From: "Terrell W. Fritz"
Date: Thu, 19 Oct 2000 19:58:25 -0600
Subject: Re: [TSSP] Progress report 18th Oct 2000
Hi Paul, At 12:45 AM 10/20/2000 +0100, you wrote: >A quick note to summarise the project status, as quite >a bit of ground has been covered lately and I just wanted >to paint a clear picture of where we're at. > >Theory notes pn1310 at > > http://www.abelian.demon.co.uk/pn1310.ps > >section 5 onwards contain simple, physically justifiable >extensions to transmission line theory in order to >accomodate the longitudinal couplings of a tesla coil. > >These extensions are self consistent and converge, in the >absence of longitudinal coupling, to the standard >transmission line solutions. The theory forms the basis >for this project's tsim simulator which is known to do a >reasonable job of predicting resonant frequencies and >the dispersion due to the longitudinal coupling. > >A number of other testable predictions emerge: > >A) Expected V/I profiles are described in pn1710 at > > http://www.abelian.demon.co.uk/pn1710/ > >and some of these examples could and should be checked by >experiment. > >B) Also pn1310 equations 7.7 and 8.4 place approximate bounds >on the resonant frequency when Cmed is either calculated from >the geometry by a laplace solver, or measured for the coil >in-situ by the 'Medhurst/Howe' method. This could be confirmed >by reference to already published measurements, and by new >measurements. > >C) Further, predictions of transimpedance, equations 9.1 and 9.2, >can be put to the test by simultaneous CW measurements of >base current and top voltage. > >Finally, there are a number of interesting consequences of >the basic theory pn1310, which may be hinted at in some of the >examples in pn1710. These may possibly have some bearing on efficient >tesla coil operation. Obviously these would be contingent on >satisfactory experimental confirmation of the points listed above. Very wonderful stuff you are doing here!! These calculations are an order of magnitude beyond anything I have seen before! I do notice that the voltage profile you calculate does have a little concavity to it on the 0 - 300 turn region: http://63.225.104.49/TeslaCoils/Misc/PaulNich/Paul.gif That's a start :-) Of course, no where near what the profile I pulled from E-Tesla5 had: http://63.225.104.49/TeslaCoils/Misc/PaulNich/ET-5.gif I assume your profiles effectively account for the radius and area at the top of the secondary. We seem to have a very large disparity there too... I am beginning to wonder if the voltage profile along the secondary needed to arrive at a given "answer" is not unique. But rather a set of profiles. The finite element analysis is straight forward but there may be a large number of odd curves that would produce the right answer. They may also be linear in that they will scale for various H/D ratios. In otherworlds, the profile in E-tesla5 may tend to give correct answers but it is just one of the odd profiles of a much larger group that would also give that answer. My voltage probe test may have naturally zeroed in on one of these erroneous profiles. I can imagine some pretty odd profiles that could be optimized by trial and error to 'produce' a correct numerical answer. Getting that odd profile to be universal would be a far greater trick but there is no theoretical reason why it could not be done. It is also possible that the scaling could be off and there is a second scale error in the program (probably the obnoxious cylindrical integration). This would open the possibility that your profile would work. There are many things about E-Tesla5 that simply "have to work". But there are a few soft spots were errors in the code could do such things. > >It has become clear that some considerable attention needs to >be paid to finding ways to produce reliable unequivocal voltage >profile measurements. Existing methods are producing conflicting >answers. Here's hoping the experimenters out there will rise >to this stiff challenge. There may be a way to prove the profiles. Basically, verify or eliminate erroneous profiles from the "set' of profiles. Hopefully, one of us will be right and we both won't have to start over :-)) I will try and confirm the technique and get back to you after I have run a few calculations. I don't think it can be proven that this will be absolute, but it will eliminate 99.9% of the erroneous profiles... > >Just few general points about theory notes and things. I've posted >up a few items lately but not had much feedback from the list >apart from Terry. Not having any other review process available >I'm completely dependant on the list members to scrutinize project >material and challenge anything that's at odds. We've seen recent >examples where faulty ideas have gone unchallenged by critical >review and thus worked their way into the common understanding. >I'm sure we're all anxious not to contribute to this and our >only safeguard is the willingness to criticize one another's >efforts. Only then can we claim to be engaged in a scientific >process. I'll be the first to admit that when the integral signs get above two in number, my eyes start to glaze over :-) But the proof is in the results and how they compare to what can be measured in real life. We can now take good measurements of many things that only a few years ago were not possible. Some of the old theories endured because they could not be proven or disproven and they were often all we had. But now theories have to stand up to real testing and we can be guided by real data to ensure we don't get lost down the wrong path. Don't be discouraged by how "quiet" we can get. I have posted many odd questions and statements to the Tesla list over the years and gotten no response. I have found that is a good sign. When you are wrong, everyone senses blood and they are all over you! :-)) > >I also believe that if someone presents an idea, the onus is >on them to be able to fully explain it properly, so if I've >failed to do that anywhere, please do point it out. "I" would like to see the capacitance per unit length graph and it's bathtube curve. Can you do more than 32 points? I worry that the top and bottom numbers may get "averaged" excessivly from their true spikyness. > >Some recent discussions turned to the subject of primary >operation spark gaps, and arc/corona loading. > >Boris wrote: > >> Hmm..Seems we are digressing bit of concentration to >> bare secondary behavior (without arcs).OK,I can live >> with that. > >I can live with that too. Although this list was set up to >carry discussions relating to a particular software project, >I've no desire at all to enforce that. Personally I'm quite >happy to see posts on other technical issues. Recent >discussions on primary gaps and secondary arc loading have >certainly sparked my interest. > >In terms of scope, I see the following topics as being >particularly relevant to tesla coiling state-of-the-art: > >* Primary gap characteristics and their bearing on efficiency. >* Primary/secondary coupling, optimisation for minimum transfer > time. >* Dependency of secondary V/I distributions on the system > geometry. >* Optimisation of geometry to achieve the most favourable > V distribution for a given stored energy. >* Effects of corona and arc loading on secondary operation. > >Currently only the 3rd and maybe the 4th item are within the scope >of the tsim/tlap modeling program but I don't see any need to >constrain the project to those. On the contrary, I think we'd >all agree that the secondary can only be understood properly in >context, so I propose that the list of topics above be taken >as a guide to scope. I think we have between us the ingredients >to make some real progress on these topics, and I for one am >feeling quite excited about the prospects. If you need something to dream about for a future project. Think about adding the streamer modeling with it's 3D and time dimensions added! You may be able to model the streamer a little by having it go right up the centerline in a 2-D finite array (that does work). But having a full secondary with 3-D streamers propagating over time is a worthy but difficult dream to fulfill. Getting those streamers to fork and adding all that randomness would be really neat! the dynamic streamer impedance messing with the secondary currents is really fun! A more near term idea is to figure out what causes those darn secondary racing arcs! Cheers, Terry > >That's all for now, > >Best regards to all, >-- >Paul Nicholson, >Manchester, UK. >--
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.