From: Paul
Date: Fri, 01 Jun 2001 20:40:56 +0100
Subject: Re: [TSSP] Secondary in motion
Terry, All Regarding the movies, you wrote: > burn the "big ones" ... onto CD ROM > I can put them on my site > These movies are historical in Tesla coiling theory... Gosh, we'd better get them right then! As you suggest, we've got a few options. Pity you're not running Linux - you could design and create these yourself, the program to produce the movie is only 75 lines, but it's a unix shell script. Have you got windows tools that will create mpeg from a data file? Although the sim code is still a building site, it's together enough for me to send you a set of working code to generate the data, at least for 'normal-ish' coils. But I agree, these movies are a milestone, and we should distill them down to a couple of definitive examples. Marc wrote: > "gimme more" comes to mind > ...could a grid be added... OK, I'll do one or two with grids and axes on, and a few other trimmings, but they'll be about 10-12 Mb. On the whole, I feel my efforts are best concentrated on getting a revision of software issued - because as long as I'm sitting on the code it prevents others from tackling the many interesting questions which now present themselves. Hmm, thinking about it, MPEG is lousy for this, for the same reason that JPEG is no good for diagrams. Do you guys have tools for stepping through animated GIFs for example? Or do you have tools to locally animate say a zip or tar archive of frame-numbered GIF images? We should check out these other options. Terry wrote: > Sometimes the question comes up if it makes a difference if the > primary and secondary are wound in the same direction, opposite > direction, counter clockwise, clockwise... It is usually assumed > that makes no difference at all. Do you see any reason to think > otherwise? Don't see any mechanism for gross difference. Reversing a winding changes the sign of the magnetic coupling, but not the capacitive coupling between pri and sec, so there is opportunity for a small difference, but the software doesn't model this. We allow for the pri-sec distributed cap, but the software takes the primary to be at earth potential as far as the E field is concerned, so any winding-sense inspired difference would be invisible to the model. Marc wrote: > ... it seems to me that at higher k's the pulses of power to the > oscillations are coming at the "wrong" times. You've probably looked at these waveforms more closely than any of us! To be honest I find it hard to make sense of the behaviour when looking at the movies - there's just so much going on, especially with the k=0.35 model. > by this i mean that the next "push" to the "ruler" is striking an > opposing "flex"? Yes, keep in mind that half the time, the primary induction is acting in a direction that will be taking energy out of the secondary, so perhaps this is what you're seeing. > i also noticed that at these points of "out of sync." pulses, there > is a corresponding sharp valley to the voltage gradient. You're at a slight disadvantage when looking at the waveforms in this detail, because I rectified the dv graph, so it's not too clear whether the primary is pushing or pulling, so to speak, > if we look at the set for k=0.35 at 70% of the second cycle, this > seems to be one of these "out of sync." pulses, a strong valley > corresponds with it in the lower 25% of the windings. then if we > follow that back to 60%, yes, I'm with you, that rapidly moving sharp notch - where the gradient changes sign, but I'll not try to make sense of it. For this kind of close examination, it would be better if the v and a non-rectified dv chart were together on the same movie. > the sharp valley moves up the coil to center, so it is traveling > down the secondary windings as the o.o.s. pulse is formed. > maybe this could be some of the cause of the differences in > potential of the windings, contributing to breakdown across them or > to the primary? There's certainly some violent changes in gradient occuring, but its hard to make sense of it, especially without the grid lines, etc. One thing we can say is that the v/turn in this high k setup is getting on for 50% higher than in the k=.20, for no extra gain in topvolts, so the k=0.35 coil would therefore have a smaller 'largest tolerable toroid'. Terry, I'm still trying to figure out the tuning on your k=0.35 setup - it's an interesting one, for sure! Terry wrote: > Using my Tesla coil tuner... primary system is 242.31kHz... > ...removed the secondary...primary ring down frequency..218kHz Ah, its the 218 kHz that I'm after. That's the real Fres of the pri. When the secondary is dropped in, things will look pretty different. Can you measure Lp directly - it should be around 19.24uH. If not, then I join you in confusion! Its the weekend now - thats official, and I'm going to spend it, first modeling this high k coil again, and then I'll make my priority to get a temporary release of the modeling code posted up. Cheers All, -- Paul Nicholson, Manchester, UK. --
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.