MMC for 24kV: Thoughts on best construction? / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

MMC for 24kV: Thoughts on best construction?

previous 1 2

Move Thread

LAN_403

J. Aaron Holmes

Mon Jul 24 2006, 03:19AM

Registered Member #477 Joined: Tue Jun 20 2006, 11:51PM
Location: Seattle, WA
Posts: 546

Hi, Terry!

Terry Fritz wrote ...

To get numbers, you have to give numbers...

Peak voltage?
Peak current?
BPS?
Primary frequency?
Primary capacitance?
Primary inductance?

I hear 200:1 and 24kV and 15nF...

Fair enough. My interest is pretty general, though. I only have a spark gap and transformer for the coil at this point, however. The rest is still being imagined! So I don't have all the numbers you're requesting. I'm basically designing for a power level that won't set my house on fire, since my garage ain't that big! I want to build a cap that'll last through a few redesigns, and am deliberately avoiding some specificity

Your guesses were correct, though: 200:1 PT; 24kV. Two strings of 20 sound good for 15nF. Definitely affordable. Regarding your question about primary inductance, you reminded me of a question I have about JavaMMC (perhaps best answered by Bart, but...): Why does it ask for "Primary Resistance", and how does one measure that? I assume this includes the spark gap, or else the sample (3 ohms) would involve one heck of a lot of wire in the primary winding!!

Hmmm... I've heard "3 ohms" passed around a lot as a typical spark gap resistance, so perhaps that's sufficient for most early design work (think I heard this in a SISG thread, actually!

)

Best Regards,
Aaron, N7OE

Hazmatt_(The Underdog)

Mon Jul 24 2006, 08:17AM

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

Here's my $.02

All I run all the time are TDK Ultrafast 2000pF Doorknob Capacitors. They work FANTASTIC. THEY DO NOT GET HOT. THEY DO NOT EVEN GET WARM.

MICA doorknob capacitors DO FAIL miserabily. The mica shorts out badly. DO NOT use old SPRAGUE mica doorknobs in a TC. I have 12x of these older capacitors and they DO NOT and WILL NOT survive pulse application. They are for DECOUPLING only. I have run the old Sprague's up to ~300W RF but they will not handle SGTC duty above that.

There ya go for doorknobs.

Matt

Steve Conner

Mon Jul 24 2006, 09:28AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Like I've said before: there are two kinds of ceramic capacitors: ones meant for bypassing and DC blocking, and ones meant for RF power in resonant circuits. Doorknobs are mostly ceramic so they also come in these two flavours. The RF power ones (surprise) work in TC tank circuits and the DC ones fail. You can usually tell the RF ones because the data specifies a kVAR rating. The DC ones have a DC working voltage.

J. Aaron Holmes

Mon Jul 24 2006, 03:45PM

Registered Member #477 Joined: Tue Jun 20 2006, 11:51PM
Location: Seattle, WA
Posts: 546

Kolas wrote ...

I saw the guy in China selling them too, he says, he's the manufacturer of the caps, and I will ask him for a data sheet now.
The reason the door knobs were considered is; they would actually be more cost effective then a bunch of series\parallel caps such as your thinking of.

I'll be interested to hear what he says. Still, 1100pF/50kV at $18 a pop works out to about $250 for a 15nF MMC. 4x40 CDE 942 150nF caps at $2.25 a pop works out to $180 for a *30nF* MMC (that people think can reasonably tolerate my 34kV peaks). So it would seem as though you literally get more "bang" for the buck that way, not the other way around as you suggest.

Though it occurs to me that the doorknob route, by virtue of being far more parallel than series, may take the high currents a bit better, enabling higher breakrates to more than make up for the smaller capacity (though some people have reported that higher breakrates only get you so far). Plus, with only a single "layer" of, e.g., 50kV doorknobs in your MMC, if one of them fails short, you aren't overvolting any of the others as you would be with series strings of caps. I don't know what the "typical" failure mode of such caps would be in a coil, but...

Regards,
Aaron, N7OE

J. Aaron Holmes

Mon Jul 24 2006, 04:15PM

Registered Member #477 Joined: Tue Jun 20 2006, 11:51PM
Location: Seattle, WA
Posts: 546

Also, perhaps I could clarify the peak string current calculation: This is just the usual peak primary current calculation divided by the number of MMC strings, yes? I.e.:

Istring = Vpeak * Sqrt(Cpri / Lpri) / #strings

So, imagining a primary of Lpri = 50uH, a 2x20-cap 942 MMC of 15nF, and a Vpeak of 34kV, I get:

Istring = 34000V * Sqrt(.000000015F / .000050H) / 2
Istring = ~290A

...which seems like plenty of headroom.

Regards,
Aaron, N7OE

EDY19

Mon Jul 24 2006, 06:27PM

Registered Member #105 Joined: Thu Feb 09 2006, 08:54PM
Location:
Posts: 408

You might check out this auction, too. It is a 35kv 30nf pulse capacitor- looks OK to me.

J. Aaron Holmes

Mon Jul 24 2006, 07:07PM

Registered Member #477 Joined: Tue Jun 20 2006, 11:51PM
Location: Seattle, WA
Posts: 546

EDY19 wrote ...

You might check out this auction, too. It is a 35kv 30nf pulse capacitor- looks OK to me.

I blew one of these to pieces about two weeks ago (seem my original post above). General opinion on the TCML seems to be that you have to be gentle with these (keep V down, breakrate low) or they croak. The problem I have with them is, as I have experienced, they're "all or nothing". If one cap blows, that's the whole cap!!

Regards,
Aaron, N7OE

Steve Ward

Mon Jul 24 2006, 07:19PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

Doorknob caps can and do work (strontium titanate types), but they only fail *one* time. Maxwells have this same problem. If you use either of these types, be very careful of the voltage reversal rating. Maxwells are often 20% reversal, which means that if you discharge one at its max rating, the negative excursion (due to LC oscillations) should only be -20% of the input voltage. With a tesla coils, you will experience basically 100% voltage reversal, so this means to be safe, you must only charge the capacitor to 50% of its rating. So even 50kV doorknob caps wont last long on a TC with 34kVDC peak charging voltage. You would need at least a 70kV rating on the doorknob caps. The maxwell's will likely need a 50-70kV rating depending on the reversal ratings.

So i say with years of experience with using doorknob caps and PP MMC caps, use the PP caps! You can even run them at their DC rating (even with 100% reversal!) and they last a long time. So for CDE MMCs, DC peak charging voltage can equal DC peak rating of the caps, but for maxwells and doorknobs, double their voltage rating.

previous 1 2

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