), you probably want your entire capbank in parallel.

Barry

Sat Apr 03 2010, 02:17PM

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

I'm doubtful that a prototype board has sufficient contact area for good low-resistance connections. They are not designed for high-current applications; there is a high probability you'll fry the proto board itself.

Note that coilgun performance is sensitive to total loop resistance in the discharge path. You want resistance to be as low as possible. Even a few milliohms will reduce the peak current. This means using short wires and soldering all connections if possible. Even screw-on connections absolutely must be very tight with lots of contact area and clean wire.

Good luck, Barry
"The majority of quotes in emails are misattributed to famous people." ~ Abraham Lincoln

Sulaiman

Sat Apr 03 2010, 04:50PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I used plain matrix board for my cap bank, 12 rows of 8 photoflash capacitors.
The wiring side shows the tinned copper wire interconnects.

I used two 160V zeners in series per row to limit/balance the charge per capacitor and connected to diagonally opposite ends of the bank to equalise stress over all the capacitors during discharge.
I've had this bank for many years with no problems (other than exploding a cheap dmm)

onidiablo

Sat Apr 03 2010, 11:41PM

Registered Member #2779 Joined: Sat Apr 03 2010, 10:51AM
Location:
Posts: 7

wow nice bank what are the specs?

onidiablo

Sun Apr 04 2010, 12:16PM

Registered Member #2779 Joined: Sat Apr 03 2010, 10:51AM
Location:
Posts: 7

ok so ive thought instead of puting all th capacitors in parrallel i would put it in in 4 rows of 3 capacitors in series to get more joules is esr a big factor in this set up? they are standard photoflash capacitors.

Saz43

Sun Apr 04 2010, 09:33PM

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

You won't get more energy by putting capacitors in series. When you series caps, you get a higher maximum voltage and less capacitance, which balances out and leaves the energy the same.

ESR does make a difference- that's why it's a good idea to avoid putting electrolytic capacitors in series if possible. However, since photoflash caps will be limited to low-to-medium amounts of energy, it might not make a noticeable difference in your case.

Myke

Sun Apr 04 2010, 11:19PM

Registered Member #540 Joined: Mon Feb 19 2007, 07:49PM
Location: MIT
Posts: 969

You do get a higher energy when putting capacitors in series. The energy stored is (c*v^2)/2 so there is the same energy stored in the caps when you put them in series or in parallel. .5*(.5F)*(2V)^2=.5*(2F)*(1V).
In series, the cap's ESR add and when they are in parallel, they just looks like resistors in series with the cap (no simplification if I am correct).
I think it's fine to put electrolytic caps in series but you need to also put balancing resistors so that they charge evenly. Some people have used series electrolytic caps for filtering high voltages and they haven't had any problems.
Like Barry said, the resistance is important but with the increased voltage from caps in series, you can get a higher peak current. As an example: 100V, 50mOhms, 1mF, 500uH results in a 134A peak decreasing to 0A in 2.22mS(not critically damped) and 200V, 100mOhms, 500uF, 500uH results in a 185A peak decreasing to 0A in 1.56mS (not critically damped).

klugesmith

Mon Apr 05 2010, 12:42AM

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

In other words, series gives you the same energy as parallel, for the same number of caps.

Banks of 12 caps in parallel, or 12 in series, or 3S4P combination, will each store 12 times the energy of a single cap. The resulting discharge dynamics will be very different if the load is unchanged, a qualification often omitted in threads on this topic. By using coils wound in the same volume with different wire gauges, you can get the same peak ampere-turns, same timing, and same damping from all 3 capacitor topologies. Then all 3 topologies will have the same current waveform and same ESR losses in each individual capacitor. (And same coilgun performance.)

Myke wrote ...
... Like Barry said, the resistance is important but with the increased voltage from caps in series, you can get a higher peak current. As an example: 100V, 50mOhms, 1mF, 500uH results in a 134A peak decreasing to 0A in 2.22mS(not critically damped) and 200V, 100mOhms, 500uF, 500uH results in a 185A peak decreasing to 0A in 1.56mS (not critically damped).

Myke's simulation examples are 1 cap and 2 caps in series.
If we put the same 2 caps in parallel: 100 V, 25 mOhms, 2 mF, 500 uH, then Barry's awesome simulator gives 192 A and first half-cycle ending at 3.12 ms. So in this underdamped case (resistance << sqrt(L/C)) , the parallel and series banks yield essentially the same peak current and magnetic field (storing about 90% of the joules that were originally in the capacitors).

The pulse lasts twice as long in the 2P case, which is not necessarily good in a coilgun.
With 2 caps in parallel and a coil with half the turns count: 100 V, 25 mOhms, 2 mF, 125 uH, the simulator gives 370 A and 1.56 ms. Exactly the same ampere-turns and timing as the 2S configuration with original coil.

Note that except for the last example, we have been changing the circuit R value as if it were entirely due to ESR of the capacitors, with no contribution from the coil.

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