Best use of 20*15000uF 400v Capacitors? / General Science and Electronics / Forums

Forums

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

« Previous topic | Next topic »

Best use of 20*15000uF 400v Capacitors?

previous 1 2

Move Thread

LAN_403

Marko

Sat Dec 16 2006, 06:57PM

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

That's a railgun - simply a perfect railgun bank.
I think it would be a waste to use them for anything else ;)

Self Defenestrate

Sun Dec 17 2006, 09:57PM

Registered Member #87 Joined: Thu Feb 09 2006, 01:36PM
Location: San Jose
Posts: 191

I'll second(or third..w/e) the railgun idea. With the size of the bank you could build, you could build a really powerful gun.

LarsE

Mon Dec 18 2006, 01:22AM

Registered Member #153 Joined: Sun Feb 12 2006, 03:44PM
Location: Sweden
Posts: 31

I will go with the railgun, now i have 24 of the caps so i think i make the bank serial-parallel with 6strings of 4 caps for 22500ÂµF 1,6KV, that will be 28,8kJ at 1,6kV.

I have seen some railguns that are held together by Steel plates, like http://kisaragi.soregashi.com/top.html (Japanese) and http://www.matthewmassey.com/SimpleHVRailgun.htm, and i wounder how the steel is affecting the performance of the railgun.
Any difference if you are using plasma armature or not?

Jasonr

Thu Jan 03 2008, 04:41AM

Registered Member #167 Joined: Mon Feb 13 2006, 06:41PM
Location: Waterloo, WI
Posts: 54

Hi,
Nice caps you got there. I would make a railgun for sure. (I mean what else is there) Sorry I have been AFK
the only problem with them is they will have a high ESR. Actually very high because of the capacitance. The biggest problem with electrolytic's because the ESR greatly slows your discharge time limiting your peak current. By putting more in parallel you reduce this, I can see why 1.6kv is appealing as it covers up inefficiencies in wiring, switching and projectile transitions.

I am transferring my domain name this week and my web site is down. However as someone mentioned my discharge time is about 1.5ms and I have have 2 banks of 50 caps 3300uf 430VDC for a total of 100 caps 860 Volts 30,508J which is about the same energy size bank. I am in the process of ripping apart my cap banks and adding more caps to get me and 87kJ bank.

How would you plan to trigger the railgun? make a spark shooter projectile trigger. PS I designed a rail for 1200volts about you current if you want me to send you the sketches.

Jason Rollette

badastronaut

Thu Jan 03 2008, 05:30AM

Registered Member #222 Joined: Mon Feb 20 2006, 05:49PM
Location:
Posts: 96

If you make a railgun with those caps, put them ALL in parallel. There is no question about it. Any configuration other than all parallel will result in a reduction of efficiency, and you'll end up with a crappy spark shooter. Somewhere on this forum I posted an arguement proving why parallel always wins.

The short story is that adding caps in series will increase your total average circuit resistance weighted by current. Adding in parallel will reduce that number. There is no compromise; lower resistance is always better, ALWAYS!!!

Higher capacitance capacitors generally have lower total average resistance than lower capacitance ones. A badly abused 5800uf 450V cap has about 45mohm resistance while one of my 3900uf 400V cap has a datasheet value of 39mohm, which has been verified by test. A 120uf 330V cap might have about 1 ohm total average resistance (weighted by current). The relationship is non-linear so you cannot extrapolate those values for your caps.

Sorry to say, but Jason is misguiding you. Yes, ESR will lower your peak current, but so will inductance. Both will also slow discharge time, but discharge time and peak current are irrelevant for such small energies.

High ESR is not a concern for railgun discharge times. Discharge times are not much of a concern for amateur railguns because most amateurs build the rail length way too long such that the current pulse is over well before the projectile exits the barrel.

You won't have to worry about discharge time unless your railgun can fire projectiles at least 1km/sec or more AND your rails are about 1 meter in length or less. That is also assuming your circuit inductance is on the order of .5uH.

The reason you want to have the lowest possible total circuit resistance is because that resistance forms a voltage divider with the equivalent resistance associated with converting electrical energy to mechanical energy. Your caps in parallel might be able to acheive 1-3mohm total circuit resistance including rail, cap, busbar, and switching losses. In comparison, the resistance associated with projectile acceleration is 10uohm or less for railguns of this size. This is why efficiency is so poor for most designs.

The only time when you might add caps in a series/parallel configuration is when your rails are so short and projectile speeds so fast that the cap bank does not discharge very much during a shot. I don't think there has ever been an amateur railgun that has had this problem because it is a sign of very high performance (for rails 12 inches or more in length). Even in that condition, the gun will still likely have better efficiency with all parallel caps (though not necessarily the highest output speed).

If you face the above condition, you may end up with a huge muzzle flash as all the projectile will exit near the peak of the current, and all the energy stored in the inductance has to be dissipated.

You also need to reduce total circuit inductance to a minimum. A value of .5uH is good, but less is generally better assuming your switch can handle the didt. It is somewhat difficult to get much less than .5uH. It is important to reduce inductance because that allows you to push the circuit resistance lower before the current starts to oscillate, which is bad for electrolytics. Even if it does oscillate, it will oscillate less with less inductance.

Also, use very low mass projectiles. One gram is about as big of a projectile you'd want to have for high efficiency. I'd say 5 grams would be the limit, but efficiency will be considerably lower for higher masses. If you use aluminum, make sure the cross sectional area of the projectile (normal to the current density) is not so small that joule heating melts the projectile. You can calculate this.

previous 1 2

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