From: Paul
Date: Tue, 12 Jun 2001 12:09:46 +0100
Subject: [TSSP] Progress at 12/05/01
The last few weeks have been quite busy, with progress on a number of fronts. Here's a quick summary of the prevailing issues. Time Domain Modeling -------------------- We've demonstrated one decent comparison with measurements, see http://www.abelian.demon.co.uk/tssp/tfss270501/ but a test at k=0.35 did not do too well - predicting too much of the higher modes, and the frequencies were well out. Probably trouble with the caps matrix when using the close cylinder primary. Still to be resolved. The 'secondary in motion' movies were entertaining, but not especially useful for figuring out what's going on. See the time domain modeling page on the website for more details. Internal Capacitance -------------------- Terry referred us to another paper by Randall Rhea, http://www.amwireless.com/archives/2000/v12n11/pg30.pdf which if nothing else, highlights how far our understanding of the coil's behaviour has come, in that most authors are content to dismiss internal capacitance. This has prompted us to consider publishing this aspect of our work, hence a flurry of activity to confirm the model's predictions for the small h/d regime in which Cint is dominant. Yesterday, I added in code to do a first-order (euphemism: a guess) adjustment for material dielectric. On Thor and Kurt's coils the result is very nice. A recent change to the Cint calculation has also improved the results, mainly at large h/d, by using an interpolation with better global properties. Unfortunately it introduces a noise at short-range so I'm having to re-write it yet again. An unequivocal measurement of the current profile of an h/d < 1.5 coil would be the best confirmation. Q-Factors --------- Still as poorly predicted as ever. But at least we now have a strategy to tackle this. Racing Arcs ----------- We may now have the tools with which to predict instantaneous voltage gradients on the secondary, but we cannot predict racing arcs unless we have an idea of the secondary breakdown gradients. These we can deter- mine by experiment if we had a means to measure top-volts. Need to investigate the influence of factors such as coil radius and wire size. Streamer Length --------------- A prerequisite for any sort of Tesla coil optimisation is a means to relate streamer length to top voltage. As above, if a reliable top-volts measurement technique could be devised, then we could gather the necessary statistics. Breakout Voltage ---------------- Something else which depends on top-volts measurements. We have the possibility of predicting breakout voltages using E-Tesla's field plotting. The method gives plausible figures, but this needs to be tested against actual measurements. Brush Discharge Impedance ------------------------- Terry has made some measurements on a brush discharge which demonstrate that a time-averaged load impedance adequately describes this type of discharge. http://www.abelian.demon.co.uk/tssp/tfcw260501/ which is a great step forward for modeling CW coils. We now need to find a way to predict those time-averaged discharge impedances - again a top-volts measurement issue. Software -------- Some new webpages, eg http://www.abelian.demon.co.uk/tssp/smap.html I'm going to move all the software docs into html format for ease of reference and update. pn1205 technical notes will be retired and replaced by a mixture of web pages, and a new technical appendix, pn1904, which will contain the math details of how the model is implemented. I'm working towards a proper release of the software, a week or two off yet I think. I'm deleting code that is only experimental or un-tested, eg the center feed and non-linear stuff. Running on Windows ------------------ Terry has tested the software under the Cygwin environment and it runs very well. The unix shell emulation is sufficiently good to allow some of the scripts to run too. As soon as the software is in a fit state, I'll ask Marco to recompile it to produce some native Windows binaries. Priorities ---------- Measurements of the higher mode frequencies on medium to large coils with h/d < 1.5, having very thin former materials. Find a way to measure peak top volts. Finally, some administrivia, List Archives ------------- TSSP list archives are now available through http://www.abelian.demon.co.uk/tssp/list/ Threads are listed in order of the date started, oldest first, and postings are listed under each thread in date order. Don't bother book- marking postings, the URLs will change. I'll fix this when I figure out how. Website ------- Project website is at http://www.abelian.demon.co.uk/tssp/ I've recently updated a lot of the pages, including the things-to-do list. Have a browse through it all. Docs ---- pn2511 and the draft pn1401 have both been revised over the weekend. Only minor changes. Still at least 2 sections to add to pn1401 before it's 'finished', but it's up on the web anyway in order to encourage feedback. Mailing List ------------ If you no longer want to receive tssp list postings, then email to majordomo@abelian.demon.co.uk and put unsubscribe tssp in the body of the message. -- Paul Nicholson, Manchester, UK. --
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.