Ring launcher problems/questions / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Ring launcher problems/questions

1 2 next

Move Thread

LAN_403

benbmw

Fri Dec 04 2015, 11:49PM

Registered Member #55052 Joined: Thu May 14 2015, 06:47PM
Location:
Posts: 12

I recently finished construction of a ring launcher and I am having some issues. I think most if the problems come from the capacitors I am using but I wanted your opinion so ill describe my setup. I am using 8 400v 3600uf electrolytic caps in a 2s4p configuration. I sourced the caps from a junk yard and im not sure how old they are but the most likely went through numerous 110 degree summers. I am using a mot controlled by a variac to charge them through a 750ohm resistor. I am switching using a rat trap to slam wedges of copper together which discharges into a 4 turn coil of 2/0 copper cable. I have also made a 10 turn coil of 8 gauge wire. My projectile is a 1/4" thick by 4.5" diameter aluminum disk. To protect my cap bank I had a Chinese 1600v 100a stud diode in anti-parallel to hopefully conduct and dissipate negative voltage spikes.

So after the first shot at only 100v the aluminum disk barely jumped no more than an inch. I tried a 300v shot and just before firing I heard some metallic "pinging" or wrinkling sounds coming from my capacitors, but I fired anyways with no better results than last time. I checked my bank because it wouldn't hold any voltage after that. I checked each side of the series config and one was at -5v while the other was at +5v. I also had removed and checked my fly back diode to find that it was short circuit in both orientations. There doesnt appear to be any physical damage to the caps or diode. How could the diode have blown with that low of voltage? Any suggestions? Would a diode in series with the work coil to completely block reverse voltage be better?

Thanks for any help,
Ben

klugesmith

Sat Dec 05 2015, 01:10AM

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

Welcome, Ben.

benbmw wrote ...
Would a diode in series with the work coil to completely block reverse voltage be better?

No. The diode would conduct until the current reaches zero. Then your capacitor bank would be sitting with the maximum possible reverse bias (assuming an underdamped RLC network).

As for destruction of a diode, how much current did you send through it, and for how long?
How did you figure those values?
Barry has an online calculator for RLC pulse discharges here:

and an online calculator for inductance of air-core coils here:

You can easily figure the resistance of your coil and interconnections, and explore some wild-ass guess values for capacitor ESR. More tricky is the magnetically coupled projectile, which will reduce the effective inductance and increase the peak discharge current.

Did you make any effort to reform your long-idle electrolytic capacitors before suddenly giving them hundreds of volts?

DerAlbi

Sat Dec 05 2015, 01:27AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Hoooly F. You are working with quite a lot of energy while you are missing fundamental principals of electronics. I suggest you massively scale down your setup to learn whats going on without damaging expensive-to-replace parts. That you startet at 100V was actually very good... but you can learn and experience your circuit flaws also with 2x 100uF @ 12V in series, smaller diodes and a smaller coil.

First problem: Series capacitors
Series capacitors will differ in capacitance and ESR and leakage. Therefore they will be charged differently and whats even worse: they discharge differently. That explains your -5V on one side of the bank after you test. This is actually a damaging condition for an electrolytic capacitor and you should only apply voltage slowly and over a high resistance. This gives the electrolyte a chance to recover the dielectricum.
Solution: put an antiparalel diode across EVERY capacitor. You need to avoid the chance that a cap can discharge below 0V.

Second: 100A-Diode.
When the Capacitors are maximal discharged, the current is at its peak (most likely not - because you are most likely ESR-bound, but accept that for now). This peak current will then flow through the diode which is verry likely to be way undersized with 100A. (datasheet?)

Third: A diode in series will not block reverse voltage.
Pease understand that a negative voltage can not be blocked by a series diode. A diode will be conductive as long as the current flows in the right direction.
Think about it this way: imagine a LC-Oscillating circuit. Now put a Diode in. The cap is initially charged then its discharging. The diode conducts, the indurctor builds up its magnetic field. Will the diode provent the capacitors to be so much discharged until its negatively charged? No! because all the time the current flow was in one direction.

Fourth: A diodes death...
is caused by over current and over heating. That there is low voltage after the Diode is shorted is only an implication of the shorted diode. I mean: how can any voltage be across a shorted component? The -5V and +5V add up to 0V. If you mean with "low voltage" the 100 or 300V, then still: it does not matter.
The "low voltage" will still be enough to cause a significant current, and the diode simply couldnt handle it.

What is actually wrong in your design and why your disc does not move as powerfull as you expected i cant say. Maybe others know more of what is going on. I am not into ring or disc launchers. It could simply be that your voltage is too low and you current does not rise fast enough.
But i dont want to encourage you to crank the voltage up... your current circuit and understanding likely cant handle that safely.

Maybe you could tell us more about what possibilies you have to measure whats going on?
Do you have an oscilloscope?
Are you able to work with LT-Spice?
Maybe a photo of your setup?
How old are you actually?
What education did you receive if you are older than 18?

Edit: damn. Klugesmith. you were fast. and you have links to usefull sites. grrr. you win! But we adress similar points.

benbmw

Sat Dec 05 2015, 01:34AM

Registered Member #55052 Joined: Thu May 14 2015, 06:47PM
Location:
Posts: 12

klugesmith wrote ...

Welcome, Ben.

benbmw wrote ...
Would a diode in series with the work coil to completely block reverse voltage be better?

and an online calculator for inductance of air-core coils here:

Thanks for the info. I think my mistake was trying to reform them in the 2s4p configuration. I just finished reforming 4 more capacitors all in parallel and tried them(keeping them in the parallel only configuration) and I got better results. It will slap a tinfoil disk against the ceiling with some force. the disk is made of tin foil folded for 16 layers and weighs about 10 grams. unfortunately the 150 gram 1/4" aluminum disk will only launch about a meter. this is much less than 1 percent efficiency. I am using the 8 gauge coil for these tests and it is sandwiched between some 3/16 plywood and held together with rubber bands temporarily. The disk starts about 1/4-3/8 away from the actual coil. Im not sure if this would affect it.

Another thing I noticed when reforming these capacitors is that when I first reformed them to 400 volts my variac was at about 17 or 18 on the dial. Now, to get it to 400 volts I need to go to 20 or 21 on the dial.

I have not made the calculation to estimate the current pulse yet.

EDIT

Thank you for all the information DerAlbi. I will upload some pictures of my setup in a bit. I think I was very unwise about how I used(and reformed) the old capacitors in series. I dont have an oscilloscope but I charged them all up individually and then measured how much they would self discharge over 30 seconds with a multimeter. This ranged from 6.5 volts to 30 volts leaked over 30 seconds. Im sure the load and charge was very unequally shared. I just turned 19 and I am currently at a community college but I am planning to transfer to a university and major in electrical engineering. I have some knowledge of electricity and circuits but obviously not that much. Enough to make a spark gap Tesla coil two summers ago. Also i have never worked with LTspice

here is the link to the album showing my setup:

DerAlbi

Sat Dec 05 2015, 02:20AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Hmmmh. Well i would like to encourage you to try yourself in LT-Spice. Its a skill you definitively must have if you want to go in this direction. (i am a university teacher in analog circuit design) LTSpice is really easy to use and basically every question you could have is easily answered by asking google that exact question

Its cheaper to experiment in LTSpice and actually safer. It just takes a little time to learn. And with will payback 10-fold.

The different leakage on your caps could vanish if you reform the caps over night. It will still not help with the different capacity... discharging series capacitors to 0V is a flawed concept in that manner.

The variation in your dial can simply be explained by voltage differences in your local power grid.

And now try to estimate your current pulse! Current is what propells your disc.. so you should have actually started there right from the beginning.
I mean.. how else you could know if your diode will be ok? Or your switch.. or your wire or... whatever you have in the circuit

Every situation you encounter in electronics will be described by numbers and can be checked against specifications. Thats the proper way to do it.
If the numbers are hard to calculate (it becomes easier over time), you use a simulation program like LTSpice. Simulation is however a science on its own and it also provided many sources of error. But better an error there than.. you know..

benbmw

Sat Dec 05 2015, 02:37AM

Registered Member #55052 Joined: Thu May 14 2015, 06:47PM
Location:
Posts: 12

Thanks again Albi. I will definitely check out LT-spice and learn to use it. I will also try reforming all my capacitors overnight. I assume the way to do this is leave them charging at their rated voltage overnight. Is this correct? Or should I bring them up slowly in increments and then, once I reach the rated voltage, leave them overnight?

benbmw

Sat Dec 05 2015, 07:53AM

Registered Member #55052 Joined: Thu May 14 2015, 06:47PM
Location:
Posts: 12

I started reforming 8 capacitors in parallel and I brought it up to 400 volts with the variac and let it sit for a few hours. After that time it was down to about 370v with no change to the variac. I measured the voltage across the 750 ohm charging resistor and it was about 20v. Also when the voltage on the variac was turned to zero, or when the charger was completely disconnected, the voltage in the capacitors dropped at a rate of about 1 volt per second. It seems that the leakage is getting worse. could this be caused by one faulty capacitor that might be beyond repair? Im not sure if I should leave them to reform overnight.

DerAlbi

Sat Dec 05 2015, 11:41AM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Hmmmmmmmmh. Dont leave them over night.
First of all reforming means a SLOW increase of voltage. So basicall you start at 50V, wait until the leakage is within specifications (Datasheet), then you dial the voltage up to 100V, wait again.. and so on.
Applying 400V directly is not reforming, its "stressing the hell out of the suckers".

An electrolytic cap is basically a aluminium foil covered in aluminium oxide. This oxide is constantly damaged by stress and reformed by a chemical reaction thats happening if there is current flow from the electrolyte to the aluminium. The holes in the dielectric oxide layer close by them selfes usually. However they can be beyond repair... why?
Imagine: a tiiiiiny little hole. (plase stay focused on dielectric layers!

) Though this hole there is 400V dropping with about 26mA. That means Power disipation! And its a LOT. its around 10W. This "W" does not only mean Watt, its means WOOOW in this case!
Tiny structures wont handle 10W of power. What happens is that the capacitor heats up locally around the dielectric failure and the electrolyte is destroyed. So the electolyte has no chance to repair the failure.

I suggest you test their leakage one by one and sort out the really bad one(s). If you end up with half the capacitors - no problem. Now your project has just become more interesting, because you need to double the efficiency to get the same results

Next time you reform some caps, use a way bigger resistor. 10K+

klugesmith

Sat Dec 05 2015, 05:43PM

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

Ben, you are right about not wanting to test or reform old capacitors in parallel. You can't tell which way the leakage current is going.

You can reform them all at the same time, using a separate resistor for each capacitor. As Der Albi said, 10 kÎ© is at the low end of recommended resistance. I think it would be fun to put a neon lamp in series with each resistor, for visual indication of capacitor currents & reminder to be conservative with the variac knob. Be sure to post a picture for us.

benbmw

Sat Dec 05 2015, 06:50PM

Registered Member #55052 Joined: Thu May 14 2015, 06:47PM
Location:
Posts: 12

Thank you for the suggestions klugesmith and DerAlbi. I will order some 20w 15kohm resistors and wait for those before continuing to reform. I dont have any high wattage resistors besides 4 200w 750ohm resistors from the original cap banks that these capacitors were in. The visual indicator is a good idea. The majority of small neon bulbs i see online are rated for less than a millliamp at 120v. Its seems that I would have to find a lamp that will handle up to 20 ma but I would still want it to be visible when the current drops as it charges.

1 2 next

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