Potential Unresolved General Solid State Tesla Issues to Investigate / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Potential Unresolved General Solid State Tesla Issues to Investigate

previous 1 2

Move Thread

LAN_403

Chris_Knight

Tue Feb 09 2016, 01:13PM

Registered Member #58280 Joined: Sat Jan 09 2016, 06:48AM
Location:
Posts: 43

Mads & Sulaiman that sounds doable. I like it.

Just a couple more issues...

Wouldn't it be easier to investigate this with a SSTC? Then no changes to the primary would have to be made and no messing around with MMC taps and/or primary taps

So it seems to me that the toughest part of this experiment is maintaining a single resonance frequency across all the different coils. Now obviously I can't go and buy 10 different spun aluminum toroids, and plus it is unlikely i would be able to find off the shelf toroids that exactly matched the capacitance I need... One option would be just to make a bunch of aluminum ducting toroids, but i feel like there would be too much variances between them. Another idea I have is to spray paint a bunch of plastic spheres with metal paint and use a large breakout point, as obviously break outs from the metal paint would be disastrous... suggestions? How would I fine tune the capacitance of the topload?

Mads Barnkob

Tue Feb 09 2016, 07:01PM

Registered Member #1403 Joined: Tue Mar 18 2008, 06:05PM
Location: Denmark, Odense C
Posts: 1968

It might not even be possible to conduct such a "only change secondary wire size and topload" experiment, it will very fast just turn into a "how much topload weight can a secondary carry".

I used cheap parts like orange pvc sewer pipe, flexible ventilation aluminium ducting dimensions.

Here I kept the same primary parameters, I kept the same secondary size, only variation was the wire diameter and number of turns, adjusted topload to get the same resonant frequency.

But I gave up after making the first, as the amount of topload needed to drop the added 20kHz for each wire size jump is just getting redicoulos, my plan was to make setups for 0.25, 0.3, 0.35, 0.4, 0.45 and 0.5mm wire, but I only got to the 0.25, 0.3 and gave up on 0.35.

However I twist and turn it, even if I start in the middle and go both ways in topload size, it just gets too different at the outer bounds and will be a large vs small topload test, and even there we already know that DRSSTCs like large toploads.

Chris_Knight

Tue Feb 09 2016, 11:17PM

Registered Member #58280 Joined: Sat Jan 09 2016, 06:48AM
Location:
Posts: 43

Hm, you're right. I did some playing around and I ended up with huge unrealistic toploads.

So I've been playing around in Java TC and just thinking on my own for about an hour or two, trying out different ideas. I think I have three different ideas that might work.

1. Maybe we've been thinking about this too much at the mico scale; considering all the variables that a thing like height changes. How about a simple question: does more windings mean longer sparks? Basically take several secondaries completely identical other than the number of turns (and thus the height) test them, and see if there is a linear relationship between turns and spark length. It's not going to be linear, but it would be interesting to see what kind of relationship, if at all, there would be. Same thing with width. I understand that we're testing H/D in a way, and we'll be changing several electrical variables, but I think that's acceptable considering tesla coils are usually designed from a physical point of view.

2. So I went back to the H/D aspect ratio idea. I thought about testing multiple different secondaries with different H/D ratios. How about when the diameter is changed, change the height to compensate for the same inductance, and thus the same frequency, as before. Of course you'll be changing the number of turns, but it would be interesting to analyze the data I think...

3. This is the most straightforward of the three. (i.e. I couldn't think of anything else) Compare the standard schmitt trigger antenna feedback system for gate drive signal generation with a 4046 PLL system in several difference coil sizes (small, medium, large secondary) and see if there are any performance differences. My hunch is that the PLL will be slightly better due to the switching delay of the standard system.

Thoughts? Am I on to something or am I completely off the track?

nzoomed

Wed Feb 10 2016, 12:29AM

Registered Member #54503 Joined: Sun Feb 22 2015, 10:35PM
Location: New Zealand
Posts: 288

Im totally lot when it comes to toploads.
Ive used javaTC when designing my coil and i see some people use two or more toroids to supposedly increase the capacitance of the top load.

I take it that it all goes by surface area?

The weird thing is that ive added an extra small topload with javaTC and it barley makes any difference to the resonant frequency?

Is there any rule of thumb when using multiple toroids?

Sigurthr

Wed Feb 10 2016, 01:52AM

Registered Member #4463 Joined: Wed Apr 18 2012, 08:08AM
Location: MI's Upper Peninsula
Posts: 597

Chris_Knight wrote ...

3. This is the most straightforward of the three. (i.e. I couldn't think of anything else) Compare the standard schmitt trigger antenna feedback system for gate drive signal generation with a 4046 PLL system in several difference coil sizes (small, medium, large secondary) and see if there are any performance differences. My hunch is that the PLL will be slightly better due to the switching delay of the standard system.

Thoughts? Am I on to something or am I completely off the track?

I've done this two times, both by changing only the driver circuit and nothing else on the coils. Even at 1.2kW there was no visible difference in output characteristics other than using a PLL (which is oscillator driven) is ground-strike tolerant because the feedback never drops out when you heavily load the output. I didn't test bridge thermals because I overengineer my bridges to the point they can run CW at unlimited duty, so there was no detectable difference in thermal performance at the point I operate.

Thinking back a bit there is a design that to my knowledge no one has been able to perfect yet for either SSTCs or DRSSTCs; the push-pull primary. The difficulty lies in having equal coupling between the two halves of the primary. I remember thinking that an interleaved primary could possibly balance out the coupling, but never actually tried it. This would let someone use a traditional Mazzilli style driver to directly drive a SSTC (assuming they got the LC to match the secondary's f0 ), this in effect would be a new type of DRSSTC; a CWDRSSTC.

Chris_Knight

Wed Feb 10 2016, 03:16AM

Registered Member #58280 Joined: Sat Jan 09 2016, 06:48AM
Location:
Posts: 43

That's interesting... what is an interleaved primary?

I feel like I've seen a bipolar tesla coil use push pull, but i might be wrong.

I don't think the Mazzilli works at the high frequencies of SSTC. Maybe something like a push pull class B would work.

That's something I am interested in working with, however that seems like a long term project. Are there any short term questions I could answer? At this point I'm leaning towards doing different length secondaries, different pri-sec coupling damned!

EDIT: Think I figured it out. Compare the audio quality and affect on tesla coil performance of the the three main methods: PWM, FM in PLL, and AM in class E. Any issues?

Sigurthr

Wed Feb 10 2016, 07:14AM

Registered Member #4463 Joined: Wed Apr 18 2012, 08:08AM
Location: MI's Upper Peninsula
Posts: 597

Re: interleaved primary;

If you take a standard 90deg SSTC primary and ad a center tap, and try to feed the DC rail from that center tap and switch the extremities to ground in alternation the "lower" half of the primary has a far lower coupling to the secondary than the "top" primary does. Current will be imbalanced between the two primary circuits, and the coil will be alternating two "on" power states.

You'd have to actually use two primaries where the space between coil A's turns is filled by coil B's turns. You then have to figure out if you wind both in the same direction and alternate phase on one coil, or wind the coils in different directions. (I'm not sure about this to be honest, and figuring out magnetics always gives me a headache.)

You can make Mazzilli work over 500KHz no problem, I've done it using an air core output transformer instead of a ferrite transformer. The parallel LC version with two input inductors works even better at high frequencies than the standard center tapped one too.

previous 1 2

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