Re: Just finished my 3.5kj cap bank & RL test results
GimpyJoe, Thu Mar 23 2006, 02:39AM

I'm not sure about the answers to the other questions, but make sure you insulate the coils really really well from each other. I have had coils arc through the layers and it's not fun to rebuild a coil from scratch when you've just finished one

Re: Just finished my 3.5kj cap bank & RL test results
Sebastian, Thu Mar 23 2006, 11:20AM

made a new coil with twice the length of wire...basically, two spirals on top of each other.

Do I understand you correctly that you want to use the 2 Cap banks for a single coil each an the coils are wound ontop of each other?

If that is the case I see a big problem with the magnetic fields canceling each other...

Re: Just finished my 3.5kj cap bank & RL test results
GimpyJoe, Thu Mar 23 2006, 01:50PM

only if they're wound in the wrong direction. Considering he already tested it I don't think there will be a problem

Re: Just finished my 3.5kj cap bank & RL test results
Sebastian, Thu Mar 23 2006, 04:39PM


True

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Thu Mar 23 2006, 06:58PM

Both spirals are in series...I'm feeding one pulse of 1400v into them.

I did find that they need to be wound in the correct direction. When I first wired it, I did a really low voltage test and the layers pushed apart...I put on a ton more epoxy and re-wired it so the fields all go in the same direction. My dual layer coil is made by wrapping a spiral on a round block of wood, covering it with epoxy, then putting on thin magazine paper and clamping another wood block on top. Then I repeated this process. The coil is a solid shunk of epoxy, but the projectiles are only a paper width away from the actual wire, so magnetic coupling is great.

I know my Qs were a bit confused...because I'm confused! I guess what I'm trying to figure out is how can math be used to explain that it will be twice as powerful now as it was on one bank, given that the capacity is half and the current is the same. To me, this means the pulse will be half as long as it was, so how many times more powerful will the magnetic field be given the dual layer coil vs single layer, twice the voltage and half the capacitance?

Re: Just finished my 3.5kj cap bank & RL test results
Marko, Thu Mar 23 2006, 07:35PM

For electrolytic caps maybe better is to wire in paralell because ESR/ESL usually limits current more than load impedance.

In paralell you will get double max current and again 2x faster discharge time, but you have higher capacitance.

As your coil behaves as load too maybe best is to test both circuits yourself.

Advantage of 1400V is that you can use triggered gap for very high speed switch (there are some small losses but fast switching may actually improve overall efficiency).

Re: Just finished my 3.5kj cap bank & RL test results
Yohan, Thu Mar 23 2006, 08:33PM

Will he have problems with a 3.5kj cap bank.?

<shakes magic 8-ball>

"Yes."

Too much power. You're going to heat the coils to the point of barrel contortion if you charge that much beyond 2~2.5kj. Problems is, you can't get the heat away fast enough without some sort of active cooling. I'm not sure how this would react for your induction setup though.

in a perfect world, would I get twice the launching power?

No. Many, many reasons why. Law of diminishing returns for one.

But, if you've got a camera capturing everything, go ahead and give her a run. I'd be interested to see the end result.

Re: Just finished my 3.5kj cap bank & RL test results
GimpyJoe, Thu Mar 23 2006, 09:35PM


It's a ring launcher, so no problem with the barrel. The problem I forsee is that with the softening of the epoxy from coil heating, the coils may touch and short.

Re: Just finished my 3.5kj cap bank & RL test results
..., Fri Mar 24 2006, 01:26AM

I would say that in a perfect world you wouldn't get exactly 2x the power. So far the only thing I can think of is the lowered coupling because of the double thickness coil and the increased esr of the circuit (paralleled caps/coils doubled it). If you had ideal caps, ideal inductor, etc you would (conservation of energy)

I see no reason why it would not work, with 3.5jk you have about 0.88calories, or the ability to raise the temp of a 100gram coil by about 15F (assuming 5% efficiency). Or it it is really small at 10g it is stil only 150F, which is still within the ratings of most wire/epoxy. I wouldn't worry about it. If you don't want for it to cool down you could epoxy it to a heatsink which should cool it very quickly. I don't see how a barrel is involved with a pancake coil.

As to arking over, I really doubt that you can get 1.5kv to ark though a piece of magnet wire regardless of the pulse shape. I have used magnet wire at 10kv/20khz and got a ton of corona but no arking over.

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Fri Mar 24 2006, 01:56AM

Thanks for all the fantastic feedback everyone. I did fire the coil today at 1260v successfully. I placed a hard disk ring on top, then a soda can, then a brick on the soda can. I fired it and the can became a disc...and to my surprise, I can launch these flattened can discs even though they have no hole in the middle...I thought a hole was required Well as soon as the weather gets better I'm going to launch some stuff outdoors and use my chrony to see what kind of output I'm getting.

After firing 3 times in a row, the coil was nice and warm, but not hot. The system is working very well, and no blown up SCRs either. I have yet to fire it at 1400v yet because my transformer doesn't put out enough voltage to charge beyond 1260v...so does this mean I need a microwave oven transformer?

Re: Just finished my 3.5kj cap bank & RL test results
GimpyJoe, Fri Mar 24 2006, 04:19AM

MOT and a variac should work nicely

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Fri Mar 24 2006, 06:27PM

Alright everyone, some numbers for you to crunch. I myself find the efficiencies hard to believe, but the spreadsheet used the same 1/2MVsquared formula across the board...I'll get some footage of things launching 'cause you've got to see it to believe it. This thing is lethal!

EDIT: Bad Data Removed, see post below

Re: Just finished my 3.5kj cap bank & RL test results
Wilson, Sat Mar 25 2006, 02:12PM

Thats an excellent efficiency. By aluminium 'disc', i presume that you don't have an iron rod protruding from you coil like in classic thompson guns?

Could we get some pics? I will be building one of these soon, from a 3kJ cap bank of computer grade caps and i would love to have a look at someone else's success b4 i start :)

Thanks,

Wilson

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Sat Mar 25 2006, 06:56PM

Nope, no central rod, and no hole in the projectile - just a 3.5" 5/16" 62g aluminum platter I cut out of scrap:

Re: Just finished my 3.5kj cap bank & RL test results
Wilson, Sun Mar 26 2006, 12:19AM

cool stuff :P

Have you tried launching the discs laterally by sandwiching it between two coils?
You'll probably get much more height due to aerodynamics if you can get it to launch in the right direction

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Sun Mar 26 2006, 11:24PM

I was right to be suspicious of those efficiencies of 25%. Like any good scientist I wanted to repeat the test and see if I got the same results, so I set up everything and got ready to chrono the launches. The speeds were much lower than I got the day before which was strange...I couldn't figure out why the chrono would mess up...

Then I discovered that I made an error when setting up for testing the first time and used reflected light instead of a direct source. What happened was the projectile projected a shadow on the reflector for the chrono, and this shadow would trigger the second sensor even though the projectile was only half way to it...so the speeds were just about doubled...

The result is that the sub 10% efficiencies posted earlier are accurate because I used direct light when taking those measurements, and the two 25% ones are not. Today I'm going to do some extensive testing, and also see what difference wiring the banks in parallel for 700v @ 14400uf instead of 1400v @ 3600uf makes. I'll also see about using the dual coil approach Wilson suggests.

[Mod edit: merged double post. To go back and add images to a post, stick them on the attachments board and then link to them when you edit.]

I just ran a bunch of tests and repeated them with acceptable deviation in the results. Here is a spreadsheet showing the average results for varying setups:





1 Layer = Single layer coil
2 Layer S = 2 layers wired in series
2 Layer P = 2 Layers wired in parallel

Re: Just finished my 3.5kj cap bank & RL test results
Wilson, Mon Mar 27 2006, 02:45AM

Interesting...the trend of results seem to hint that an Al disc with 2 parallel coils would yield a higher efficiency.

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Mon Mar 27 2006, 07:15PM

That seems to be the case, and it makes sense too because in parallel the current is doubled and the pulse is therefore shortened. These things launch so fast that a long pulse is wasteful since the projectile is long gone. I'm unable to do a 3 layer parallel for fear of destroying my SCRs with excess current. At some point adding layers would be ineffective because the bottom layers are further from the disc and have less magnetic coupling.

Re: Just finished my 3.5kj cap bank & RL test results
Wilson, Wed Mar 29 2006, 01:34AM

what SCRs are you using atm? puk or stud?
I've bought 4 studs for use in my launcher, put i'm not sure if their 3kA surge rating will be sufficient for a 2-3kJ bank.

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Wed Mar 29 2006, 02:08AM

Mine are the 57/20 SKKH in the attached datasheet



I am using 3 in parallel...I have 3 more and will add them for 6 total in parallel for safety since I'm barely making it with 3.

Re: Just finished my 3.5kj cap bank & RL test results
Quantum Singularity, Wed Mar 29 2006, 03:54AM

Congrats... 357ft/sec wow! I got to get me a chrono.

A few things a was thinking about... its definately about quick pulses as my 120uF cap charged at a few kJ is probably very similar to yours at the same power level (although I am sure different size coils and different triggering will definately make them not perfectly equal for a given power level). And I beleive my 1 cap has about .1 ohm ESR wich is probably higher than your cap bank. Its true the esr will half when paralleling two caps, but that doesnt mean the current will double. Current is V/R. There is other R in the circuit than just the ESR so the circuit resistence will not quite half, and the current will then not quite double. I want to try different series/parallel configs to figure out wich is better. Remeber, the induction disk launcher wants maximum di/dt. So the common railgun consesus that a longer less current pulse is better than a shorter high current pulse doesnt hold true here. Heck even my 4uF caps would launch a light weight copper ring pretty high with the first coil I ever made (totaly non-optomized). So I am a bit hesitent on how my new 9100uF caps will work but it seems FastMHz has proved that multi-thousand uF caps will do quiet well with this design. Keep up the good work!

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Wed Mar 29 2006, 07:14PM

It would be neat to find out the ESR of my bank...how would I measure that? Mine is actually wired 4S4P so individual cap loads shouldn't be too bad.

Your 120uf charged to about 7650 volts will be about 3.5kj.

Your caps will work well...from my experimentation and testing, I would recommend wiring your caps 6S2P so you get 1500v @ 3033uf...having that extra cap is hard to fit in...or even 4S3P for 1000v @ 6825uf since you're using thicker wire.

Re: Just finished my 3.5kj cap bank & RL test results
Quantum Singularity, Thu Mar 30 2006, 01:24AM

Did you get a good deal on those scr's? The few similar models I found for sale are $50+ per each. I dont think I could afford 6 of them! I got another scr question I'll post on the general electronics forum.

I only charged my big cap above 5kV once. That is the really mangled disk thats on my EMP Ringlauncher webpage. It also kinda messed up the weak launching coil I was using at the time. The disk in my vatar the ended up in the roof might have been that high or even higher though, I didnt have a meter at the time so there wasnt any way to tell. But judging by how dish-shaped it warped and comparing to my other disks it was definately around the 7kV mark or more.

I am not sure about the ESR of lytics... I am guessing that would have to be on a spec sheet somewhere from manufacturer, if it was ever listed. I was told by a few people mine was .1 ohm on my Maxwell, and that would be correct with the facotry specified 100kA max discharge (10kV/.1ohm=100kA).

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Thu Mar 30 2006, 02:25AM

I bought them off a fellow forum member a while back...your best bet is surplus, which is what I think mine were anyway.

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Sun Apr 02 2006, 05:32PM

Interesting question: why doesn't steel launch? I placed a steel disc on my coil and fired it and it didn't budge! Would this happen with any ferrous material?

Re: Just finished my 3.5kj cap bank & RL test results
Electroholic, Sun Apr 02 2006, 06:17PM

You use Alu disk because you want low resistance, light weight, so the repelling force generated from eddy current can accelerate it.
Now, if the disk is ferromagnetic, the coil will suck it down when fired.
Notice we put the slug behind the coil when firing reluctance type coil guns?

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Sun Apr 02 2006, 07:26PM

Makes perfect sense...that's so obvious now I feel stupid

Re: Just finished my 3.5kj cap bank & RL test results
FastMHz, Fri Apr 07 2006, 07:07PM

Got my new diodes from Mouser, rated for 25a @ 1600v, far higher than I'll need while charging my bank. So I did a test to see how it was all working and put, for the first time, the 1/4" thick aluminum disc on my latest and greatest dual layer coil, put a soda can on that, and then a brick on the soda can. I fired it at only 1000v since I was indoors, which in my bank is 1.8kj. The brick flew 6 feet into the air, and the can was 1/4 it's original height. It was so loud it scared me - I thought something blew up...I'll get my brother to take some footage soon so everyone here can see what computer grade electrolytics can really do