Strontium Titanate versus Mica for VTTC tank cap / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Strontium Titanate versus Mica for VTTC tank cap

Move Thread

LAN_403

J. Aaron Holmes

Fri Jan 26 2007, 06:18PM

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

With the sudden availability of inexpensive brand new strontium titanate doorknob caps from China, the temptation to use these for a VTTC tank cap is irresistable. But what are the drawbacks? I know that Steve Ward and others have had success with these, yet mica continues to be recommended.

I would think that, for short duty (and particularly if one is using staccato), the doorknobs would be ok even at higher powers, however if one was expecting long runs at 60pps, heating might become a major concern (just as with the tube). I'm wondering if Steve and/or other VTTC'ers have experienced any issues with doorknob temperature stability resulting in detuning during operation? It just seems like very low pps would be the only way to keep things in tune at higher powers when using ceramics (I suppose this is true of SGTCs too, but you don't hear of doorknob use so often there...)

...the mica transmitting caps in the 1000+pF @ 20+kV range seem extraordinarily expensive when purchased from reputable sources, and thus far I haven't managed to find any suitable ones at hamfests.

My tube collection is growing; I've now also collected a 3-1000Z. It really wants to be in a coil!

Regards,
Aaron, N7OE

Steve Ward

Fri Jan 26 2007, 09:50PM

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

It completely depends on 3 factors. The peak voltage in the circuit, the frequency of the resonance, and the dissipation factor of the dielectric. I know that SrTi dielectric has a higher dissipation factor than mica, but im not sure how much. When i used a single 2nF 40kV doorknob on my last tube coil, it ran slightly warm with the 12A line current (using staccato) to give you some idea. My dual 833A coil used a series parallel arrangement of 4 of these caps (double the current capability). These caps ran pretty much at room temp when pumping out 2 foot sparks at 30bps for long run times. Personally, i find this to be quite acceptable. I have used both the SrTi dielectric caps from TDK (the UHV series), and MuRata (the white ones). They seem basically identical from what i can tell. Something of importance may be to use a very good voltage safety margin with these capacitors. I dont think that the 40kV rating was overly engineered, given that the capacitor will see *at least* 5kVAC, plus some other voltage depending on the Q of the loaded tank circuit (might be something like 2-3X the plate voltage). I think i may run some simulations to see what is actually expected.

I have bought good mica caps on ebay from time to time for around 20-30 bucks for a 3nF 10kV unit.

Hazmatt_(The Underdog)

Fri Jan 26 2007, 09:58PM

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

I can't say how well ceramic doorknobs work in a VTTC because that's CW operation compared to a SGTC. Doorknobs fair very well in SGTC's. They do not exhibit much heating at the power level I am working at. They handle very high currents in short times. My tank currents were in the 300A range with 9 2000pF 40KVDC TDK cpas.

I switched over to ceramics since my mica doorknobs are in poor shape. The older Sprague mica's will work for you at 300W input, but after that they started shorting badly in pulse discharge. They are good for decoupling, but have some very definate limits when it comes to pulse discharge.

I would like to test them on a VTTC, but in looking at the work required versus output, it doesn't seem reasonable to me. So I'm going to put that one off for the future.

J. Aaron Holmes

Fri Jan 26 2007, 11:06PM

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

Thanks, Steve. Yeah, I thought about just putting a few doorknobs in series/parallel like you mentioned. Your experience makes me feel better about trying that now. 'yishdz' on eBay seems to be very reasonable, price-wise.

Hazmatt_(The Underdog) wrote ...

I would like to test them on a VTTC, but in looking at the work required versus output, it doesn't seem reasonable to me. So I'm going to put that one off for the future.

I don't know why, but ever since seeing a relatively low-powered VTTC at a hamfest (5...maybe 6-inch sparks), I've been dying to build one. The strangest part is, whenever I look at a VTTC, I feel like I'm looking at technology older than what's in an SGTC. I don't know why that is!

Regards,
Aaron, N7OE

Hazmatt_(The Underdog)

Sat Jan 27 2007, 12:36AM

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

Let me put it this way, my plan is to build an 811A push-pull VTTC. That's going to have to wait until I can devote the energy to it because of the way I want to do it. I want to optimize it and operate it in an impedance matched output, rather then driving the plates to cherry red with 4" sparks.

I'd really like to have one too, I just want to do a good design job and be thoughtful about its design.

J. Aaron Holmes

Sat Jan 27 2007, 12:58AM

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

I see. Yep; from what I've heard, efficiency in tube coils is a challenge. Personally, for a first tube coil, I'm sticking with the "easy" way

Steve Conner

Sat Jan 27 2007, 01:03AM

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

I think most VTTC builders do end up getting close to optimal by trial and error. The ultimate problem is that streamer loading is a non-linear function that nobody knows how to model, so you can't optimise Tesla coils by simulation. You can sort of do it with disruptive coils and DRSSTCs, because the short pulse length lets you make a lot of simplifying assumptions. But a VTTC is just an impedance matching network between a tube plate and a bunch of plasma, so there's no getting round the fact that you need to know the impedance of the plasma to design it, but there's no way of simulating it. So trial and error is as good a way as any.

Those glowing plates and 4" sparks are just a consequence of the facts that tubes aren't very efficient and it takes a lot of power to make sparks with CW. If you get 70% efficiency with tubes you are doing great. At the frequencies VTTCs typically run (low 100s of kHz) you can get over 95% with MOSFETs without even really trying.

Where tubes really come into their own is that the tube will still be 70% efficient at 30MHz, as opposed to the MOSFET H-bridge which would explode instantly.

Dr. Drone

Sat Jan 27 2007, 04:16PM

Registered Member #290 Joined: Mon Mar 06 2006, 08:24PM
Location:
Posts: 1673

J. Aaron Holmes

Sat Jan 27 2007, 05:59PM

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

I'm really glad to hear folks haven't had stability issues with the SrTi caps. I'll definitely take that route for the first coil. They're just too cheap to pass up, even brand new!

To Chris: That makes a few folks working on really big VTTCs now. Can't wait to see the results!! For me, I'll be happy just to get something working

I was excited to find the 3-1000Z though; it has a plate voltage rating of 6kV! I'll probably stick to the 3-500Z for my first coil, though. Got one of those a short bit ago, which is what got me thinking about building a VTTC after many years of interest.

Thanks all!
Aaron, N7OE

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