Capacitor Woes / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Capacitor Woes

previous 1 2

Move Thread

LAN_403

Proud Mary

Tue Sept 15 2009, 12:39AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

There is no cure for not selecting the right parts for the job. Over- rate every thing by a good margin and your problems will most likely disappear.

Questions here often revolve around dollars, sometimes thousands of dollars, over a year.

Herr Zapp

Tue Sept 15 2009, 04:18AM

Registered Member #480 Joined: Thu Jul 06 2006, 07:08PM
Location: North America
Posts: 644

Bird-Dawg -

A few questions and comments:

1. In your first post, you mention 470pF, 30KV USSR surplus ceramic capacitors, but the eBay link you provided shows .01uF, 10KV wound film capacitors. Please clarify which type of caps you actually used.

2. The .03uF, 35KV Maxwell pulse caps (37667) you mention have been available on the surplus market for quite a while, and many have been used in Tesla coils. However, a fair number of these caps have failed in coils powered by 15KV, 60ma NSTs. The consensus seems to be that they fail by overvoltage (voltage reversal), and that two must be used in series for reliability. Two in series yield a capacitance value that's suitable for a 15/60 static-gap TC, but is much too small for best results with a 15/120 NST power supply.

3. The surplus Maxwell 37667's were removed from medical laser systems as they approached the end of their "rated life". However, you have no idea how close to failure these "well-used" caps actually are. They are RATED at 3.0E08 charge/discharge cycles, but there is no way to determine how many cycles the caps have actually seen, or how close to failure they may actually be. Are you willing to risk the money on buying two (or four) of these capacitors, and having no idea whatsoever how long they'll run before catastrophic failure? They have a decent peak current rating (25K amps), but the rated voltage reversal is only 15%.

4. Cornell-Dubilier 942 series film caps are well-proven in severe Tesla coil use, and can be purchased for around $3.50 ea in quantity. In a TC powered by a 15/120 NST power supply, a correctly-designed MMC using the C-D caps is indestructable. You'll need about 40 caps for a 15/60 power supply, and about 80 caps for a 15/120 power supply.

5. The C-D 942 .15uf, 2KV caps are rated at 432A peak current, and 13.5A RMS current. Compare those ratings with ANY of the other HV film caps offered on eBay, and you'll see that there is NO comparison. Most of the eBay film caps won't even show the peak current values because these are not really designed for pulse service (they're DC filter/coupling caps).

Do the cap-cost math, and consider the potential reliability problems with the surplus Maxwell caps.

Regards,
Herr Zapp

Bird-Dawg

Wed Sept 16 2009, 12:01AM

Registered Member #2360 Joined: Sun Sept 13 2009, 05:43PM
Location: Kennesaw, Ga USA
Posts: 14

Updated the link with the correct one.
Well with you guys help i guess i will get the cde caps. I try to go for robustness will building my coils over engenering evrything and expecting the worse. I just had some pre conceved notion in my head that the cde caps where not nearly as durable as the cermic and maxwells (wich Herr-Zapp is right that is quite a gamble for the money) but when you do the math the cdes win. Thanks for all yalls help, you guess have gone beyond the call of duty with answering my question.

P.S. are those bleed resitors on the caps really needed?

Arcstarter

Wed Sept 16 2009, 03:12AM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

Yay CDE!

Bleeders should be used. They can also balance the voltage in each cap, which is a good thing, but furthermore they protect you from having a cap or two with enough voltage to give you a little zap. As much as i doubt it, there *could* always be the strangest circumstances, and you could fibrillate your heart, or kill yourself. No telling. Nerve damage is more likely. Also, it will make you jerk if you touched a charged cap. You can hit something and break a bone, dislocate a joint, break something of value... Not to mention fire from a stupidly placed cap bank.

Life is too fragile to put yourself in immediate danger just to save 3-7 bucks, but if you do not have some fun in your life, are you really living in the first place?

In short, continue building fun things, just spend a little extra to prevent nerve damage or something else.

Hazmatt_(The Underdog)

Wed Sept 16 2009, 08:28AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Yea, I'd say 500K resistors across the caps would be good. I got a hell of a shock long ago after the bank was off from a single cap I had removed from a string.

Typically with the transformer in parallel with the cap you don't have this issue because the cap discharges through the transformer. I recommend doing it this way for safety.

Herr Zapp

Wed Sept 16 2009, 04:37PM

Registered Member #480 Joined: Thu Jul 06 2006, 07:08PM
Location: North America
Posts: 644

Bird-Dawg -

To determine the appropriate resistor value for bleeder resistors used on an MMC, do a little research on "R-C time constant". You will want to select a resistor value that will reduce the voltage across each capacitor to a relatively safe value in a reasonable amount of time (several seconds) when the power is turned off, without drawing too much current when the coil is operating.

The bleeder resistors must also be rated for the appropriate voltage. The Vishay VR-37 series resistors are rated at 3500 volts vs the 350 or 500V ratings of common metal film or carbon composition resistors, yet only cost a few cents more.

A typical value for MMC bleeder resistors is 10 megohms.

Regards,
Herr Zapp

previous 1 2

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