Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 11 May 2001 19:53:19 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: [TSSP] Time domain modeling Reply-To: tssp@abelian.demon.co.uk Hi All, Another weekend looms, and I'll resist the temptation of the Corby Teslathon in order to get a bit more done on the simulator software. Malcolm, you were right when you wrote: > at a glance, k looks to be pretty low for that system. ? when referring to that Thor waveform. Found a coding error that messed up the pri-sec coupling. The modes have the right phase too, now, so we get a waveform that looks more like a real coil - I'll put up a gif file later in the weekend - the code is in pieces right now. It's amazing how many different ways there are to setup the solenoid operator [A], I'm on my third attempt now - they all work but I'm trying to keep program code in 1:1 correspondence with the math in pn1401 so that it can be validated by inspection, and its a case of finding a reasonable compromise between pn1401 clarity and program speed. For some of the routines I've ended up with two versions, a slow one which is inspectably correct, and a fast one which is incomprehensible. A command line option toggles between the two, so that the outputs can be checked for agreement. Anyway, yesterday I found an intelligible construction for F and K which is O(N^3) and today I made it work in the code, so there's light at the end of the tunnel. Finding the complex frequencies of the normal modes turns out to be quite reliable - even when the system is heavily loaded at the top. It starts to struggle at very low k-factors since it is trying to distinguish between two very closely spaced modes, but the k values in question are much lower than normally used on TCs. Other than that, the det(A-1) appears very well behaved and the usual gradient descent methods work fine. Has an interesting landscape - I'll include some maps in pn1401. The technique of inverse iteration, as a means of finding the normal mode spatial distributions, works beautifully on the [A] operator, very stable, and converges in just 1 or 2 iterations, providing gamma is computed accurately enough - it wont let me get away with being sloppy with the determinant. The method involves picking a random vector and passing it 'backwards' through the operator, and what emerges is a vector that's a lot less like a random vector and more like the eigenvector. Take that vector and send it round again, and so on. Potential trouble here with closely spaced modes - the iteration can converge to the wrong mode, so you have to detect this and start again with a different random vector. I'm having to model at around 200 elements, in order to compute det(A-1) in reasonable time, say less than 1 second. The matrix triangulation is O(N^3). The summation loops to form the solenoid operator A are also O(N^3), but F and K can be computed just once, apart from a single factor involving the series resistance Rs, which is frequency dependant. The description of the coil in terms of a Fredholm integral is turning out to be quite fruitful, mathematically speaking, meaning that from time to time, nice things pop out of the equations - things that make you want to stop and get a beer. For example, the solenoid operator [A] can be decomposed into the product of two other operators, [A] = w^2 [L][C] where [C] is a function of the Cint, Cext, Ctor, Ctop, and [L] is a function of the secondary mutual inductance profile. Thus we end up with a characteristic equation of det(w^2 [L][C] - [1]) = 0 where [1] is the identity operator. Thus in going from the lumped case to the distributed case, we replace the lumped reactances with their corresponding operators - operators which are differential in the time domain, and integral in the spatial domain. All in all, quite a simple and pleasant system to work with. As regards load impedance, we must decide between two ways to model this, and the decision must be left to nature. The question is this: Can we get away with a time-averaged load impedance which is a function of the top voltage averaged over a cycle or more, or must we come to terms with an impedance which is changing significantly within a cycle. Repeat this question with 'load impedance' replaced with 'gap resistance' and 'top voltage' replaced by 'primary current'. Both these need answering by experiment. The matter boils down to whether or not the top voltage and primary current are sine waves, or are the peaks and troughs flattened or peaky? Scope traces of top voltage waveform at various CW power levels within the brush discharge regime, plus a control waveform at a level below breakout. The waves can be LP filtered at say 20*Fres and we can FT the trace data to determine the harmonic distortion. If low enough, we can use a time averaged load impedance, rather than an instantaneous load conductance, which represents a big reduction of CPU effort in the time domain modeling. Anyway, enough waffle, Cheers for now, -- Paul Nicholson, Manchester, UK. -- Message-Id: <4.1.20010514220431.017f9a38@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Measured vs. Operating coil coupling? Cc: Bcc: X-Attachments: In-Reply-To: <3AFC351F.EA7A70C7@abelian.demon.co.uk> References: Message-Id: <4.1.20010514220431.017f9a38@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi All, I am wondering if the measured (at say 60Hz AC) primary to secondary coupling is the same as the coupling an operating coil sees? What makes me wonder if they are different is the current profile along the secondary at 60Hz AC may be different than the operating profile such as example 8 at: http://www.abelian.demon.co.uk/tssp/pn1710/ Perhaps the 60Hz AC and operating secondary current profiles are very close. However, if there is any difference, it seems like the coupling would also be affected buy the different secondary current profiles in the two cases. Perhaps there is a resonant or dynamic coupling coefficient that is different from the low frequency steady state coupling? I imagine they would not be too different at any rate, but the effect may be real... The effect of the top toroid as a shorted turn came up on the Tesla list and I copied it below for reference... Cheers, Terry >From: "Terry Fritz" >To: tesla@pupman.com >Subject: Re: Toroid Eddy Current? >Hi Ray, > >I think there are three factors at work that diminish the shorted turn >effect for toroids. > >1. The coupling is not real high. I used MandKV31 (coupling calculation >program*) to find the coupling for a 22 inch C-C toroid with an 8 inch cord >that was 10 inches above a 30 inch long 10.25 Dia. secondary with 1000 >turns. The coupling was only 0.028. > >* >http://hot-streamer.com/TeslaCoils/Programs/Programs.htm > >2. Toroids are not real good conductors. Corregated dryer pipe and other >aluminum materials are faily resistive to RF currents. Not sure how much >difference that makes but... > >3. There is not much current near the top of the coil. As Paul's TSSP >project and my tests have shown.** Thus, there is not a lot of nearby >current to couple to the toroid. > >** >http://hot-streamer.com/TeslaCoils/MyPapers/NSVPI/NVSPI.htm >http://www.abelian.demon.co.uk/tssp/pn1710/ > >I think all three of these have some affect and add up to only a small Eddy >current effect from the toroid. The actual numbers and all are a bit fuzzy >but it could probably all be figured out with a fair amount of precision. > >Cheers, > > Terry > > >At 09:58 PM 5/13/2001 -0700, you wrote: >>Doesn't the toroid form a single shorted turn that would dampen the >>secondary by lowering the Q? It seems that the toroid is close enough to the >>secondary to have an effect. An interesting test could be performed if >>someone had a working coil with an identical spare toroid. That spare toroid >>could be cut with a band saw radially toward the center stopping couple of >>inches from the middle. Probably 3 or 4 (maybe more) equally spaced cuts >>would be adequate and maybe dabs of epoxy to stabilize the cuts. Then >>install the modified toroid, retune the coil and compare the results with >>the original toroid. >> >>Maybe I'm off base but I have never seen this discussed and if the Q of the >>material for the secondary coil form matters, seems this just might. >> >>Ray >> Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 14 May 2001 22:20:21 -0600 From: "Terrell W. Fritz" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Sender: terrellf@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: [TSSP] Measured vs. Operating coil coupling? Reply-To: tssp@abelian.demon.co.uk Hi All, I am wondering if the measured (at say 60Hz AC) primary to secondary coupling is the same as the coupling an operating coil sees? What makes me wonder if they are different is the current profile along the secondary at 60Hz AC may be different than the operating profile such as example 8 at: http://www.abelian.demon.co.uk/tssp/pn1710/ Perhaps the 60Hz AC and operating secondary current profiles are very close. However, if there is any difference, it seems like the coupling would also be affected buy the different secondary current profiles in the two cases. Perhaps there is a resonant or dynamic coupling coefficient that is different from the low frequency steady state coupling? I imagine they would not be too different at any rate, but the effect may be real... The effect of the top toroid as a shorted turn came up on the Tesla list and I copied it below for reference... Cheers, Terry >From: "Terry Fritz" >To: tesla@pupman.com >Subject: Re: Toroid Eddy Current? >Hi Ray, > >I think there are three factors at work that diminish the shorted turn >effect for toroids. > >1. The coupling is not real high. I used MandKV31 (coupling calculation >program*) to find the coupling for a 22 inch C-C toroid with an 8 inch cord >that was 10 inches above a 30 inch long 10.25 Dia. secondary with 1000 >turns. The coupling was only 0.028. > >* >http://hot-streamer.com/TeslaCoils/Programs/Programs.htm > >2. Toroids are not real good conductors. Corregated dryer pipe and other >aluminum materials are faily resistive to RF currents. Not sure how much >difference that makes but... > >3. There is not much current near the top of the coil. As Paul's TSSP >project and my tests have shown.** Thus, there is not a lot of nearby >current to couple to the toroid. > >** >http://hot-streamer.com/TeslaCoils/MyPapers/NSVPI/NVSPI.htm >http://www.abelian.demon.co.uk/tssp/pn1710/ > >I think all three of these have some affect and add up to only a small Eddy >current effect from the toroid. The actual numbers and all are a bit fuzzy >but it could probably all be figured out with a fair amount of precision. > >Cheers, > > Terry > > >At 09:58 PM 5/13/2001 -0700, you wrote: >>Doesn't the toroid form a single shorted turn that would dampen the >>secondary by lowering the Q? It seems that the toroid is close enough to the >>secondary to have an effect. An interesting test could be performed if >>someone had a working coil with an identical spare toroid. That spare toroid >>could be cut with a band saw radially toward the center stopping couple of >>inches from the middle. Probably 3 or 4 (maybe more) equally spaced cuts >>would be adequate and maybe dabs of epoxy to stabilize the cuts. Then >>install the modified toroid, retune the coil and compare the results with >>the original toroid. >> >>Maybe I'm off base but I have never seen this discussed and if the Q of the >>material for the secondary coil form matters, seems this just might. >> >>Ray >> Delivered-To: terrellf@mail-dnvr.uswest.net Date: Tue, 15 May 2001 17:54:29 +1200 From: "Malcolm Watts" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Organization: Massey University at Wellington, NZ Subject: Re: [TSSP] Measured vs. Operating coil coupling? X-mailer: Pegasus Mail for Win32 (v3.12) Reply-To: tssp@abelian.demon.co.uk Hi Terry, On 14 May 01, at 22:20, Terrell W. Fritz wrote: > Hi All, > > I am wondering if the measured (at say 60Hz AC) primary to secondary > coupling is the same as the coupling an operating coil sees? What > makes me wonder if they are different is the current profile along the > secondary at 60Hz AC may be different than the operating profile such > as example 8 at: > > http://www.abelian.demon.co.uk/tssp/pn1710/ > > Perhaps the 60Hz AC and operating secondary current profiles are very > close. However, if there is any difference, it seems like the > coupling would also be affected buy the different secondary current > profiles in the two cases. Perhaps there is a resonant or dynamic > coupling coefficient that is different from the low frequency steady > state coupling? I imagine they would not be too different at any > rate, but the effect may be real... According to my measurements there is little difference. That's not to say that there isn't a measureable difference - just that if there is it appears to be small. I normally measure k off the beat envelope or by doing the spectral sweep in the coupled system with a high Z source across the tank circuit. > The effect of the top toroid as a shorted turn came up on the Tesla > list and I copied it below for reference... That topic has cropped up a few times in the past. I note that you stated the simulation assumed that the terminal was 10" above the windings. The k you found was entirely reasonable for that degree of separation but is much too low for many coils whose terminals are mounted just a couple of inches above the windings. It should definitely be subject ot modelling in my opinion. Also, the modelling should investigate the profiles when it is a shorted turn and when it is slitted. Curious that the frequency differences are not greater in practice than one would think. Something else for Paul to choke over. Regards, malcolm > Cheers, > > Terry > > > >From: "Terry Fritz" > >To: tesla@pupman.com > >Subject: Re: Toroid Eddy Current? > > >Hi Ray, > > > >I think there are three factors at work that diminish the shorted > >turn effect for toroids. > > > >1. The coupling is not real high. I used MandKV31 (coupling > >calculation program*) to find the coupling for a 22 inch C-C toroid > >with an 8 inch cord that was 10 inches above a 30 inch long 10.25 > >Dia. secondary with 1000 turns. The coupling was only 0.028. > > > >* > >http://hot-streamer.com/TeslaCoils/Programs/Programs.htm > > > >2. Toroids are not real good conductors. Corregated dryer pipe and > >other aluminum materials are faily resistive to RF currents. Not > >sure how much difference that makes but... > > > >3. There is not much current near the top of the coil. As Paul's > >TSSP project and my tests have shown.** Thus, there is not a lot of > >nearby current to couple to the toroid. > > > >** > >http://hot-streamer.com/TeslaCoils/MyPapers/NSVPI/NVSPI.htm > >http://www.abelian.demon.co.uk/tssp/pn1710/ > > > >I think all three of these have some affect and add up to only a > >small Eddy current effect from the toroid. The actual numbers and > >all are a bit fuzzy but it could probably all be figured out with a > >fair amount of precision. > > > >Cheers, > > > > Terry > > > > > >At 09:58 PM 5/13/2001 -0700, you wrote: > >>Doesn't the toroid form a single shorted turn that would dampen the > >>secondary by lowering the Q? It seems that the toroid is close > >>enough to the secondary to have an effect. An interesting test could > >>be performed if someone had a working coil with an identical spare > >>toroid. That spare toroid could be cut with a band saw radially > >>toward the center stopping couple of inches from the middle. > >>Probably 3 or 4 (maybe more) equally spaced cuts would be adequate > >>and maybe dabs of epoxy to stabilize the cuts. Then install the > >>modified toroid, retune the coil and compare the results with the > >>original toroid. > >> > >>Maybe I'm off base but I have never seen this discussed and if the Q > >>of the material for the secondary coil form matters, seems this just > >>might. > >> > >>Ray > >> > > Delivered-To: terrellf@mail-dnvr.uswest.net Date: Tue, 15 May 2001 11:46:57 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Measured vs. Operating coil coupling? Reply-To: tssp@abelian.demon.co.uk Terry, I agree with your comments to Ray, but I'll add a few of my own... 4. With a large toroid, the top current can be a large proportion of the base current, eg 70%, so that there can be a sizeable current available for induction at the top. 5. The induced currents in a shorted turn, toroid, or ground sheet are not in themselves a bad thing - in fact they can be quite useful if carefully managed. Its the I^2R loss in the shorted loop that really counts, and whether or not the shorted loop destroys the Q of the coil depends on the impedance match between the loop resistance and the rest of the system. For any shorted loop there will be a worst possible loop resistance that will have, for the particular k factor involved, the max impact on system Q, and that worst resistance will be the matched resistance. If the loop R is below or above this, the impact on Q will be reduced accordingly. Thus a shorted loop which has a very low R will only have a tiny effect on Q, even though the induced current can be quite high, as seen by the apparent reduction of the secondary inductance. To demonstrate this qualitatively I've generated a set of response curves for a resonating solenoid coupled to a shorted turn: http://www.abelian.demon.co.uk/tssp/tmp/eddy-coupling.gif The green curve is the worst-case scenario - the resonator is well matched to the shorted loop resistance of 1K ohms and the Q is very low. The blue and red traces show the response with a loop resistance of 100K ohms and 10 ohms respectively, and you can see that, when we are well away from the 'matched' resistance, the effect of the shorted turn on the Q is much reduced. So the moral is - if you're going to have a shorted turn, make it a good one, eg the nice wide conductor of the toroid. There is an advantage to this ... 6. The regime in which shorted loop R is well below the 'matched' R is commonly used for screening in RF circuits, eg the screening cans around an RF coil, or the silver plated inner surface of a cavity resonator. The beauty of this is that the big currents stirred up in the screen emit their own mag field, of a polarity which tends to cancel out the total mag field outside of the screen, in other words, although the screen material is non-magnetic, the eddy currents that it carries do in effect act as a magnetic shield, preventing the mag field from extending beyond the screen. 7. I'm currently exploring the controlled use of eddy currents in a ground sheet - you may recall that I layed out some 30m^2 of sheet Ali under my coil, in a pattern which avoided induction of circulating currents, in an attempt to raise the Q by more efficiently catching the external displacement currents of the coil. Well the rise of Q was small - certainly not the vast improvement I hoped for. I have a coil with an AC resistance of around 40 ohms, which forms a resonator with an effective loss of around 150 ohms. I can account for another 50 of those, so there are still around 50 ohms unaccounted for in the loss budget. I'm assuming now that this missing loss is due to eddy currents stirred up in the soil beneath the coil. My next experiment will be to lay out the ground sheet as a continuous conductor, to see if the 'screening can' effect will noticeably improve the system Q by cancelling out the mag field extending down into the ground. > the current profile along the secondary at 60Hz AC may be different > than the operating profile Yes, I've prepared two graphs showing the V/I profiles of Marco's Thor system when driven by a CW voltage applied to the (untuned) primary terminals, as modeled by tssp: http://www.abelian.demon.co.uk/tssp/tmp/tviplot.thor-p.60.gif http://www.abelian.demon.co.uk/tssp/tmp/tviplot.thor-p.66030.gif for drive at 60Hz and at 66.03kHz respectively. Clearly we can expect the k factor at the operating frequency to differ from that measured with low frequency currents, although of course the mutual inductance is the same in both cases since that is frequency-independent and is fixed by the geometry of the coils. As you know, tssp's time domain simulator is currently under test, so hopefully over the next few months we can turn up some reliable statements about how the effective operating k relates to the mutual inductance M. Note that the familiar formula k=M/sqrt(LpLs) only applies to coupled resonant circuits where the two separate coils are tuned to the same frequency, and when the system is driven near to that frequency. Away from resonance, or with mis-matched resonators, the k is a more complicated expression, so we are on shaky ground if we try to talk about k at 60Hz - better to stick with M. On the whole, I'd expect a modest Fres shift - say of order 1% down, on making a radial cut through the toroid, but no noticeable difference in performance. I believe this has been reported by those who have taken the trouble to test this. Just to add one final piece of speculation, there is some possibility that the shorted turn of the toroid may be of some slight benefit, in that it's resistance may dampen the higher resonant modes rather more than the two desired operational modes either side of Fres, thus perhaps usefully suppressing the unpleasant consequences of higher mode excitation, eg racing arcs and difficult quenching. Plenty of research to be done on this topic. By all means forward these comments to the pupman list if you wish. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Tue, 15 May 2001 16:46:16 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: [TSSP] Multimode CW excitation Reply-To: tssp@abelian.demon.co.uk Hi All, Just an off-the-cuff idea for your amusement, Say, P watts of CW base fed into a secondary gives a peak top volts V. Half that power, P/2 watts, would then give V/sqrt(2) Suppose you then setup a typical primary winding, resonated with the usual Cpri, but with the gap shorted. Modest k factor. Now the single mode splits into two, nearby frequencies, call them f+ and f-. Drive the base of the coupled system with P/2 watts at each resonance f+ and f-, for a total drive of P watts. Now, with the assumption that we still get V/sqrt(2) volts in each mode at the coil top, this means that (f+ - f-) times each second, the two modes reinforce at the top to give 2V/sqrt(2) volts, the net result being a peak voltage for our P watts which is sqrt(2) times that which we had with single mode CW excitation. Then suppose we add a second 'primary' resonator... Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Wed, 16 May 2001 08:50:06 +1200 From: "Malcolm Watts" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Organization: Massey University at Wellington, NZ Subject: Re: [TSSP] Multimode CW excitation X-mailer: Pegasus Mail for Win32 (v3.12) Reply-To: tssp@abelian.demon.co.uk Hi Paul, The trick of splitting primary energy between two resonators to obtain SQRT(2)V difference between them can be simply done by using a second "passive" resonator as a ground mirror. I once built a mini system to do it. On 15 May 01, at 16:46, Paul wrote: > Hi All, > > Just an off-the-cuff idea for your amusement, > > Say, P watts of CW base fed into a secondary gives a peak > top volts V. > > Half that power, P/2 watts, would then give V/sqrt(2) > > Suppose you then setup a typical primary winding, resonated > with the usual Cpri, but with the gap shorted. Modest k factor. > Now the single mode splits into two, nearby frequencies, > call them f+ and f-. That assumes that there is an amplitude change (of course) and in the normal course of events for that to occur, there is energy exchange between the two resonant circuits at a rate dictated by k (I know you know all this - just thinking aloud). But it is the effective amplitude modulation of either circuit which when examined spectrally, shows f+ and f- to be present (with no Fr). But in the time domain, the resonator is still oscillating at Fr, albeit with a sinusoidally varying amplitude. If I were to monitor an AM signal from any source (or DSBSC to be more accurate), I would see the same things in the frequency domain and the time domain. Anyway...... > Drive the base of the coupled system with P/2 watts at each > resonance f+ and f-, for a total drive of P watts. > > Now, with the assumption that we still get V/sqrt(2) volts > in each mode at the coil top, this means that (f+ - f-) times > each second, the two modes reinforce at the top to give > 2V/sqrt(2) volts, the net result being a peak voltage > for our P watts which is sqrt(2) times that which we had with > single mode CW excitation. The two modes do not always reinforce though. they go through a cycle which causes a peak coincidence at one point, and a cancellation at another on a regular basis. Hence the amplitude modulation. Am I reading your suggestion correctly? > Then suppose we add a second 'primary' resonator... > > Cheers, > -- > Paul Nicholson, > Manchester, UK. Regards, Malcolm Message-Id: <4.1.20010515174050.017fdee0@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Re: [TSSP] Measured vs. Operating coil coupling? Cc: tesla@pupman.com Bcc: X-Attachments: In-Reply-To: <3B010921.33D35D5@abelian.demon.co.uk> References: <4.1.20010514220431.017f9a38@pop.dnvr.qwest.net> Message-Id: <4.1.20010515174050.017fdee0@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Paul, At 11:46 AM 5/15/2001 +0100, you wrote: >Terry, > >I agree with your comments to Ray, but I'll add a few of my own... > >4. With a large toroid, the top current can be a large proportion of >the base current, eg 70%, so that there can be a sizeable current >available for induction at the top. So we could see say 10 amps in the toroid. Interesting since we do not often think of the toroid as a current loop. However, perhaps we should!! > >5. The induced currents in a shorted turn, toroid, or ground sheet >are not in themselves a bad thing - in fact they can be quite useful >if carefully managed. Its the I^2R loss in the shorted loop that >really counts, and whether or not the shorted loop destroys the Q of >the coil depends on the impedance match between the loop resistance >and the rest of the system. For any shorted loop there will be a >worst possible loop resistance that will have, for the particular k >factor involved, the max impact on system Q, and that worst resistance >will be the matched resistance. If the loop R is below or above this, >the impact on Q will be reduced accordingly. Thus a shorted loop which >has a very low R will only have a tiny effect on Q, even though the >induced current can be quite high, as seen by the apparent reduction >of the secondary inductance. To demonstrate this qualitatively I've >generated a set of response curves for a resonating solenoid coupled >to a shorted turn: > > http://www.abelian.demon.co.uk/tssp/tmp/eddy-coupling.gif > >The green curve is the worst-case scenario - the resonator is well >matched to the shorted loop resistance of 1K ohms and the Q is very >low. The blue and red traces show the response with a loop resistance >of 100K ohms and 10 ohms respectively, and you can see that, when we >are well away from the 'matched' resistance, the effect of the >shorted turn on the Q is much reduced. > >So the moral is - if you're going to have a shorted turn, make it a >good one, eg the nice wide conductor of the toroid. There is an >advantage to this ... What a fascinating set of curves! It is not a big deal to add an outer aluminium tape ring to a toroid made from dryer duct to vastly decrease the resistance of the loop. Dryer duct toroid tend to have strangely poorer performance than smooth toroid. Perhaps this is a truly new concern for toroid builders! I also note that 5% frequency shift!! E-Tesla does not consider such things affecting resonant frequencies... I feel some experiments coming on :-)) snip... > >Yes, I've prepared two graphs showing the V/I profiles of Marco's >Thor system when driven by a CW voltage applied to the (untuned) >primary terminals, as modeled by tssp: > > http://www.abelian.demon.co.uk/tssp/tmp/tviplot.thor-p.60.gif > http://www.abelian.demon.co.uk/tssp/tmp/tviplot.thor-p.66030.gif > >for drive at 60Hz and at 66.03kHz respectively. Clearly we can expect >the k factor at the operating frequency to differ from that measured >with low frequency currents, although of course the mutual inductance >is the same in both cases since that is frequency-independent and is >fixed by the geometry of the coils. > >As you know, tssp's time domain simulator is currently under test, so >hopefully over the next few months we can turn up some reliable >statements about how the effective operating k relates to the mutual >inductance M. Note that the familiar formula k=M/sqrt(LpLs) only >applies to coupled resonant circuits where the two separate coils are >tuned to the same frequency, and when the system is driven near to >that frequency. Away from resonance, or with mis-matched resonators, >the k is a more complicated expression, so we are on shaky ground if >we try to talk about k at 60Hz - better to stick with M. > >On the whole, I'd expect a modest Fres shift - say of order 1% down, >on making a radial cut through the toroid, but no noticeable >difference in performance. I believe this has been reported by those >who have taken the trouble to test this. E-Tesla tends to have a little unaccounted for error in the sub 1% area. Perhaps this is the key! > >Just to add one final piece of speculation, there is some possibility >that the shorted turn of the toroid may be of some slight benefit, in >that it's resistance may dampen the higher resonant modes rather more >than the two desired operational modes either side of Fres, thus >perhaps usefully suppressing the unpleasant consequences of higher >mode excitation, eg racing arcs and difficult quenching. Plenty of >research to be done on this topic. Yep!!!! > >By all means forward these comments to the pupman list if you wish. I did! > >Cheers, >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Tue, 15 May 2001 17:54:41 -0600 From: "Terrell W. Fritz" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Cc: tesla@pupman.com X-Sender: terrellf@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: [TSSP] Measured vs. Operating coil coupling? Reply-To: tssp@abelian.demon.co.uk Hi Paul, At 11:46 AM 5/15/2001 +0100, you wrote: >Terry, > >I agree with your comments to Ray, but I'll add a few of my own... > >4. With a large toroid, the top current can be a large proportion of >the base current, eg 70%, so that there can be a sizeable current >available for induction at the top. So we could see say 10 amps in the toroid. Interesting since we do not often think of the toroid as a current loop. However, perhaps we should!! > >5. The induced currents in a shorted turn, toroid, or ground sheet >are not in themselves a bad thing - in fact they can be quite useful >if carefully managed. Its the I^2R loss in the shorted loop that >really counts, and whether or not the shorted loop destroys the Q of >the coil depends on the impedance match between the loop resistance >and the rest of the system. For any shorted loop there will be a >worst possible loop resistance that will have, for the particular k >factor involved, the max impact on system Q, and that worst resistance >will be the matched resistance. If the loop R is below or above this, >the impact on Q will be reduced accordingly. Thus a shorted loop which >has a very low R will only have a tiny effect on Q, even though the >induced current can be quite high, as seen by the apparent reduction >of the secondary inductance. To demonstrate this qualitatively I've >generated a set of response curves for a resonating solenoid coupled >to a shorted turn: > > http://www.abelian.demon.co.uk/tssp/tmp/eddy-coupling.gif > >The green curve is the worst-case scenario - the resonator is well >matched to the shorted loop resistance of 1K ohms and the Q is very >low. The blue and red traces show the response with a loop resistance >of 100K ohms and 10 ohms respectively, and you can see that, when we >are well away from the 'matched' resistance, the effect of the >shorted turn on the Q is much reduced. > >So the moral is - if you're going to have a shorted turn, make it a >good one, eg the nice wide conductor of the toroid. There is an >advantage to this ... What a fascinating set of curves! It is not a big deal to add an outer aluminium tape ring to a toroid made from dryer duct to vastly decrease the resistance of the loop. Dryer duct toroid tend to have strangely poorer performance than smooth toroid. Perhaps this is a truly new concern for toroid builders! I also note that 5% frequency shift!! E-Tesla does not consider such things affecting resonant frequencies... I feel some experiments coming on :-)) snip... > >Yes, I've prepared two graphs showing the V/I profiles of Marco's >Thor system when driven by a CW voltage applied to the (untuned) >primary terminals, as modeled by tssp: > > http://www.abelian.demon.co.uk/tssp/tmp/tviplot.thor-p.60.gif > http://www.abelian.demon.co.uk/tssp/tmp/tviplot.thor-p.66030.gif > >for drive at 60Hz and at 66.03kHz respectively. Clearly we can expect >the k factor at the operating frequency to differ from that measured >with low frequency currents, although of course the mutual inductance >is the same in both cases since that is frequency-independent and is >fixed by the geometry of the coils. > >As you know, tssp's time domain simulator is currently under test, so >hopefully over the next few months we can turn up some reliable >statements about how the effective operating k relates to the mutual >inductance M. Note that the familiar formula k=M/sqrt(LpLs) only >applies to coupled resonant circuits where the two separate coils are >tuned to the same frequency, and when the system is driven near to >that frequency. Away from resonance, or with mis-matched resonators, >the k is a more complicated expression, so we are on shaky ground if >we try to talk about k at 60Hz - better to stick with M. > >On the whole, I'd expect a modest Fres shift - say of order 1% down, >on making a radial cut through the toroid, but no noticeable >difference in performance. I believe this has been reported by those >who have taken the trouble to test this. E-Tesla tends to have a little unaccounted for error in the sub 1% area. Perhaps this is the key! > >Just to add one final piece of speculation, there is some possibility >that the shorted turn of the toroid may be of some slight benefit, in >that it's resistance may dampen the higher resonant modes rather more >than the two desired operational modes either side of Fres, thus >perhaps usefully suppressing the unpleasant consequences of higher >mode excitation, eg racing arcs and difficult quenching. Plenty of >research to be done on this topic. Yep!!!! > >By all means forward these comments to the pupman list if you wish. I did! > >Cheers, >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Wed, 16 May 2001 11:03:14 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Multimode CW excitation Reply-To: tssp@abelian.demon.co.uk Malcolm Watts wrote: > The two modes do not always reinforce though. they go through a cycle > which causes a peak coincidence at one point, and a cancellation at > another on a regular basis. Hence the amplitude modulation. Am I > reading your suggestion correctly? Yes, I think so. I'm suggesting putting a two-tone (ie DSBSC) CW feed into the coil base, after having first made the coil resonate at both component frequencies by coupling in a freewheeling primary resonator, thus obtaining a factor sqrt(2) increase in peak volts, due to the power compression of the beat envelope. This in addition to the usual CW burst-mode power compression. > > Then suppose we add a second 'primary' resonator... by which I mean sharing the drive power over multiple frequencies, so that from time to time they will all come together in phase to create a substantial peak. Sort of a half-way house between the trad coil and the CW coil. Not really a serious suggestion unless a way can be found to generate the multi-tone signal which doesn't involve the driver also seeing a mega current pulse from time to time. While we're on the subject, the traditional use of a two-tone test signal is to measure distortion on comms links, eg RF linear amps, etc. Maybe a two-tone test can be devised to measure the amount of non-linearity caused by brush discharges. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Thu, 17 May 2001 18:25:39 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Measured vs. Operating coil coupling? Reply-To: tssp@abelian.demon.co.uk Hi Terry, All, Terrell W. Fritz wrote: > ...Dryer duct toroid tend to have strangely poorer performance > than smooth toroid. Perhaps this is a truly new concern for > toroid builders! It would be interesting to know where abouts the average toroid sits in relation to the curves in http://www.abelian.demon.co.uk/tssp/tmp/eddy-coupling.gif in other words, is a dryer-duct toroid more resistive than the 'worst case', ie somewhere between the green curve and the blue curve, or is it less resistive, ie between the green curve and the red curve? If the former, then reducing the resistance may make things worse. And how about a spun aluminium toroid - where does that sit? In view of the small Fres shift reported by those who have tried cutting the toroid, it may be nearer the blue than the red. Plenty of room for useful experiments with this. Eg, take a toroid with a radial cut. Measure Q and Fres with a range of resistances bridging the cut. Should be able to form a graph showing the locus of the response peaks (imagine a U-shaped curve joining the peaks of the three curves in the image, touching the peaks of all the intermediate response curves. Question boils down to whether or not the resulting curve shows the whole U, or maybe just the left hand half, ie with zero bridging resistance the response doesn't go as far down as the green curve, and never gets up the other side because the toroid resistance itself is too high. It should be possible to demonstrate a U-shaped locus of the coupling by doing the above expt using the primary, with a variable resistance across the primary terminals (no gap, no cap), and CW feed into the base. Most interesting for me is the possibility of preventing the magnetic field from going into the ground - but that test will have to wait till the weather improves. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 18 May 2001 10:53:11 +1200 From: "Malcolm Watts" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Organization: Massey University at Wellington, NZ Subject: Re: [TSSP] Measured vs. Operating coil coupling? X-mailer: Pegasus Mail for Win32 (v3.12) Reply-To: tssp@abelian.demon.co.uk Hi Paul, There's no doubt in my mind where a spun toriod lies (leaving oxide layers aside): On 17 May 01, at 18:25, Paul wrote: > Hi Terry, All, > > Terrell W. Fritz wrote: > > > ...Dryer duct toroid tend to have strangely poorer performance > > than smooth toroid. Perhaps this is a truly new concern for > > toroid builders! > > It would be interesting to know where abouts the average toroid > sits in relation to the curves in > > http://www.abelian.demon.co.uk/tssp/tmp/eddy-coupling.gif > > in other words, is a dryer-duct toroid more resistive than the > 'worst case', ie somewhere between the green curve and the > blue curve, or is it less resistive, ie between the green curve and > the red curve? If the former, then reducing the resistance may make > things worse. And how about a spun aluminium toroid - where does that > sit? In view of the small Fres shift reported by those who have tried > cutting the toroid, it may be nearer the blue than the red. My guess is nearer the red. A cut terminal would be on a par loss- wise with a good quality shorted turn wouldn't it? > Plenty of room for useful experiments with this. Eg, take a toroid > with a radial cut. Measure Q and Fres with a range of resistances > bridging the cut. Should be able to form a graph showing the locus of > the response peaks (imagine a U-shaped curve joining the peaks of the > three curves in the image, touching the peaks of all the intermediate > response curves. Question boils down to whether or not the resulting > curve shows the whole U, or maybe just the left hand half, ie with > zero bridging resistance the response doesn't go as far down as the > green curve, and never gets up the other side because the toroid > resistance itself is too high. > > It should be possible to demonstrate a U-shaped locus of the > coupling by doing the above expt using the primary, with a > variable resistance across the primary terminals (no gap, no cap), and > CW feed into the base. > > Most interesting for me is the possibility of preventing the > magnetic field from going into the ground - but that test will > have to wait till the weather improves. Well I ran my very first "proper" coil on a stainless steel benchtop about 20 years ago. While its performance was not stunning (it followed the wonderful recipe mentioned on the builder's list of a single heavy turn in the primary and b-all turns in the secondary :) it was a bit of a puzzle as to why it worked as well as it did until I realized that the conductive benchtop was basically reflecting energy rather than absorbing. the primary was positioned less than three inches above the bench surface. And just for the record, its relatively poor performance was unchanged when it was run on a wooden desktop instead. The benchtop was earthed and used as an earth for the coil. Regards, malcolm > Cheers, > -- > Paul Nicholson, > Manchester, UK. > -- > Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 18 May 2001 20:52:22 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: [TSSP] Time domain simulation Reply-To: tssp@abelian.demon.co.uk Hi All, Latest output is at http://www.abelian.demon.co.uk/tssp/tmp/thor-p.gif as you can see there is still a slight problem with the top voltage. Never ceases to amaze me that after umpteen pages of arithmetic, the first 2/3rds of which are in pn1401, we actualy get a recognisable waveform - rather than a random noise :) Won't be long before we're looking for some quantitative test comparisons. I've posted up an extensively revised pn1401, first 17 pages only. The math is nicer now. Download through http://www.abelian.demon.co.uk/tssp/pn1401.html in which the sections on in/out impedance have been moved to the end of a slightly revised pn2511, at http://www.abelian.demon.co.uk/tssp/pn2511.html Still a long way to go with the docs - we now have 14.5k lines of code and only 4.5k lines of documentation. -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 18 May 2001 16:36:46 -0400 From: "Metlicka Marc" Sender: owner-tssp@abelian.demon.co.uk To: "tssp@abelian.demon.co.uk" X-Mailer: Mozilla 4.7 [en] (Win98; U) X-Accept-Language: en Subject: [TSSP] leader formation Reply-To: tssp@abelian.demon.co.uk paul, all i was looking through some coil footage i had taken from the triggered gap on a small coil and discovered some very interesting "freeze frame" effects showing leader and streamer formation. the avi is a little large (1.5 meg) but can be viewed well with "real player" and using the slider one can follow frame by frame a streamer starting with an energy burst, or ion cloud, and then to leader and on to streamer. one can also follow the energy being drained from one stream by another and some very interesting effects as the safety gap fires when i go across the limits of the dimmer. i can send to any that wish to see it, it is well worth the look? i think i was lucky in capturing this at the right time and place, also, i didn't know at the time, but there was a pinhole in the toroid facing the camera that holds most of the events in the right area for viewing? just thought i would offer to share this with any that are interested in leader formation and streamer actions. marc m. Message-Id: <4.1.20010518161206.01800260@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Re: [TSSP] Time domain simulation Cc: Bcc: X-Attachments: In-Reply-To: <3B057D76.D3649F8A@abelian.demon.co.uk> References: Message-Id: <4.1.20010518161206.01800260@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Paul, At 08:52 PM 5/18/2001 +0100, you wrote: >Hi All, > >Latest output is at > > http://www.abelian.demon.co.uk/tssp/tmp/thor-p.gif > It looks great!! I note the higher order harmonic affects are now accounted for!! If you have the math tools ( :-)) you may want to run an FFT of the waveform. That would confirm that the "right stuff" is there. > >Won't be long before we're looking for some quantitative test >comparisons. :-))) snip... Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 18 May 2001 16:23:04 -0600 From: "Terrell W. Fritz" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Sender: terrellf@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: [TSSP] Time domain simulation Reply-To: tssp@abelian.demon.co.uk Hi Paul, At 08:52 PM 5/18/2001 +0100, you wrote: >Hi All, > >Latest output is at > > http://www.abelian.demon.co.uk/tssp/tmp/thor-p.gif > It looks great!! I note the higher order harmonic affects are now accounted for!! If you have the math tools ( :-)) you may want to run an FFT of the waveform. That would confirm that the "right stuff" is there. > >Won't be long before we're looking for some quantitative test >comparisons. :-))) snip... Cheers, Terry Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sat, 19 May 2001 18:20:51 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Time domain simulation Reply-To: tssp@abelian.demon.co.uk Terrell W. Fritz wrote: > If you have the math tools ( :-)) you may want to run an > FFT of the waveform. Already done, since we arrived here via the discrete spectrum of normal modes. I've assembled some bits and pieces in http://www.abelian.demon.co.uk/tssp/tmod.html which includes a mode spectrum. Will be interesting to see what happens to this spectrum when we crank up the k factor. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sat, 19 May 2001 21:29:05 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: [TSSP] Time domain simulation - top discharge Reply-To: tssp@abelian.demon.co.uk Hi All, Latest waveforms, http://www.abelian.demon.co.uk/tssp/tmp/thor-p.300kv.gif without breakout, and a tentative attempt at a discharge from the topload, occuring at 200kV in http://www.abelian.demon.co.uk/tssp/tmp/thor-p.200kv.gif In both cases the base current is not to be relied on. -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sat, 19 May 2001 22:16:47 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] leader formation Reply-To: tssp@abelian.demon.co.uk Hi Marc, I can handle AVI formats with the following video codecs, IBM Ultimotion (ULTI) depth 16. JPEG (JPEG) depth 24. Motion JPEG (MJPG) depth 24. Intergraph JPEG (IJPG) depth 24. Microsoft Video 1 (CRAM) depth 8 and 16. Radius Cinepak (CVID) depth 24. Intel Indeo 3.1 (IV31) depth 24. Intel Indeo 3.2 (IV32) depth 24. Intel Raw YUV (YUV9) depth 24. Creative CYUV (CYUV) depth 16. Uncompressed (RGB ) depth 4, 8, 16 or 24. Run length encoded (RLE8) depth 8. Editable MPEG (XMPG) depth 24. > the avi is a little large (1.5 meg) Perhaps you can edit out the significant bits? The file is a little large to send through email - perhaps I can download from your web site? Cheers, -- Paul Nicholson, Manchester, UK. -- Message-Id: <4.1.20010519155152.017f6e30@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Re: [TSSP] leader formation Cc: Bcc: X-Attachments: In-Reply-To: <3B06E2BF.D640C020@abelian.demon.co.uk> References: <3B0587DE.B08F9572@orwell.net> Message-Id: <4.1.20010519155152.017f6e30@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Paul, I have Marc's video posted at: http://hot-streamer.com/temp/triggered.avi My codec's only show a black image but mine may be old... BTW - your lastest plots are super neat!!! Cheers, Terry At 10:16 PM 5/19/2001 +0100, you wrote: >Hi Marc, > >I can handle AVI formats with the following video codecs, > > IBM Ultimotion (ULTI) depth 16. > JPEG (JPEG) depth 24. > Motion JPEG (MJPG) depth 24. > Intergraph JPEG (IJPG) depth 24. > Microsoft Video 1 (CRAM) depth 8 and 16. > Radius Cinepak (CVID) depth 24. > Intel Indeo 3.1 (IV31) depth 24. > Intel Indeo 3.2 (IV32) depth 24. > Intel Raw YUV (YUV9) depth 24. > Creative CYUV (CYUV) depth 16. > Uncompressed (RGB ) depth 4, 8, 16 or 24. > Run length encoded (RLE8) depth 8. > Editable MPEG (XMPG) depth 24. > >> the avi is a little large (1.5 meg) > >Perhaps you can edit out the significant bits? The file is a little >large to send through email - perhaps I can download from your web >site? > >Cheers, > >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sat, 19 May 2001 15:53:59 -0600 From: "Terrell W. Fritz" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Sender: terrellf@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: [TSSP] leader formation Reply-To: tssp@abelian.demon.co.uk Hi Paul, I have Marc's video posted at: http://hot-streamer.com/temp/triggered.avi My codec's only show a black image but mine may be old... BTW - your latest plots are super neat!!! Cheers, Terry At 10:16 PM 5/19/2001 +0100, you wrote: >Hi Marc, > >I can handle AVI formats with the following video codecs, > > IBM Ultimotion (ULTI) depth 16. > JPEG (JPEG) depth 24. > Motion JPEG (MJPG) depth 24. > Intergraph JPEG (IJPG) depth 24. > Microsoft Video 1 (CRAM) depth 8 and 16. > Radius Cinepak (CVID) depth 24. > Intel Indeo 3.1 (IV31) depth 24. > Intel Indeo 3.2 (IV32) depth 24. > Intel Raw YUV (YUV9) depth 24. > Creative CYUV (CYUV) depth 16. > Uncompressed (RGB ) depth 4, 8, 16 or 24. > Run length encoded (RLE8) depth 8. > Editable MPEG (XMPG) depth 24. > >> the avi is a little large (1.5 meg) > >Perhaps you can edit out the significant bits? The file is a little >large to send through email - perhaps I can download from your web >site? > >Cheers, > >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sat, 19 May 2001 23:32:56 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] leader formation Reply-To: tssp@abelian.demon.co.uk Terrell W. Fritz wrote: > I have Marc's video posted at: > > http://hot-streamer.com/temp/triggered.avi > > My codec's only show a black image but mine may be old... Thanks Terry, I get: /pan# file trig* triggered.avi: Microsoft RIFF, AVI data AVI Video Codec: Unknown VCR2(56435232) Sorry Marc, can't read this file. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sun, 20 May 2001 00:19:46 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk One of the difficulties ahead is that by adding in the primary resonator we now have quite a few extra dimensions to the resonator configuration space. The virtual secondary database is spanned by seven dimensions: h, b, h/d, awg, sr, td, and tb. We must add another five dimensions to accomodate flat spiral and conical primaries, and goodness knows how many are required to describe the gap and cap. As a means of zeroing in on an 'optimum system' I was thinking of trying the following method: Preparation: Start by generating say 1000, more or less randomly chosen designs. Selection: Compute the performance of all the coils and select the best say 100. Discard the remaining 900. Propagation: Take the design parameters of the successfull 100 and intermix them to produce another 900 offspring. Mutation: Randomly tweak by a small amount the parameters of some of the systems. Loop: Back to the selection step. Begins with a fairly high level of mutation, gradually reducing over say 100 generations. Would need about 100k coils to be computed, that's around 10 times the vsd effort, but surely considerably less than visiting at least 2 or 3 points on perhaps 12 or 15 axes. Not that any of the results would be particularly realistic, as I don't suppose we can take into account the breakout characteristics of the toploads involved. Is there any hope of being able to calculate a reasonable estimate of the voltage at which a particular topload will discharge? If not then any attempt at optimisation is just a piece of recreational mathematics. Food for thought. Comments welcome. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sat, 19 May 2001 20:40:45 -0400 From: "Metlicka Marc" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.7 [en] (Win98; U) X-Accept-Language: en Subject: Re: [TSSP] leader formation Reply-To: tssp@abelian.demon.co.uk i downloaded it and ran it through windows media player, this works very well and are able to set zoom size to get a good clarity. sorry for the trouble with the format? marc "Terrell W. Fritz" wrote: > > Hi Paul, > > I have Marc's video posted at: > > http://hot-streamer.com/temp/triggered.avi > > My codec's only show a black image but mine may be old... > > BTW - your latest plots are super neat!!! > > Cheers, > > Terry > > At 10:16 PM 5/19/2001 +0100, you wrote: > >Hi Marc, > > > >I can handle AVI formats with the following video codecs, > > > > IBM Ultimotion (ULTI) depth 16. > > JPEG (JPEG) depth 24. > > Motion JPEG (MJPG) depth 24. > > Intergraph JPEG (IJPG) depth 24. > > Microsoft Video 1 (CRAM) depth 8 and 16. > > Radius Cinepak (CVID) depth 24. > > Intel Indeo 3.1 (IV31) depth 24. > > Intel Indeo 3.2 (IV32) depth 24. > > Intel Raw YUV (YUV9) depth 24. > > Creative CYUV (CYUV) depth 16. > > Uncompressed (RGB ) depth 4, 8, 16 or 24. > > Run length encoded (RLE8) depth 8. > > Editable MPEG (XMPG) depth 24. > > > >> the avi is a little large (1.5 meg) > > > >Perhaps you can edit out the significant bits? The file is a little > >large to send through email - perhaps I can download from your web > >site? > > > >Cheers, > > > >-- > >Paul Nicholson, > >Manchester, UK. > >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sun, 20 May 2001 12:51:42 -0400 From: "Metlicka Marc" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.7 [en] (Win98; U) X-Accept-Language: en Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk paul, all Paul wrote: > Not that any of the results would be particularly realistic, as I > don't suppose we can take into account the breakout characteristics of > the toploads involved. Is there any hope of being able to calculate a > reasonable estimate of the voltage at which a particular topload will > discharge? i may be way off in my thinking, but could, using my triggered gap to force gap conduction, i use a variac to bring the voltage to an nst up slowly, use a one set sized toroid and find the input voltage where the toroid just breaks out. then use another toroid of close diameter but of a larger dryer duct, then bring the voltage up slowly until it just breaks out. then use a toroid of larger overall diameter and do the same. this may give some starting data and then maybe the roc's of each could be calculated into a form of database to find the mean breakout voltages of different roc's? since i could do this with the same coil profile and only the topload being changed, would this help in any way? i cover all of my dryer duct toroids i build with a layer of aluminum tape to try to smooth, and i have been having good luck coating the duct with drywall mud then sanding smooth, spraying with polyurethane, then aluminum tape for a smooth toroid almost as good as a spun toroid. i could do this with the three or five control topload in this testing? if you, and all on the tssp project, think this data could generate a basis of topload breakout voltage, i would be happy to try. marc If not then any attempt at optimisation is just a piece of > recreational mathematics. > > Food for thought. Comments welcome. > > Cheers, > > -- > Paul Nicholson, > Manchester, UK. > -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 09:16:53 +1200 From: "Malcolm Watts" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Organization: Massey University at Wellington, NZ Subject: [TSSP] Something of Interest X-mailer: Pegasus Mail for Win32 (v3.12) Reply-To: tssp@abelian.demon.co.uk Hi all, Being the type to get bored easily when there is nothing to do and not wishing to start a major series of expts (for which I'm currently ill-equipped laboratory-wise at the moment), I played around with a couple of resonators driven CW from a very low Z signal generator and scope. Briefly, two of the resonators were around 20" long, one with a diameter of 9" and the other about 4.5" (about 700 turns closewound), and the third resonator about 4" but wound with about 1700 turns of rather finer wire. I don't have the details with me at work so please forgive the lack of specifics. What I observed will give rise to a series of expts and msmts when I get back from the States. The generator is capable of 3VRMS with a source impedance of 0.5 Ohms resistive. Among other things, I noted a *drop* in Q when placing a terminal on resonator #3 (usually one sees a rise and I infer skin/proximity losses are most probably responsible due to the small wire size) but of all three, this was the best radiator with no terminal. I placed a 2' fluoro lamp on each resonator in turn and with this particular resonator, the lamp lit for 3 or 4 resonant modes (higher order). The lamp could be ignited by holding about 6" away from the tuned coil and was still running 3' or so away with some judicious retuning of the sig gen. There is some interesting data to be mined out of all this and I expect some interesting running results as well, both with and without terminals. I decided to flag this as I won't have an opportunity to do anything about it for several weeks and someone else on the list might. While conventional wisdom concentrates on high secondary Q's, there is some really interesting stuff to be gleaned from secondaries that pack a lot of L into a relatively small volume. Tx line effects abound. Regards, malcolm Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 09:51:27 +1200 From: "Malcolm Watts" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Organization: Massey University at Wellington, NZ Subject: [TSSP] Re: Something of Interest X-mailer: Pegasus Mail for Win32 (v3.12) Reply-To: tssp@abelian.demon.co.uk BTW, All resonators were run with an untuned primary of about 12uH coupling into them at a about k = 0.1. Direct connection of the sig gen to the coil base produced little of interest. Malcolm ------- Forwarded message follows ------- From: Malcolm Watts To: tssp@abelian.demon.co.uk Subject: Something of Interest Send reply to: M.J.Watts@massey.ac.nz Date sent: Mon, 21 May 2001 09:16:54 +1200 Hi all, Being the type to get bored easily when there is nothing to do and not wishing to start a major series of expts (for which I'm currently ill-equipped laboratory-wise at the moment), I played around with a couple of resonators driven CW from a very low Z signal generator and scope. Briefly, two of the resonators were around 20" long, one with a diameter of 9" and the other about 4.5" (about 700 turns closewound), and the third resonator about 4" but wound with about 1700 turns of rather finer wire. I don't have the details with me at work so please forgive the lack of specifics. What I observed will give rise to a series of expts and msmts when I get back from the States. The generator is capable of 3VRMS with a source impedance of 0.5 Ohms resistive. Among other things, I noted a *drop* in Q when placing a terminal on resonator #3 (usually one sees a rise and I infer skin/proximity losses are most probably responsible due to the small wire size) but of all three, this was the best radiator with no terminal. I placed a 2' fluoro lamp on each resonator in turn and with this particular resonator, the lamp lit for 3 or 4 resonant modes (higher order). The lamp could be ignited by holding about 6" away from the tuned coil and was still running 3' or so away with some judicious retuning of the sig gen. There is some interesting data to be mined out of all this and I expect some interesting running results as well, both with and without terminals. I decided to flag this as I won't have an opportunity to do anything about it for several weeks and someone else on the list might. While conventional wisdom concentrates on high secondary Q's, there is some really interesting stuff to be gleaned from secondaries that pack a lot of L into a relatively small volume. Tx line effects abound. Regards, malcolm ------- End of forwarded message ------- Message-Id: <4.1.20010520204358.01809240@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Re: [TSSP] Genetic optimisation Cc: Bcc: X-Attachments: In-Reply-To: <3B06FF92.84B5A34D@abelian.demon.co.uk> References: Message-Id: <4.1.20010520204358.01809240@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Paul, At 12:19 AM 5/20/2001 +0100, you wrote: snip... that all sounded goo to me! > >Not that any of the results would be particularly realistic, as I >don't suppose we can take into account the breakout characteristics of >the toploads involved. Is there any hope of being able to calculate a >reasonable estimate of the voltage at which a particular topload will >discharge? If not then any attempt at optimisation is just a piece of >recreational mathematics. If one inputs a top load voltage into E-Tesla6, it will create a field stress array in volts per unit distance. Thus, if you know the breakdown voltage of air, you can find the torroid voltage that will give that field stress. Voltage plot: http://hot-streamer.com/temp/MattD1.gif Voltage stress plot: http://hot-streamer.com/temp/MattD2.gif Basically, it seems to follow very closely the sphere's breakdown voltage of 3MV/m (3,000,000 volts per meter radius). Since the area of interest is very close to the top load's surface, the surrounding stuff does not have a great effect. However, that is assuming the breakdown voltage is the same at DC, 60Hz, and 350kHz. So you can find the voltage differentials per unit distance in the arrays of a field stress analysis and come up with the volts per meter stress in various places. The question then is when does air initially breakdown at say 300kHz? The fact that the voltage "suddenly" builds up rather than is a continuous stress is also a factor. "Pulsed" CW coils have very long swordlike streamers as opposed to pure CW coils due to this effect. Still thinking on this.... Cheers, Terry > >Food for thought. Comments welcome. > >Cheers, > >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Sun, 20 May 2001 20:57:23 -0600 From: "Terrell W. Fritz" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Sender: terrellf@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Hi Paul, At 12:19 AM 5/20/2001 +0100, you wrote: snip... that all sounded goo to me! > >Not that any of the results would be particularly realistic, as I >don't suppose we can take into account the breakout characteristics of >the toploads involved. Is there any hope of being able to calculate a >reasonable estimate of the voltage at which a particular topload will >discharge? If not then any attempt at optimisation is just a piece of >recreational mathematics. If one inputs a top load voltage into E-Tesla6, it will create a field stress array in volts per unit distance. Thus, if you know the breakdown voltage of air, you can find the torroid voltage that will give that field stress. Voltage plot: http://hot-streamer.com/temp/MattD1.gif Voltage stress plot: http://hot-streamer.com/temp/MattD2.gif Basically, it seems to follow very closely the sphere's breakdown voltage of 3MV/m (3,000,000 volts per meter radius). Since the area of interest is very close to the top load's surface, the surrounding stuff does not have a great effect. However, that is assuming the breakdown voltage is the same at DC, 60Hz, and 350kHz. So you can find the voltage differentials per unit distance in the arrays of a field stress analysis and come up with the volts per meter stress in various places. The question then is when does air initially breakdown at say 300kHz? The fact that the voltage "suddenly" builds up rather than is a continuous stress is also a factor. "Pulsed" CW coils have very long swordlike streamers as opposed to pure CW coils due to this effect. Still thinking on this.... Cheers, Terry > >Food for thought. Comments welcome. > >Cheers, > >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 09:48:19 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] leader formation Reply-To: tssp@abelian.demon.co.uk Metlicka Marc wrote: > i downloaded it and ran it through windows media player, > this works very well and are able to set zoom size to get a > good clarity. > sorry for the trouble with the format? Not your fault - I guess we are victims of Gates greed. Perhaps it can be saved in a different format - mpeg maybe? Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 10:14:13 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk > Paul wrote: > > Is there any hope of being able to calculate a > > reasonable estimate of the voltage at which a particular topload > > will discharge? > Metlicka Marc wrote: > ... could, using my triggered gap to force gap conduction, i use a > variac to bring the voltage to an nst up slowly, use a one set sized > toroid and find the input voltage where the toroid just breaks out... Yes, if there was some way to measure the top voltage, say to 5 or 10% accuracy. I'm sure there are approximate formulas based on radius of curvature, etc. Would be interesting to see how they compared with some measured breakout voltages. No doubt the amount of surface rough- ness will make a contribution. > this may give some starting data and then maybe the roc's of each > could be calculated into a form of database to find the mean breakout > voltages of different roc's? Yes, we'd end up with a table of roc and breakout volts, but maybe it would only be valid for a particular height above ground, etc. So many factors involved - needs some careful thought. > since i could do this with the same coil profile and only the topload > being changed, would this help in any way? Yes, I think it would, but really it would need several people to perform these measurements on their different coils in order to build up an overall picture. First step would be to settle on a way of measuring the top volts, so that all the results can be fairly compared and placed together into a composite table. > i could do this with the three or five control topload in this > testing? if you, and all on the tssp project, think this data could > generate a basis of topload breakout voltage, i would be happy to try. I think this would be a good thing to do. Maybe the following strategy? 1/ Devise a standardised way to measure the peak topvolts. Something that anybody can setup without buying any expensive kit. Must be repeatable and reasonably accurate, say 5% to 10%. 2/ Persuade lots of coilers to measure their systems with whatever toploads they have available. 3/ Compile the results into a table. At that point it should be clear whether we have a pattern to the table, or whether the results are all over the place. If there *is* a pattern, then we can extract the desired empirical formula. 4/ Use the formula to predict the breakout volts of a coil that hasn't been measured, and then test it. Marc, IMO this could well be an important and very worth while thing to do, not just for the tssp requirements, but for coilers in general. Maybe this study has already been done? What do the others on the list think? Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 10:32:51 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Something of Interest Reply-To: tssp@abelian.demon.co.uk Hi Malcolm, > Among other things, I noted a *drop* in Q when placing a terminal on > resonator #3 (usually one sees a rise and I infer skin/proximity losses > are most probably responsible due to the small wire size) Yes, usually Q goes up, I guess due to reduced AC resistance at the lower frequency of the loaded coil. But if the wire is so thin that it's radius is smaller than the skin depth at *both* frequencies, then the R will be the same for both, and wL/R will apply, hence the reduced Q of the loaded thin-wire coil. > but of all three, this was the best radiator with no terminal. Don't know what to make of that. > ... While conventional wisdom concentrates on high secondary Q's, > there is some really interesting stuff to be gleaned from secondaries > that pack a lot of L into a relatively small volume. Yes, high Q or high transimpedance - which to go for to get the best performance? Perhaps there is an optimum mix of the two? Not sure what we can conclude from the fluoro indications - as mentioned in another thread we could do with a neat way to measure top volts. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 10:34:55 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Re: Something of Interest Reply-To: tssp@abelian.demon.co.uk Malcolm Watts wrote: > BTW, All resonators were run with an untuned primary of about 12uH > coupling into them at a about k = 0.1. Direct connection of the sig > gen to the coil base produced little of interest. Zin at the primary terminals will be quite a lot lower than the base Zin, so I guess with the base drive the coil would not be taking much power from the source. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 11:10:15 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Terrell W. Fritz wrote: > ... > Voltage stress plot: > http://hot-streamer.com/temp/MattD2.gif Very nice. Could you produce a variation on E-Tesla which, when *given* a V-profile generates the gradient plot as above, and then explores it to find the spot with the highest gradient, which we take to be the weakest link - the 'Robinson spot' perhaps :). Then by scaling the gradient of the Robinson spot up to the breakdown value of air, say 10kV/cm, we get the maximum top volts before breakout. > ... However, that is assuming the breakdown voltage is > the same at DC, 60Hz, and 350kHz. Indeed. You would be calculating the *static* breakdown voltage with no account of the rate of change, but its a start. Comparison with the tabulated results of Marc's proposed project would be interesting, and might reveal the extent to which dV/dt is a factor. > ... The question then is when does air initially breakdown at > say 300kHz? Pity Jim Lux isn't on this list - I'm sure he could provide some essential background to this discussion. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 11:40:39 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Something of Interest Reply-To: tssp@abelian.demon.co.uk Malcolm wrote: > > but of all three, this was the best radiator with no terminal. I wrote: > Don't know what to make of that. After more thought, perhaps this coil resulted in the best impedance match between the primary terminals and the very low CW source impedance. Something to add to the every-growing list of things to be modeled. With TCs, it always seems to be impossible to change *one* thing at a time. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 07:44:13 -0500 From: "Bert Hickman" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mozilla-Status2: 00000000 Organization: Stoneridge Engineering X-Mozilla-Draft-Info: internal/draft; vcard=0; receipt=0; uuencode=0; html=0; linewidth=0 X-Mailer: Mozilla 4.75 [en] (Win95; U) X-Accept-Language: en Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Paul, Terry, and all, Avalanche breakdown is an essential "first step" in the sequence of corona, streamer, and leader formation. The E-field necessary to achieve avalanche breakdown in air at Tesla Coil frequencies (10e4 - 10e6) is apparently similar to that at DC - approximately 30 kV/cm. However, under RF excitation, the local E-field around the HV terminal can be significantly enhanced through the interaction of the terminal and any surrounding space charge regions. Even if the RF voltage peaks are initially insufficient to initiate avalanche breakout, they may generate waves of positive and negative ions around the output terminal. Once the surrounding space charge regions are created, they will increase the "effective" E-field seen at the HV terminal, thereby reducing the terminal voltage required to initiate breakdown. After a number of RF cycles, field enhancement becomes sufficient to initiate breakdown at a lower peak terminal voltage than for the DC case. The effect of space charges on RF breakdown is presented in Loeb ("Fundamental Processes of Electrical Discharge, McGraw-Hill/Chapman & Hall, 1939), reflecting the work of L. E. Reukema. Reukema used sphere gaps and relatively short gaps (0.25 - 2.5 cm). He found a progressive lowering of breakdown voltage, by as much as 17% (about 25 kV/cm), as the frequency was increased between 20 kHz and 60 kHz. However, no further breakdown voltage reductions were seen between 60 kHz and 425 kHz. Yuri P. Raizer also arrives at a similar conclusion but from a somewhat different slant ("Gas Discharge Physics, Springer-Verlag, Corrected Edition - 1997) - he estimates the breakdown field in room temperature for relatively long gaps (>6 cm) is about 26 kV/cm. So... barring precise measurements, a good starting point for estimating E-field for terminal breakout at Tesla Coil frequencies might be about 25 - 26 kV/cm. -- Bert -- -- Bert Hickman Stoneridge Engineering Email: bert.hickman@aquila.net Web Site: http://www.teslamania.com Paul wrote: > > Terrell W. Fritz wrote: > > > ... > > Voltage stress plot: > > http://hot-streamer.com/temp/MattD2.gif > > Very nice. Could you produce a variation on E-Tesla which, when > *given* a V-profile generates the gradient plot as above, and then > explores it to find the spot with the highest gradient, which we > take to be the weakest link - the 'Robinson spot' perhaps :). Then > by scaling the gradient of the Robinson spot up to the breakdown > value of air, say 10kV/cm, we get the maximum top volts before > breakout. > > > However, that is assuming the breakdown voltage is > > the same at DC, 60Hz, and 350kHz. > > Indeed. You would be calculating the *static* breakdown voltage > with no account of the rate of change, but its a start. Comparison > with the tabulated results of Marc's proposed project would be > interesting, and might reveal the extent to which dV/dt is a factor. > > > ... The question then is when does air initially breakdown at > > say 300kHz? > > Pity Jim Lux isn't on this list - I'm sure he could provide some > essential background to this discussion. > > Cheers, > -- > Paul Nicholson, > Manchester, UK. > -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Mon, 21 May 2001 17:18:12 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Hi Bert, Somehow you always seem to drop in an informative post. >... [Reukema] found a progressive lowering of breakdown voltage, > by as much as 17% (about 25 kV/cm), as the frequency was increased > between 20 kHz and 60 kHz. However, no further breakdown voltage > reductions were seen between 60 kHz and 425 kHz. That sounds promising. Does the 425 kHz limit imply things change above that frequency, or is that just where his measurements stopped? > ...[Raizer] estimates the breakdown field in room temperature for > relatively long gaps (>6 cm) is about 26 kV/cm. Are we OK to extrapolate this figure to gaps of order a metre or so? > So... barring precise measurements, a good starting point for > estimating E-field for terminal breakout at Tesla Coil frequencies > might be about 25 - 26 kV/cm. That sounds very reasonable. Perhaps Terry could use this value to calculate the breakdown voltage of a test setup and compare it with the peak top voltage measured in an actual run? If they came out in the same ballpark then we would have the basis of a two pronged strategy - Marc's gathering of experimental results, plus Terry's field calculations - which between them might be able to bracket the actual values to a range narrow enough to permit meaningful modeling. With a bit of luck we might expect a reasonable match for smooth toroids, with progressively lower measured breakdowns for rougher surfaces, at which point Terry will invent a fiddle factor to adjust for the toroid material. While we're on the subject of genetic methods, a few background links, http://www.rennard.org/alife/english/gavintrgb.html http://cmp.ameslab.gov/~jrmorris/ga/ http://ai.bpa.arizona.edu/~mramsey/ga.html http://www.hao.ucar.edu/public/research/si/pikaia/tutorial.html Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Tue, 22 May 2001 07:38:47 -0500 From: "Bert Hickman" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mozilla-Status2: 00000000 Organization: Stoneridge Engineering X-Mozilla-Draft-Info: internal/draft; vcard=0; receipt=0; uuencode=0; html=0; linewidth=0 X-Mailer: Mozilla 4.75 [en] (Win95; U) X-Accept-Language: en Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Paul, I only wish I had more time to devote towards supporting the excellent work you're doing! My responses are interspersed below... Paul wrote: > > Hi Bert, > > Somehow you always seem to drop in an informative post. > > >... [Reukema] found a progressive lowering of breakdown voltage, > > by as much as 17% (about 25 kV/cm), as the frequency was increased > > between 20 kHz and 60 kHz. However, no further breakdown voltage > > reductions were seen between 60 kHz and 425 kHz. > > That sounds promising. Does the 425 kHz limit imply things change > above that frequency, or is that just where his measurements stopped? Although Reukema limited his tests to 425 kHz, the space charge enhancement mechanism appears to reach a limit at about 60 kHz. Although experimentally verified to remain relatively constant at about 26 kV/cm from 60 kHz through 425 kHz, it probably applies to higher frequencies in the range applicable to Tesla Coils. Different mechanisms come into play when we extend this to microwave and optical breakdown in air. > > > ...[Raizer] estimates the breakdown field in room temperature for > > relatively long gaps (>6 cm) is about 26 kV/cm. > > Are we OK to extrapolate this figure to gaps of order a metre or > so? > Apparently so. Initiation of breakdown is a threshold avalanche process ultimately determined by the relation of electron creation versus their removal. In an electronegative gas such as air (predominated by the oxygen component), electrons are rapidly removed by the formation of negative ions. This removes them from further participation in avalanche multiplication. Per Raizer, avalanche multiplication in air at STP (the onset of breakdown) has a lower E-field limit of about 26 kV/cm once we are in the "long spark" regime. A long spark (>6 cm) is characterized by multiple avalanches in an evolutionary sequence: streamer flash(es) --> leader propagation (fed by groups of streamers) --> spark (if leader bridges the gap). If the local E-field never reaches the 26 kV/cm threshold, the initial electron avalanche that begins the process cannot occur irrespective of frequency or gap length. > > So... barring precise measurements, a good starting point for > > estimating E-field for terminal breakout at Tesla Coil frequencies > > might be about 25 - 26 kV/cm. Yes, assuming a smooth HV terminal... > > That sounds very reasonable. Perhaps Terry could use this value to > calculate the breakdown voltage of a test setup and compare it with > the peak top voltage measured in an actual run? If they came out in > the same ballpark then we would have the basis of a two pronged > strategy - Marc's gathering of experimental results, plus Terry's > field calculations - which between them might be able to bracket > the actual values to a range narrow enough to permit meaningful > modeling. > > With a bit of luck we might expect a reasonable match for smooth > toroids, with progressively lower measured breakdowns for rougher > surfaces, at which point Terry will invent a fiddle factor to > adjust for the toroid material. > > While we're on the subject of genetic methods, a few background > links, > > http://www.rennard.org/alife/english/gavintrgb.html > http://cmp.ameslab.gov/~jrmorris/ga/ > http://ai.bpa.arizona.edu/~mramsey/ga.html > http://www.hao.ucar.edu/public/research/si/pikaia/tutorial.html > > Cheers, > -- > Paul Nicholson, > Manchester, UK. > -- Best regards, -- Bert -- -- Bert Hickman Stoneridge Engineering Email: bert.hickman@aquila.net Web Site: http://www.teslamania.com Delivered-To: terrellf@mail-dnvr.uswest.net Date: Wed, 23 May 2001 17:45:45 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Hi Bert, All, Bert Hickman wrote: > I only wish I had more time to devote... I wish you had too - things always seem a lot clearer after you've put them into words. Looks like 26kV/cm is the figure to work with at normal TC frequencies, - comes as some relief that the value is not severely frequency dependent, things would be tricky if so! If I've understood things right, the leader formation begins and continues as long as this gradient can be maintained just ahead of the leader. I guess the significant threshold involves meeting this value at the surface of the smooth toroid and once a leader begins to form, its sharp point will then ensure that the leader forms rapidly for quite some distance - even though the 'background' field from the topload would, by itself, fall below the 26kV/cm threshold only a little way from the surface. Subject to the toroid having enough charge available to support that formation. What stops the leader formation? I guess either it hits earth or it runs out of charge - the toroid is depleted and the 26kV/cm cannot be maintained at the tip? So a big toroid would be reluctant to break out (modest surface gradient), but it would throw a long streamer as soon as it did (lots of charge available)? I'm afraid I've got some more questions! > A long spark (>6 cm) is characterized by multiple avalanches in an > evolutionary sequence: streamer flash(es) --> leader propagation > (fed by groups of streamers) --> spark (if leader > bridges the gap). Can we assume that this whole sequence takes place in a timescale short compared with an RF cycle - I suppose thats so because if not there would be a bigger frequency dependence? When the HT falls away, do things recombine and settle down sufficiently that on the next half cycle there is no 'memory' of the previous half cycle? Ultimately what I'm fishing around for is some confidence that some acceptable and realistic account can be taken of the breakout thresholds, otherwise attempts at non-linear time domain modeling will founder on that point. I feel as though we are on top of the technical matter of computing the response and now, quite suddenly it seems, we are up against this more difficult problem of finding a load conductance function which provides an acceptable summary description of the breakout dynamics. If I've got things right, then Terry should be able to calculate quite easily the top voltage at which streamers should suddenly start to form, and we might also be able to calculate an estimate of streamer length too (as a function of topvolts). Given those two separate figures (or functions) we would then have the choice of optimising for max topvolts or max streamer length, using the same genetic software but with two different merit functions. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Wed, 23 May 2001 17:58:59 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: [TSSP] Magnifiers, Modeling, and Macros Reply-To: tssp@abelian.demon.co.uk Hi All, A few bits and pieces... Does anyone want a section on 'magnifiers' in pn1401, or should that be deferred to a later document? Using the operators already defined in pn1401 as building blocks, the magnifier - and for that matter, arbitrary extra coils - can be 'connected in' to the analysis simply by stringing together the appropriate operators (as long as the extra coils are far enough away from one another that their E and B fields don't interact significantly). Eg for a standard magnifier, the cross-coupling between the second and third coils involves the top-drive operator of the 2nd coil and the base drive operator of the 3rd coil (both equ 2.16 in latest pn1401), along with the individual solenoid operators. You end up with an operator for the whole thing, which you then submit to the normal mode analysis in the usual way. It all seems to come out quite neatly, but I'm not particularly bothered about writing this up at the moment. The software organisation is quite different now, with most of the code in a common library module tlib.c, and the various simulation programs just call out routines from tlib as needed. Eg there are functions which return each of the operators (or their kernel functions) as matrices, and there are other library routines to add and multiply them, and to solve inverses and so on. Other routines locate the complex frequencies of the modes of a given operator, and another computes the eigenfunctions. All this means that from now on its fairly easy to throw together an ad-hoc program to suit a particular configuration or experiment. Hopefully this will encourage others to use the software in ways of their own devising, as its quite impossible for one person, given just a single lifetime, to explore the available permutations. BTW, the library and simulation programs should compile under windows, because I've avoided doing anything unix-specific in the code.The fancy multi-cpu stuff was taken out ages ago, Marco made tsim compile on his PC, and I don't think I've added anything in since which will have broken that port. The only thing that won't go to windows is the shed load of plotting and analysis scripts - unix shell scripts which apply standard tools to the data files produced by the simulation programs. I'm not sure what to do about these. Maybe if there is an 'excel' expert out there, we could team up to produce some macros which will take care of basic plotting and so on. Others might be harder, eg I'm just doing a script which reads the simulator time domain output file and creates an mpeg movie of the secondary voltage profile - I don't think you can combine diverse tools in this way under windows? Anyway, one way or another, we should try to find ways to make the tssp software more accessible, otherwise when I get bored with all this and go back to astronomical computing (I've a pulsar detection suite to finish off later this year), I suspect this code will just sit on a website gathering dust. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Wed, 23 May 2001 11:50:10 -0700 (PDT) From: "boris petkovic" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Hi Paul,All, > Looks like 26kV/cm is the figure to work with at > normal TC frequencies, > - comes as some relief that the value is not > severely frequency > dependent, things would be tricky if so! --- Surely,~26-28KV/cm critical surface field would be all right to assume for TC freqs (and standard atm.conditions) Definitely,if such field appears for a certain time ,it will get show to start happening per see. But,what counts the most here is one thing:TIME... By this I don't mean just time of TC waveform exceeding this treshold ,but also time pauses between 2 primary condensers bangs-->in other words some depedence on BPS (belive it or not!). John Freau experiments with smooth toroids are very valuable when we speak of this matter. He has noticed smaller bang size (innital cap energy) needed for visual spark breakout with higher BPS rate than with lower BPS rate. Few months ago,I contacted with M.Denicolai and he said he would make controled experiments on his thor in order to investigate such phenomenon this summer . Something else you should know as well.With about 400 KV peak voltage on his coil Marco was unable to get breakout of his 20*150 cm smooth toroid. Only with Ep~10 J he started to observe first faint streamers coming from the terminal. Question improved E-Tesla program should answer to: What max. toroid surface field corresponds to 400 KV voltage on THOR secondary? ---- > > If I've understood things right, the leader > formation begins and > continues as long as this gradient can be maintained > just ahead of the > leader. I guess the significant threshold involves > meeting this value at > the surface of the smooth toroid and once a leader > begins to form, its > sharp point will then ensure that the leader forms > rapidly for quite > some distance - even though the 'background' field > from the > topload would, by itself, fall below the 26kV/cm > threshold only a little > way from the surface. Subject to the toroid having > enough charge > available to support that formation. What stops the > leader formation? > I guess either it hits earth or it runs out of > charge - the toroid is > depleted and the 26kV/cm cannot be maintained at the > tip? So a big > toroid would be reluctant to break out (modest > surface gradient), but > it would throw a long streamer as soon as it did > (lots of charge > available)? > > I'm afraid I've got some more questions! > > > A long spark (>6 cm) is characterized by multiple > avalanches in an > > evolutionary sequence: streamer flash(es) --> > leader propagation > > (fed by groups of streamers) --> spark (if leader > > bridges the gap). > > Can we assume that this whole sequence takes place > in a timescale short > compared with an RF cycle - I suppose thats so > because if not there > would be a bigger frequency dependence? --- I think we can,althought some scopings on 250 Khz+ coils I saw in past ,displayed not so small differences. --- > > When the HT falls away, do things recombine and > settle down sufficiently > that on the next half cycle there is no 'memory' of > the previous half > cycle? > > Ultimately what I'm fishing around for is some > confidence that some > acceptable and realistic account can be taken of the > breakout > thresholds, otherwise attempts at non-linear time > domain modeling will > founder on that point. I feel as though we are on > top of the technical > matter of computing the response and now, quite > suddenly it seems, we > are up against this more difficult problem of > finding a load conductance > function which provides an acceptable summary > description of the > breakout dynamics. > > If I've got things right, then Terry should be able > to calculate quite > easily the top voltage at which streamers should > suddenly start to form, > and we might also be able to calculate an estimate > of streamer length > too (as a function of topvolts). Given those two > separate figures (or > functions) we would then have the choice of > optimising for max topvolts > or max streamer length, using the same genetic > software but with two > different merit functions. > > Cheers, > -- > Paul Nicholson, > Manchester, UK. > -- Regards, boris __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/ Delivered-To: terrellf@mail-dnvr.uswest.net Date: Wed, 23 May 2001 23:26:02 -0500 From: "Bert Hickman" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mozilla-Status2: 00000000 Organization: Stoneridge Engineering X-Mozilla-Draft-Info: internal/draft; vcard=0; receipt=0; uuencode=0; html=0; linewidth=0 X-Mailer: Mozilla 4.75 [en] (Win95; U) X-Accept-Language: en Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Hi Paul, all, Paul wrote: > > Hi Bert, All, > > > If I've understood things right, the leader formation begins and > continues as long as this gradient can be maintained just ahead of the > leader. I guess the significant threshold involves meeting this value at > the surface of the smooth toroid and once a leader begins to form, its > sharp point will then ensure that the leader forms rapidly for quite > some distance - even though the 'background' field from the > topload would, by itself, fall below the 26kV/cm threshold only a little > way from the surface. Subject to the toroid having enough charge > available to support that formation. What stops the leader formation? > I guess either it hits earth or it runs out of charge - the toroid is > depleted and the 26kV/cm cannot be maintained at the tip? So a big > toroid would be reluctant to break out (modest surface gradient), but > it would throw a long streamer as soon as it did (lots of charge > available)? > You understand correctly. Once formed, a leader acts as a "lossy" extension of the toroid. Initial breakout occurs preferentially when the HV terminal is positively polarized. And, once an initial cathode-directed leader has formed, it's conductivity immediately begins to decrease, reducing the effective potential at the leader tip. This is due to ion recombination and channel cooling through thermal diffusion and radiation. High channel conductivity can only be maintained by passing additional displacement current into the leader capacitance and transfering additional charge into the streamers that fan outwards from the leader tip. A leader will continue to propagate only if the HV terminal voltage is increased sufficiently to overcome channel losses so as to maintain an E-field of at least 26 kV/cm at the leader tip. If the terminal voltage continues to rise at a sufficiently rapid rate, the leader will continue to advance (albeit in jumps). Leader propagation ceases as we approach peak terminal voltage and dv/dt approaches zero. Now if the terminal voltage begins to decline (as we pass the first positive voltage peak), channel current can now actually reverse direction as we begin to "pull" charge out of the region surrounding the leader tip, and as we discharge channel capacitance. While the resulting displacement current will not (usually) extend the leader, it does help maintain leader conductivity until we begin the NEXT positive half cycle. > I'm afraid I've got some more questions! > > > A long spark (>6 cm) is characterized by multiple avalanches in an > > evolutionary sequence: streamer flash(es) --> leader propagation > > (fed by groups of streamers) --> spark (if leader > bridges the gap). > > Can we assume that this whole sequence takes place in a timescale short > compared with an RF cycle - I suppose thats so because if not there > would be a bigger frequency dependence? > The timescale of a given "step" is much shorter than an RF cycle - in fact, multiple steps can occur on the way towards the peak positive voltage in a given RF cycle. While leader growth will preferentially occur on the rising half of positive half-cycles, displacement current heating (maintaining channel conductivity) will occur during both positive and negative cycles. > When the HT falls away, do things recombine and settle down sufficiently > that on the next half cycle there is no 'memory' of the previous half > cycle? There's certainly a "memory" effect between successive RF cycles, and also between "beats" (primary-secondary energy transfer cycles) due to displacement current heating. There's also ample empirical evidence of memory effects between successive "bangs" (at least in in disruptive systems running at breakrates of ~100 BPS or above). This may be due to residual imbalances in the space charge surrounding the toroid from the previous discharges, or preferential reignition of the "warm" path left by the previous leader's root. > > Ultimately what I'm fishing around for is some confidence that some > acceptable and realistic account can be taken of the breakout > thresholds, otherwise attempts at non-linear time domain modeling will > founder on that point. I feel as though we are on top of the technical > matter of computing the response and now, quite suddenly it seems, we > are up against this more difficult problem of finding a load conductance > function which provides an acceptable summary description of the > breakout dynamics. > > If I've got things right, then Terry should be able to calculate quite > easily the top voltage at which streamers should suddenly start to form, > and we might also be able to calculate an estimate of streamer length > too (as a function of topvolts). Given those two separate figures (or > functions) we would then have the choice of optimising for max topvolts > or max streamer length, using the same genetic software but with two > different merit functions. Sounds like a good starting point... BTW, some pictures showing the bright blue "glow" of collective streamer flashes and streamers extending beyond the tips of leaders can be seen in the excellent pictures taken by Mike Hammer and Sue Gaeta - check out: http://www.aquila.net/bert.hickman/photos/corona2.jpg http://www.aquila.net/bert.hickman/photos/P1300073.jpg > > Cheers, > -- > Paul Nicholson, > Manchester, UK. > -- Best regards, -- Bert -- -- Bert Hickman Stoneridge Engineering Email: bert.hickman@aquila.net Web Site: http://www.teslamania.com Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 25 May 2001 07:56:35 +0100 From: "Paul" Sender: root@abelian.demon.co.uk To: terrellf@uswest.net X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: E-Tesla6 anomoly Hi Terry, Browsing throught the list archives - came across your recent post, Referring to http://hot-streamer.com/temp/MattFeature.gif > In one case the rail is inside the sphere and in the other it is > outside. There is a "blip" (Microsoft = feature :-))) as it passes > through. shouldn't the strike rail be outside the Gauss surface for correct Fres? If inside - flux collected by the ring will make it to earth without your program accounting for it - hence the discontinuity. Why not use a Gauss cylinder - the sphere thing seems a lot of hard work and I don't see what you gain? BTW, you seem to have about three or more email addresses, is there a 'best' one to use? twftesla@qwest.net terrellf@uswest.net terry@hot-streamer.com Looking forward to breakout voltage estimates from E-Tesla! Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 25 May 2001 10:46:15 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: [TSSP] Loss due to ultrasonic mechanical vibration of secondary? Reply-To: tssp@abelian.demon.co.uk Hi All, Fishing around for additional loss mechanisms for a secondary coil, has anyone ever looked into the possibility that mechanical vibration of the winding at the RF frequency could be responsible for some part of the overall dissipation? Perhaps a winding which is less than perfectly rigid might cause minute inelastic deformations of the winding/former, especially in something like sonotube which presumably damps sound vibrations rather well. Just something to ponder. Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 25 May 2001 11:56:52 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk boris petkovic wrote: > Surely,~26-28KV/cm critical surface field would be all > right to assume for TC freqs (and standard > atm.conditions) Yes, I'll go with that for now. > But,what counts the most here is one thing:TIME... > By this I don't mean just time of TC waveform > exceeding this treshold ,but also time pauses between > 2 primary condensers bangs-->in other words some > depedence on BPS (belive it or not!). > John Freau experiments with smooth toroids are very > valuable when we speak of this matter. > He has noticed smaller bang size (innital cap energy) > needed for visual spark breakout with higher BPS rate > than with lower BPS rate. Much smaller? 5%, 20% ? Looks like the function that delivers breakout potential must have BPS as a parameter. > Few months ago,I contacted with M.Denicolai and he > said he would make controled experiments on his thor > in order to investigate such phenomenon this summer . Ah good. I'm hoping Marco will also take some more Q measurements, too, he has a 'missing loss' similar to mine - around 30% not accounted for. > Question improved E-Tesla program should answer to: > What max. toroid surface field corresponds to 400 KV > voltage on THOR secondary? Definately. E-Tesla's field plotting will turn out to be the key to all this. Terry? > > Can we assume that this whole sequence takes place > > in a timescale short > > compared with an RF cycle - I suppose thats so > > because if not there > > would be a bigger frequency dependence? > --- > I think we can,althought some scopings on 250 Khz+ > coils I saw in past ,displayed not so small > differences. > --- We could still produce useful conclusions if we restricted to coils with Fres < say 200 kHz ? To begin with anyway. If we can demonstrate agreement with large low-Fres coils, that would give a baseline from which to explore the higher Fres regime. Looks like we need to devise experiments to measure the breakout potentials from large(ish) coils with smooth toroids at low BPS and compare these with ET predictions. Hopefully, in this regime, some of the more subtle time-dependent features described by Bert will be less intrusive and we can get a good account of the dynamics just with a simple threshold. Lets see if can achieve say +/-10% agreement. Getting well outside my field, this. Marc and Terry, over to you! Cheers, -- Paul Nicholson, Manchester, UK. -- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 25 May 2001 11:26:28 -0700 (PDT) From: "boris petkovic" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk > > John Freau experiments with smooth toroids are > very > > valuable when we speak of this matter. > > He has noticed smaller bang size (innital cap > energy) > > needed for visual spark breakout with higher BPS > rate > > than with lower BPS rate. > > Much smaller? 5%, 20% ? Looks like the function > that > delivers breakout potential must have BPS as a > parameter. --- Strange might seem at first sight,but in light of experimental evidence we must accept BPS rate influence .About percantages and functionaI depedance I can't tell anything specific. There must be some kind of free cold discharge always present when local TC electrical field exceeds 28KV/cm ,but the problem appears to me that in some cases ,even in a complete dark, we cannot see any signs of such discharge around toroid. There is my beleif today that difference between 50 BPS and 100 BPS is more expressed than between 100 BPS and 300 BPS,and almost no difference between 1000 BPS and 300 BPS. Careful experiments must be made for final decison. --- > > Question improved E-Tesla program should answer > to: > > What max. toroid surface field corresponds to 400 > KV > > voltage on THOR secondary? > > Definately. E-Tesla's field plotting will turn out > to be the > key to all this. --- Agreed. --- > > I think we can,althought some scopings on 250 Khz+ > > coils I saw in past ,displayed not so small > > differences. > > --- > > We could still produce useful conclusions if we > restricted > to coils with Fres < say 200 kHz ? To begin with > anyway. --- Paul,the best we can do here at moment to assume simple average impedance of corona arc connecting top of resonator (somethig Terry has done with figures 1 pf/' ,200 kohm). Growing of arc during TC operation is almost impossible to simulate.Even the moment of transition to stable leader state in case of TC is wraped in enigma. Personally,I have seen scoping of one particular coil's top current that has shown incredible chaos degree when discharged into air by thin wire put on top terminal. That's the why I give up in advance of any try to get deep into generalisation of TC arcs dynamics. Concentration to pre-disharging era of TC secondary stay my main interest (unusual to hear that from the mouth of spark junky) --- > If we can demonstrate agreement with large low-Fres > coils, > that would give a baseline from which to explore the > higher Fres regime. --- Experiments,experiments ,and only experiments .. -- > > Looks like we need to devise experiments to measure > the > breakout potentials from large(ish) coils with > smooth toroids > at low BPS and compare these with ET predictions. > --- Positive Regards, boris __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/ Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 25 May 2001 11:37:44 -0700 (PDT) From: "boris petkovic" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk Subject: Re: [TSSP] Loss due to ultrasonic mechanical vibration of secondary? Reply-To: tssp@abelian.demon.co.uk --- Paul wrote: > Hi All, > > Fishing around for additional loss mechanisms for a > secondary coil, > has anyone ever looked into the possibility that > mechanical vibration > of the winding at the RF frequency could be > responsible for some part > of the overall dissipation? > > Perhaps a winding which is less than perfectly rigid > might cause > minute inelastic deformations of the winding/former, > especially in > something like sonotube which presumably damps sound > vibrations > rather well. > > Just something to ponder. > --- This is a very brave idea and worth of checking. Definitely,secondary system experince some losses via sound energy.How much ,hard to tell. Heck,I think you're on a good track! Cheers, Boris __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/ Message-Id: <4.1.20010525133005.0183ce00@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Re: [TSSP] Genetic optimisation Cc: Bcc: X-Attachments: In-Reply-To: <3B0BE939.AC4F28A2@abelian.demon.co.uk> References: <4.1.20010520204358.01809240@pop.dnvr.qwest.net> <3B08E987.9674148@abelian.demon.co.uk> <3B090D9D.A9C591A3@aquila.net> <3B093FC4.4AF0FC51@abelian.demon.co.uk> <3B0A5DD7.EDF57F8D@aquila.net> Message-Id: <4.1.20010525133005.0183ce00@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Paul, Bert's 26kV/cm number is very good. When the voltage on a sphere of 1cm radius reaches 26kV it will arc outward. It is a linear function so a 2cm radius sphere is 52kV and a three is 78kV... Us high altitude folks have a bit differnt numbers due to the thin air up here ;-)) My spark gap distance are differnt than most poples since I live at 5400 feet. The chart at: http://hot-streamer.com/TeslaCoils/Misc/SGapVolt.jpg "may" be useful here... A toroid's surface is a bit different than a sphere and has the toroid diameter as an additional factor. E-Tesla6 can do this calculation for a given case but does not provide a simple equation. However there is probably one for this or it can be derived. I suspect the answer is very close to 26kV per cm of cord diameter. Cheers, Terry At 05:45 PM 5/23/2001 +0100, you wrote: >Hi Bert, All, > >Bert Hickman wrote: > >> I only wish I had more time to devote... > >I wish you had too - things always seem a lot clearer after you've put >them into words. > >Looks like 26kV/cm is the figure to work with at normal TC frequencies, > - comes as some relief that the value is not severely frequency >dependent, things would be tricky if so! > >If I've understood things right, the leader formation begins and >continues as long as this gradient can be maintained just ahead of the >leader. I guess the significant threshold involves meeting this value at >the surface of the smooth toroid and once a leader begins to form, its >sharp point will then ensure that the leader forms rapidly for quite >some distance - even though the 'background' field from the >topload would, by itself, fall below the 26kV/cm threshold only a little >way from the surface. Subject to the toroid having enough charge >available to support that formation. What stops the leader formation? >I guess either it hits earth or it runs out of charge - the toroid is >depleted and the 26kV/cm cannot be maintained at the tip? So a big >toroid would be reluctant to break out (modest surface gradient), but >it would throw a long streamer as soon as it did (lots of charge >available)? > >I'm afraid I've got some more questions! > >> A long spark (>6 cm) is characterized by multiple avalanches in an >> evolutionary sequence: streamer flash(es) --> leader propagation >> (fed by groups of streamers) --> spark (if leader > bridges the gap). > >Can we assume that this whole sequence takes place in a timescale short >compared with an RF cycle - I suppose thats so because if not there >would be a bigger frequency dependence? > >When the HT falls away, do things recombine and settle down sufficiently >that on the next half cycle there is no 'memory' of the previous half >cycle? > >Ultimately what I'm fishing around for is some confidence that some >acceptable and realistic account can be taken of the breakout >thresholds, otherwise attempts at non-linear time domain modeling will >founder on that point. I feel as though we are on top of the technical >matter of computing the response and now, quite suddenly it seems, we >are up against this more difficult problem of finding a load conductance >function which provides an acceptable summary description of the >breakout dynamics. > >If I've got things right, then Terry should be able to calculate quite >easily the top voltage at which streamers should suddenly start to form, >and we might also be able to calculate an estimate of streamer length >too (as a function of topvolts). Given those two separate figures (or >functions) we would then have the choice of optimising for max topvolts >or max streamer length, using the same genetic software but with two >different merit functions. > >Cheers, >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 25 May 2001 13:40:10 -0600 From: "Terrell W. Fritz" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Sender: terrellf@pop.dnvr.qwest.net X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Subject: Re: [TSSP] Genetic optimisation Reply-To: tssp@abelian.demon.co.uk Hi Paul, Bert's 26kV/cm number is very good. When the voltage on a sphere of 1cm radius reaches 26kV it will arc outward. It is a linear function so a 2cm radius sphere is 52kV and a three is 78kV... Us high altitude folks have a bit differnt numbers due to the thin air up here ;-)) My spark gap distance are differnt than most poples since I live at 5400 feet. The chart at: http://hot-streamer.com/TeslaCoils/Misc/SGapVolt.jpg "may" be useful here... A toroid's surface is a bit different than a sphere and has the toroid diameter as an additional factor. E-Tesla6 can do this calculation for a given case but does not provide a simple equation. However there is probably one for this or it can be derived. I suspect the answer is very close to 26kV per cm of cord diameter. Cheers, Terry At 05:45 PM 5/23/2001 +0100, you wrote: >Hi Bert, All, > >Bert Hickman wrote: > >> I only wish I had more time to devote... > >I wish you had too - things always seem a lot clearer after you've put >them into words. > >Looks like 26kV/cm is the figure to work with at normal TC frequencies, > - comes as some relief that the value is not severely frequency >dependent, things would be tricky if so! > >If I've understood things right, the leader formation begins and >continues as long as this gradient can be maintained just ahead of the >leader. I guess the significant threshold involves meeting this value at >the surface of the smooth toroid and once a leader begins to form, its >sharp point will then ensure that the leader forms rapidly for quite >some distance - even though the 'background' field from the >topload would, by itself, fall below the 26kV/cm threshold only a little >way from the surface. Subject to the toroid having enough charge >available to support that formation. What stops the leader formation? >I guess either it hits earth or it runs out of charge - the toroid is >depleted and the 26kV/cm cannot be maintained at the tip? So a big >toroid would be reluctant to break out (modest surface gradient), but >it would throw a long streamer as soon as it did (lots of charge >available)? > >I'm afraid I've got some more questions! > >> A long spark (>6 cm) is characterized by multiple avalanches in an >> evolutionary sequence: streamer flash(es) --> leader propagation >> (fed by groups of streamers) --> spark (if leader > bridges the gap). > >Can we assume that this whole sequence takes place in a timescale short >compared with an RF cycle - I suppose thats so because if not there >would be a bigger frequency dependence? > >When the HT falls away, do things recombine and settle down sufficiently >that on the next half cycle there is no 'memory' of the previous half >cycle? > >Ultimately what I'm fishing around for is some confidence that some >acceptable and realistic account can be taken of the breakout >thresholds, otherwise attempts at non-linear time domain modeling will >founder on that point. I feel as though we are on top of the technical >matter of computing the response and now, quite suddenly it seems, we >are up against this more difficult problem of finding a load conductance >function which provides an acceptable summary description of the >breakout dynamics. > >If I've got things right, then Terry should be able to calculate quite >easily the top voltage at which streamers should suddenly start to form, >and we might also be able to calculate an estimate of streamer length >too (as a function of topvolts). Given those two separate figures (or >functions) we would then have the choice of optimising for max topvolts >or max streamer length, using the same genetic software but with two >different merit functions. > >Cheers, >-- >Paul Nicholson, >Manchester, UK. >-- Message-Id: <4.1.20010525141449.01866e28@pop.dnvr.qwest.net> To: "Paul" From: "Terrell W. Fritz" Subject: Re: E-Tesla6 anomoly Cc: Bcc: X-Attachments: In-Reply-To: <3B0E0223.BA7B139@abelian.demon.co.uk> References: Message-Id: <4.1.20010525141449.01866e28@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Paul, At 07:56 AM 5/25/2001 +0100, you wrote: >Hi Terry, > >Browsing throught the list archives - came across your recent >post, > >Referring to http://hot-streamer.com/temp/MattFeature.gif > >> In one case the rail is inside the sphere and in the other it is >> outside. There is a "blip" (Microsoft = feature :-))) as it passes >> through. > >shouldn't the strike rail be outside the Gauss surface for >correct Fres? If inside - flux collected by the ring will make it >to earth without your program accounting for it - hence the >discontinuity. Yes, I will forward you a followup post on this. > >Why not use a Gauss cylinder - the sphere thing seems a lot >of hard work and I don't see what you gain? I originally tried the cylinder but It got to be a mess. Jim Monty also tried the cylinder at first (despite my warning not to :-))) and he ended up going the sphere route too... You would still run into a crash with a large top load diameter and a small high strike rail. There is always some situation that the strike rail can crash in. Jim though of using section of cylinders and such but he never went forward with that. for me the sphere just seemed natural and only one set of messy equations was needed without having to worry with the top and bottom and corners. It is history now :-)) > >BTW, you seem to have about three or more email addresses, is >there a 'best' one to use? > > twftesla@qwest.net > terrellf@uswest.net > terry@hot-streamer.com twftesla@qwest.net - is the account I use just for the Tesla list moderation but I get mail there too. Best to avoid it, but no big deal. terrellf@qwest.net (identical to terrellf@uswest.net, they got bought out and changed the domain name. uswest.net will stop working in about 6 months so everything should be qwest.net now.) This is my main mail and most stuff should come here. terry@hot-streamer.com - Is the mail I use for the hot-streamer server. I avoid it since I have to stretch over to the other keyboard :-) but it works fine too. They are all check all the time so I get the message regardless. terrellf@qwest.net is the best one. > >Looking forward to breakout voltage estimates from E-Tesla! Finally working on this now :-)) I have gotten a bit behind in mails and such do to work but things are calm now and catching up. Great work you are doing with the program and stuff. I need to get the latest stuff and catch up on studying it. I will work on getting it in the Windows form so those trapped behind the Windows can play too :-)) I have the next 10 days off work so playtime is upon me :-)))) Cheers, Terry > >Cheers, >-- >Paul Nicholson, >Manchester, UK. >-- Delivered-To: terrellf@mail-dnvr.uswest.net Date: Fri, 25 May 2001 21:35:37 +0100 From: "Paul" Sender: owner-tssp@abelian.demon.co.uk To: tssp@abelian.demon.co.uk X-Mailer: Mozilla 4.07 [en] (X11; I; Linux 2.2.4 i686) Subject: Re: [TSSP] Loss due to ultrasonic mechanical vibration of secondary? Reply-To: tssp@abelian.demon.co.uk boris petkovic wrote: > This is a very brave idea and worth of checking. > Definitely,secondary system experince some losses via > sound energy.How much ,hard to tell. > Heck,I think you're on a good track! And if the tube happens to have an acoustic resonance close to the electrical Fres... you have a potentially very lossy tube. Maybe Terry should add some weights to the ends of his lossy sonotube coil. Plenty of experiments suggest themselves... Measuring the mechanical vibrations in the presence of large E and B fields could be tricky. Transfer the vibrations along a taute fishing line to an acoustic pickup some distance away perhaps. How about in reverse. Have the secondary carry a few amps DC and expose it to ultrasonic sound at Fres. Look for induced EMF superimposed on the DC across the secondary terminals. Amazing how Tesla coiling continues to present endless possibilities for worthwhile investigations. Cheers, -- Paul Nicholson, Manchester, UK. -- Message-Id: <4.1.20010525143136.01943e70@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Re: [TSSP] Loss due to ultrasonic mechanical vibration of secondary? Cc: Bcc: X-Attachments: In-Reply-To: <3B0E29E7.6C458B1@abelian.demon.co.uk> References: Message-Id: <4.1.20010525143136.01943e70@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Paul, I notice in my solid state gap work that you can hear the current going through the wires (everything else is dead silent). Magnetostriction I think is the term and there is the magnetic attraction between wires with opposing currents. By knowing the current profile and the distance and mass of the conductors, one could work out the force and how far they move. I am sure the secondary vibrates a bit but I would think the currents in the secondary would cause far less vibration. My SS coil does not seem to vibrate the secondary but it would be at 350kHz which is hard to hear :-)) I have never heard of polyurethane cracking or wear in secondary wiring do to vibration of this type. There is also corona around the coils and stuff here and there (like off the tope edge of the secondary) that is a loss that may be significant. the field stress plots coming up will show these points. Cheers, Terry At 10:46 AM 5/25/2001 +0100, you wrote: >Hi All, > >Fishing around for additional loss mechanisms for a secondary coil, >has anyone ever looked into the possibility that mechanical vibration >of the winding at the RF frequency could be responsible for some part >of the overall dissipation? > >Perhaps a winding which is less than perfectly rigid might cause >minute inelastic deformations of the winding/former, especially in >something like sonotube which presumably damps sound vibrations >rather well. > >Just something to ponder. > >Cheers, >-- >Paul Nicholson, >Manchester, UK. >-- Message-Id: <4.1.20010525134600.0185b550@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Re: [TSSP] Genetic optimisation Cc: Bcc: X-Attachments: In-Reply-To: <20010523185010.16122.qmail@web11803.mail.yahoo.com> References: <3B0BE939.AC4F28A2@abelian.demon.co.uk> Message-Id: <4.1.20010525134600.0185b550@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi Boris, I remember Marco's coil as having a power supply problem that may have affected this: http://www.pupman.com/listarchives/2000/November/msg00117.html I forget the details but that may have affected his toroid breaking out at that time. Marco's system is very good for this since his is so well controlled. I have his system running now so I'll report back. I may have to use a fairly large grid size and adjust things to get good resolution at the toroid surface... Catching up on mail now ;-)) Cheers, Terry At 11:50 AM 5/23/2001 -0700, you wrote: > >Hi Paul,All, > >> Looks like 26kV/cm is the figure to work with at >> normal TC frequencies, >> - comes as some relief that the value is not >> severely frequency >> dependent, things would be tricky if so! >--- >Surely,~26-28KV/cm critical surface field would be all >right to assume for TC freqs (and standard >atm.conditions) >Definitely,if such field appears for a certain time >,it will get show to start happening per see. >But,what counts the most here is one thing:TIME... >By this I don't mean just time of TC waveform >exceeding this treshold ,but also time pauses between >2 primary condensers bangs-->in other words some >depedence on BPS (belive it or not!). >John Freau experiments with smooth toroids are very >valuable when we speak of this matter. >He has noticed smaller bang size (innital cap energy) >needed for visual spark breakout with higher BPS rate >than with lower BPS rate. >Few months ago,I contacted with M.Denicolai and he >said he would make controled experiments on his thor >in order to investigate such phenomenon this summer . > >Something else you should know as well.With about 400 >KV peak voltage on his coil Marco was unable to get >breakout of his 20*150 cm smooth toroid. >Only with Ep~10 J he started to observe first faint >streamers coming from the terminal. >Question improved E-Tesla program should answer to: >What max. toroid surface field corresponds to 400 KV >voltage on THOR secondary? >---- > > >> >> If I've understood things right, the leader >> formation begins and >> continues as long as this gradient can be maintained >> just ahead of the >> leader. I guess the significant threshold involves >> meeting this value at >> the surface of the smooth toroid and once a leader >> begins to form, its >> sharp point will then ensure that the leader forms >> rapidly for quite >> some distance - even though the 'background' field >> from the >> topload would, by itself, fall below the 26kV/cm >> threshold only a little >> way from the surface. Subject to the toroid having >> enough charge >> available to support that formation. What stops the >> leader formation? >> I guess either it hits earth or it runs out of >> charge - the toroid is >> depleted and the 26kV/cm cannot be maintained at the >> tip? So a big >> toroid would be reluctant to break out (modest >> surface gradient), but >> it would throw a long streamer as soon as it did >> (lots of charge >> available)? >> >> I'm afraid I've got some more questions! >> >> > A long spark (>6 cm) is characterized by multiple >> avalanches in an >> > evolutionary sequence: streamer flash(es) --> >> leader propagation >> > (fed by groups of streamers) --> spark (if leader >> > bridges the gap). >> >> Can we assume that this whole sequence takes place >> in a timescale short >> compared with an RF cycle - I suppose thats so >> because if not there >> would be a bigger frequency dependence? >--- >I think we can,althought some scopings on 250 Khz+ >coils I saw in past ,displayed not so small >differences. >--- > >> >> When the HT falls away, do things recombine and >> settle down sufficiently >> that on the next half cycle there is no 'memory' of >> the previous half >> cycle? >> >> Ultimately what I'm fishing around for is some >> confidence that some >> acceptable and realistic account can be taken of the >> breakout >> thresholds, otherwise attempts at non-linear time >> domain modeling will >> founder on that point. I feel as though we are on >> top of the technical >> matter of computing the response and now, quite >> suddenly it seems, we >> are up against this more difficult problem of >> finding a load conductance >> function which provides an acceptable summary >> description of the >> breakout dynamics. >> >> If I've got things right, then Terry should be able >> to calculate quite >> easily the top voltage at which streamers should >> suddenly start to form, >> and we might also be able to calculate an estimate >> of streamer length >> too (as a function of topvolts). Given those two >> separate figures (or >> functions) we would then have the choice of >> optimising for max topvolts >> or max streamer length, using the same genetic >> software but with two >> different merit functions. >> >> Cheers, >> -- >> Paul Nicholson, >> Manchester, UK. >> -- > >Regards, >boris > >__________________________________________________ >Do You Yahoo!? >Yahoo! Auctions - buy the things you want at great prices >http://auctions.yahoo.com/ Message-Id: <4.1.20010525154307.00a0e4e0@pop.dnvr.qwest.net> To: tssp@abelian.demon.co.uk From: "Terrell W. Fritz" Subject: Toroid field stress Cc: Bcc: X-Attachments: In-Reply-To: <3B0EC219.AF920927@abelian.demon.co.uk> References: <20010525183744.39856.qmail@web11801.mail.yahoo.com> Message-Id: <4.1.20010525154307.00a0e4e0@pop.dnvr.qwest.net> X-Eudora-Signature: <> Hi All, I finally got away from work issues and can get back to doing something important :-)) I gave the Thor a trial test run: http://hot-streamer.com/temp/tsspTest1.gif http://hot-streamer.com/temp/tsspTest2.gif Hmmm, that looks a little odd so I'll have to work on it :-) Just getting the bugs worked out here... The setup file is: http://hot-streamer.com/temp/tsspoutput.txt The 500 x 500 array voltage stress output file with the voltage in V/cm (400kV top load voltage) is at: http://hot-streamer.com/temp/tsspsoutput.xls The "XLS" Excel file is really just a space delimited array of numbers so any data program can import it. I import them into MathCad to make the graphs. Excel would explode on a file this big :-)) This is just the first try here and it looks a bit messed up but I am finally working on it :-)) I have no idea how much space there is between to top of Marco's secondary and the centerline of his toroid so I am guess there. I think I got the toroid cord diameter wrong too... Cheers, Terry