[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

*To*: tesla-at-pupman-dot-com*Subject*: Re: Vortex gap loss measurements*From*: "Daniel Boughton" <daniel_boughton-at-yahoo-dot-com> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)*Date*: Sun, 03 Sep 2000 18:37:24 -0600*Delivered-To*: fixup-tesla-at-pupman-dot-com-at-fixme

Malcom: It is obvious that the gap cannot be described via linear resitive model(see my additional comments to Gary). Plasma is a very unique thing. I have worked for years with RF generators, impedance matching networks to plasma chambers for silicon etching. With a rich variety of gas mixtures, pressures, temperatures, volumes, and frequencies a wide variety of impedance responses are dynamically encountered (primarily capacitive). The matching networks are automated to analyze the load (chamber) impedance by detecting the foward and reverse VSWR and adjust the capacitors (vacuum variable) for max power transfer. The chamber was in a constant flux (as far as impedance goes) when ignited. Even so, a mathematical model of behaviour was derived which acurately predicted the behaviour across all variables. This mathematical model was loaded in the firmware to automate the match. The result was to have precise tuning in a very fast time response (due to microprocessor calculation speed and servo motion to adjust the capacitor). I say all of this for a few reasons: (1.) a linear resistive model cannot purely describe the development of plasma in the gap as you pointed out. (2.) I do not believe that any mathematical model of a single static gap under an air blast of quantified pressure has been developed to date which adequately describes and predicts plasma behaviour in the gap. The circuit (external to the gap)response has been the only point analyzed to date (see Morecroft). (3.) Treating the primary RLC circuit as a lumped element analysis falls woefully short of any real description of circuit phenomena and must be described as distributed. I have found that my rotary spark gap with about 8KW powering it is a tremendous broadcaster of EM energy. The metal shielding I have in my lab actually has plasma balls about 1 cm in diameter form in the cracks between them when running off tune (no sparks issuing from the secondary). The math model must also account for this radiated power loss. Treating it as a lumped element ignores other richer gap phenomena as well. Now as to a linear decrimenting model, the equation for it is quite easy to develop given Gary's data (simple 1st degree equation and we know the envelope and intercepts so a simple direct variation RT equation will account for his data if you haven't already see his scope images) but the gap phenomena as a circuit controller is much more complex. For instance the arc has its own inductive elements as well as capacitive (how much energy is stored in the magnetic field of the inductance and how much in the electrostatic field of the capacitance and what is the resultant impedance? How does impedance change as a function of time? Does it swing from capacitive to inductive? Does it resonate at its own (assumably high frequency and how does this impact the operation of the circuit). Now it is quite clear that if you force an RLC response the circuit will respond to the logaritmic equations (any 2nd year EE student will have done this experiment in lab and derived the equations). One experiment might be to use a plasma tube in place of a gap under strict controls and monitor the conductance (and admitance) of the plasma (although RF plasma engineers have been doing this for years) as the variables are processed. You noted in your message that "The gap is dissipative but that is where the similarities with resistance end. I analysed this and wrote a note on it several years ago." How do you mean dissipative? Do you mean heat loss? Secondly, don't take me wrong (I agree as you can tell from my previous statements) but how did you determine that the gap (I think you mean arc)is dissipative and that is where similarities end? How did you conduct your experiment and do you still have data? Regards, Dan --- Tesla list <tesla-at-pupman-dot-com> wrote: > Original poster: "Malcolm Watts" > <M.J.Watts-at-massey.ac.nz> > > On 2 Sep 00, at 18:36, Tesla list wrote: > > > Original poster: "Daniel Boughton" > <daniel_boughton-at-yahoo-dot-com> > > > > Gary: > > > > This is a very interesting experiment. First, when > you > > measured the slope during the ring down, did you > see > > the same size decrement of each succesive > oscillation? > > Is this what you mean by linear as oppossed to > > logarithmic? Traditional wisdom purports that the > > decay is according to V(t)=V(i)e^-rt where V(i) is > the > > forced initial potential on the capacitor. Your > > results are very interesting however in that it > flies > > in the face of convention. The derived equation > must > > be something like V(t)= V(i)*-krt. I wonder if > without > > the secondary it is linear due to resistive losses > > only. Without the secondary the additional > absorption > > of energy via the secondary mutual inductance is > > missing which accounts for the linear decay? Also > what > > I found interseting was that with the gap distance > the > > slope remained constant. I would have expected > greater > > gap resistance at further distances but it seems > that > > the plasma provides a constant resistance no > matter > > how wide the spark gap is set (within reason of > > course-I 'm sure at a foot the resistance would be > > substantial as compared to 300 mil). > > The linear decrement of a ringing RLC circuit which > has a gap > in series with it was discovered by Stone circa > 1914. The > Corums mention it in their literature but > unfortunately never > went on to use the information in their modelling. > The linear > decrement is entirely due to the gap > characteristics. An RLC > circuit by itself produces only a logarithmic > decrement. This > clearly shows that modelling the gap as a > resistancwe does not > work. You cannot apply the classic time constant > equations to > this situation. The gap is dissipative but that is > where the > similarities with resistance end. I analysed this > and wrote a > note on it several years ago. It is important to > note > (ultimate pedantry) that you *cannot* ascribe a > value for Q to > the primary if the primary includes a gap. You can > compare > various primaries with each other by comparing the > ringdown > slope (gentler is obviously better). > > Regards, > Malcolm > > > > __________________________________________________ Do You Yahoo!? Yahoo! Mail - Free email you can access from anywhere! http://mail.yahoo-dot-com/

- Prev by Date:
**Re: Vortex gap loss measurements** - Next by Date:
**Coil turn ?** - Prev by thread:
**Re: Vortex gap loss measurements** - Next by thread:
**Re: Vortex gap loss measurements** - Index(es):