Capacitor options

[Ewhaz]

My capacitor overheating has me worried. I'm looking into other options as it seems that aluminum electrolytic are hard to find in any high voltage applications above 500v. It seems the only capacitors I can find over that range are only available for direct purchase from suppliers of specialty capacitors. This is frustrating. I also have to keep my costs down so that I have room to experiment with.

Can I go with ANY type of electrolytic capacitor, such as paper in oil or does this have to be aluminum foil? Where did you all find your capacitors at?

[Aaron]

cap heating

The capacitor heating seems a little odd. I've never had any capacitors warm no matter how big or small my setup has been to charge caps. It may even be dangerous.

How many uf's again at 500v? You can find ones from old microwaves in that range at .1-1uf range.

[Tecstatic]

Electrolytic capacitors overheat, when the ripple current exceeds the specifications.

If your application has short high current pulses, then from my experience with switch mode power supplies, the capacitance is not the most significant figure, it is the volume of the capacitors. Preferably several smaller caps than one large.

I once read, that for this purpose ( short high current pulses and high frequencies ) two capacitors of the same volume has aprox. the same ripple current capacity, although their capacitance differs.

You can offload your present capacitors by paralleling some capacitors with good pulse properties.

Another option is polypropylene capacitors (MFP), you can get them up to approx. 1uF/ 3kV. They have good pulse properties. Ceramic capacitors can be even better.

Use ebay to buy capacitors, if your budget is low. You can set up a search for your specific needs and get emails with the matches. I have got capacitors for nearly nothing this way. www . ebay.de is a great site for electronic components.

Ebay tips. I can see two successful ways to buy cheaply:

1. some sellers auctions gets no bids, you can offer the minimum bid and hope to be the only bidder, it happens.

2. If you really want an item, then identify your maximum bid, and make the bid 3-7 seconds before the auction ends. Surprisingly often I have got the item for a fraction of my bid. Here it is a must, that your computer uses a clock server for time synchronisation.

4 gigantic High Voltage Elko, each 31,5 µF / 1600V Siemens (high quality) were sold for a total 8,39 euros in December. But those are not seen so often. Image attached. I have even seen another Siemens cap 1,8kV weight 7 Kg (as I remember).

4 Stk HV Elko 31,5 µF 1600V Siemens bei eBay.de: Bauteile (endet 12.12.07 19:26:08 MEZ)

There are also many offers for ceramic Tesla capacitors (for Tesla coils).

The attached data sheet are for capacitors from China sold on ebay.de at reasonable prices.

Good luck :-)

[theremart]

Another note on Capactiors

Do not hook them up backwards

amazing smoke came out of mine when I did...

I agree getting hot is a bad thing.... I was told to treat capacitors like batteries, don't put a AA battery in for the job of a Car Battery.. But I can also see Aarons point of not having good quality capacitors.

I would like to be able to store energy in a better place than the battery or a capacitor... I even considered compressed air, or heat...

Lead acid batteries do loose energy. Seems like a shame to loose 10% of your hard earned energy overnite.

mart

[Ewhaz]

Well heres the thing, it's a 3.3 uf capacitor rated at 450V but it still overheats. However it did seem small when I got it, I was expecting a big capacitor but it was the small aluminum canister type, about the diameter of a AAA battery and about 1/2 inch long.

I also notice that if I touch the capacitor, that is the shell, and touch other parts of the circuit I'll get a tingle going through me. It's still charging, however before I found the faulty 555 it was simply wide open and dumping strait into the battery. Only after I fixed that and it started pulsing properly did it start to get hot again.

I'd thought about putting some in parallel or series to help divide the load between them, I was just researching that today.

Edit:
Well, just for kicks I paralleled another cap in there, instantly they are both running much cooler but still getting warm.

[ren]

Id like to make a comment directed towards your input on capacitor selection Tecstatic. I noticed recently on EFVT 2 something that I didnt pick up before. John has two capacitors on some of his devices. I noticed the stubby one and the large one on the motor generator, but I recently saw another on the small schoolgirl with the 4 timing wheels (think?)

I was wondering if they might be there to accomodate a storage of high potential AND high current. Sort of like one dumps at 14v 1/2 amp and the other at 200v 1/10 amp. You get the idea.

[Peter Lindemann]

Caps Not Happy

Well heres the thing, it's a 3.3 uf capacitor rated at 450V but it still overheats. However it did seem small when I got it, I was expecting a big capacitor but it was the small aluminum canister type, about the diameter of a AAA battery and about 1/2 inch long.

I also notice that if I touch the capacitor, that is the shell, and touch other parts of the circuit I'll get a tingle going through me. It's still charging, however before I found the faulty 555 it was simply wide open and dumping strait into the battery. Only after I fixed that and it started pulsing properly did it start to get hot again.

I'd thought about putting some in parallel or series to help divide the load between them, I was just researching that today.

Edit:
Well, just for kicks I paralleled another cap in there, instantly they are both running much cooler but still getting warm.

Ewhaz,

Thought I'd chime in here. There are a number of issues here. The first one is whether or not the capacitor is actually functioning up to its stated specifications. I don't know what you are charging the caps up to, but leakage currents (internal discharge) always increase as you approach the upper limit of the dielectric material. The second major factor, which is often overlooked, is the internal construction of the capacitor. Caps that are designed as a "filter cap" are not meant to be discharged quickly. Electrolytic caps are all made of foil and plastic sheets rolled up in the can. The question is, how are the plates connected to the terminals? If the plates are connected to the terminals in one location at the end of the roll, discharging the cap rapidly produces INDUCTANCE in the turns of the roll. This will also heat the cap and slow down the discharge speed. Caps that are designed for rapid discharge are rolls of foil and plastic sheets where all of the layers of one plate are connected together on one side of the roll, and all of the layers of the other plate are connected together on the other side. Now the charge in the deepest layers of the cap can come straight out the side instead of having to go around all of the rolled layers. This is the kind of cap construction you need because it has no inductance and can charge and discharge instantly. Typically, this kind of capacitor is called a "photo flash" capacitor and is designed for instant discharge to run Xenon Flash tubes in photographic equipment.

Here is a link to a site that sells this kind of capacitor.

Photo Flash Aluminum Electrolytic Capacitors - Capacitor Industries

I hope this helps.

Peter

[Aaron]

photo flash caps

When I built my first "dual battery charger" SG, I needed 330uf 600v caps because I wanted to build everything 100% perfect to the schematic..I think that is what they were.

Anyway, I went to a photo lab at a local grocery store and asked them if I could have a bag full of discarded disposable cameras for a science project. They gave me about a dozen of them. Inside are 120uf to 300uf+ caps at a 120v up to a few hundred volts. They're free if you get them this way. I hope this helps.

[Ewhaz]

Thank you everybody, I appreciate the help!

Peter, Thank you especially it's always nice to see you contributing, especially to a struggling experimenter. I did have one question though, since you have worked closely with Bedini and Bearden in the past, you might know more about this application than most. I wanted to follow the schematic as closely as possible for my proof of concept (This is the circuit from page 46 of the free energy generation, circuits and schematics book) How critical is the capacitance of the capacitor?

I've read other peoples information on capacitors and how it's not too be too low and create a 'fluffy' charge, or too big that it fries the battery. How much leeway do I have with this capacitor as far as it's ratting? In other parts of the book they show capacitance any where from 1uf to 23000uf. Can I go with a 100 to 200uf cap without seriously distorting the operation of the charger?

I was actually already looking into that website, thank you for providing me with a link to the right capacitors. I'll send them an email and see what I can get.

[ren]

Aaron I find it interesting to note that you have achieved good results with a large capacitance cap bank. I note that John has some rather large cap banks too, probably in the range of 1-2 farads.

Then I see schematics calling for much lower. Is there a direct link between induction (and collapse) of a coil and optimal capacitance / cap size?

[Aaron]

caps and discharging

Hi Ren,

Probably the strongest charge I ever got was with large capacitance discharged with a mechanical switch.

Using 1uf or 0.1uf range, the charge was too fluffy.

A few hundred uf's were ok with small batts like 1.7ah 6v batts...takes a little while.

198,000uf at about 15v discharges are what gave the strongest load powering charges I ever got. Also, that was the only way I got the batts to charge on their own after disconnecting....like up to 1 hour!

[Peter Lindemann]

All Sizes

Ewhaz,

John experimented with every size capacitor you can imagine. I agree with Aaron. The systems that charged the batteries the best were the systems with the large capacitors discharging across a MECHANICAL CONTACTER!! The biggest system John used had three 1 Farad capacitors wired in series, for a total of 333,000uf. The system charged the cap for about one and a half seconds, up to about 15 volts, and discharged it through #4 wire to the batteries. The contacter was a copper strip on a Delrin wheel and the sliding contacts were silver tipped bronze strap.

I don't see any pictures of this exact set-up on the internet at this point, but this machine ran continuously for 8 weeks. There was one Wal-Mart garden battery on the front running the systems, and 5 batteries on the back. 4 of the batteries on the charging side stayed there all the time. One garden battery on the back and one garden battery on the front were switched every 12 hours (twice a day) for 60 days in a row.

Peter

[Ewhaz]

Alright, so big capacitors are not the issue really, but rather that the capacitors are capable of discharging very quickly without inducing it's own inductance.

The difficulty I am having is trying to get all 3 parameters together in one place, the 3.3 uf, the +400v and the electrolytic type capacitor. The flash capacitors all seem to be only about 360v maximum. It seems to me that while the circuits do not generate all that great amount of energy, everything is way overrated. I always assumed it was because the radiant energy would damage the componants even though they are running well within their range. However I've heard contradictory stories about this one way and the other.

I choose the self oscillator setup because it just made more sense to me. There is less to worry about and less parts capable of wearing out. Later after I get this proof of concept out of the way, then I am free to experiment in other ways as well. However, with wanting to be 'scientific' I'm trying my damnedest to stay within the parameters in the schematic but things like the capacitor and other things are driving be batty trying to find exactly what the schematic calls for.

[lighty]

Alright, so big capacitors are not the issue really, but rather that the capacitors are capable of discharging very quickly without inducing it's own inductance.

Inducing it's own inductance? Huh?


Anyway- in order to get fast pulse discharge response you would have to limit capacitors impedance elements as much as possible; namely ESR and ESL. Buying a low ESR and/or low ESL capacitors should help, however for most amateurs and low budget applications large capacitance/low impedance capacitors are simply too expensive and too rare.

In order to cheaply lower the impedance it would be very beneficial to form a large capacitors by symmetrically paralleling a larger number of smaller value capacitors. So, simply calculate your desired capacitance and divide it to a number of standard value smaller capacitors. You could of course measure peak frequency response and calculate a combination of multiple different values capacitors to combine but that's beside the capabilities of most amateurs and their measurement equipment. Of course many smaller capacitors would introduce a greater leak current but for this purpose it shouldn't matter.

Symmetrically paralleling means that leads from multiple capacitors should all be of exact same size and length and terminate in a single point.

So, to keep it short. If I were you I would parallel as much smaller value capacitors as possible. I would try to get low ESR and possible low ESL kind but if it's not possible (or if it's too expensive) I wouldn't lose any sleep over it. I would also pay attention to have all the capacitors the same value, model and manufacturer. I would then test all the capacitors for possible defective ones just to be on the safe side. After that, I would pay attention to have as thick conductors (copper or silvered copper) as possible and to cut them all to the same (as short as possible) length. I would then connect them all to a single point and from that point I would connect my low ESR/ESL capacitor to the rest of the system (by using as thick and as short conductor as possible).

I did it in the past when my budget limited my options but the results were very convincing indeed. It should work for Bedini type capacitive dischargers as well. It's much simpler to use bigger capacitors and simply ignore somewhat larger capacitor impedance but if you want sharper discharger impulses and distributed power dissipation you should strive to have as low ESL/ESR as possible.


Oh, and please note that every single capacitor in that setup must have adequate voltage rating for the targeted charge/discharge since paralleling them won't change their voltage breakthrough rating.


Hope this helps.


P.S.
BTW- electrolytic capacitors are the worst kind for fast discharges. Oil filled ones or specially made polystyrene ones are much better choice.

[Peter Lindemann]

Diagram on Page 46

Thank you everybody, I appreciate the help!

Peter, Thank you especially it's always nice to see you contributing, especially to a struggling experimenter. I did have one question though, since you have worked closely with Bedini and Bearden in the past, you might know more about this application than most. I wanted to follow the schematic as closely as possible for my proof of concept (This is the circuit from page 46 of the free energy generation, circuits and schematics book) How critical is the capacitance of the capacitor?

I've read other peoples information on capacitors and how it's not too be too low and create a 'fluffy' charge, or too big that it fries the battery. How much leeway do I have with this capacitor as far as it's ratting? In other parts of the book they show capacitance any where from 1uf to 23000uf. Can I go with a 100 to 200uf cap without seriously distorting the operation of the charger?

I was actually already looking into that website, thank you for providing me with a link to the right capacitors. I'll send them an email and see what I can get.

Dear Ewhaz,

The schematic on page 46 is a GENERIC circuit. In other words, there is nothing special or significant about it. Work on these circuits dates back to 2003. John and I built one of the first of these type on a breadboard I brought over to the shop. The very first one used a neon bulb wired from the gate to the anode of the SCR to dump the cap, instead of the 555 timer and the rest. The neon bulbs would run for a while, then polarize in the electrostatic field and stop working. So, the timer discharge is more reliable. The H11D1 is the high voltage version of the opto-isolator and is the only one that works in these conditions. The SCR is the best solid-state device to use for the discharge because it has the lowest voltage drop across the junction. That said, John tried a lot of different SCRs in these circuits, and they all behaved slightly differently. The 3.3uf cap was a special cap John had left over from one of his audio amplifier production models. You can see two of the yellow cylinders at this link, in the photo of "The Real McCoy"

20 Bedini

These capacitors are fairly large for their capacitive value and were good to about 300 volts, I think. The Real McCoy would charge them to about 170 volts before discharge. In fact, The Real McCoy has the same output circuit as the diagram on page 46, but runs a miniature mono-pole motor with a Cole Switch on the front end instead of the Self-Oscillator. The point is, John tried every combination you can imagine.

None of the circuits were Over Unity. Every circuit had losses. The only systems that ran OU were the ones that continuously charged and discharged the batteries repeatedly. Only after long study did we draw the conclusion that it was the BATTERY that was operating OU, not the circuits!

The bottom line, and I have said this over and over, do not expect these solid-state chargers to be OU. They aren't!!! The OU of the SSG appears as mechanical energy on the shaft!

On a side note, I just found a picture of John's machine with the 3 one Farad cap discharge system at: Welcome to Bedini Technology The photo is near the bottom of the long page.

Also, the simplest way to solve your cap problem is to use four of them in a series/parallel set-up. Make two sets of two of the caps in series, and parallel the two sets. This will give you a capacitor that has the same capacitance as the original one, but it will be rated at twice the voltage. That should solve your heating problem.

Good luck,

Peter