[Prev][Next][Index][Thread]
DC on Tesla coil
----------
From: Malcolm Watts [SMTP:MALCOLM-at-directorate.wnp.ac.nz]
Sent: Wednesday, March 25, 1998 3:47 PM
To: Tesla List
Subject: Re: DC on Tesla coil
Hi Viv,
Here's a few thoughts for what it's worth:
> From: Vivian [SMTP:V.C.Watts-at-btinternet-dot-com]
> Sent: Tuesday, March 24, 1998 2:46 PM
> To: Tesla List
> Subject: DC on Tesla coil
>
> Hi All,
> On Saturday I was demonstrating my TC to a good friend when due to a
> pull on the ground cable unbeknown to me my secondary became disconnected
> from earth. When firing it up sparks ensued from around the base and from
> Primary to Secondary. So I shut it down very quickly. Now when I
> re-connected the secondary to the ground I got a large spark just before
> contact. So the secondary had stored charge. Question is was it a static
> charge in the sense of formed by the insulation of the coil as proposed on
> the list or is their a tendency for a level of DC to be built up but
> naturally bleed away by normal connection to ground. If a spark ensues to
> an object and the connection is terminated at the peak of the Oscillation
> then will a charge be retained?
I would think so. I think energy stored in the coil dielectrics would
explain your observations. I too have experienced this. I think
everyone would have in fact.
> Before the idea is dismissed outright one might like to consider another
> experience I have had. I use to work for a company making automatic test
> equipment to test bare circuit boards (no components fitted) for shorts and
> opens. We charged up a track on the circuit board to 500V or so and
> measured how long it took at a known source impedance. If it took longer
> than expected it was either directly shorted to something else or was
> leaking via high resistance to another trace. Times measured were in nano
> seconds and detection of shorts to another trace measuring only 0.02 pf
> could be resolved or resistance's up to 1000MOhms.
>
> One of the power planes on the board would be grounded. The problem we
> experienced was that if the circuit board had multiple power planes, testing
> one plane i.e. charging it up to 500v would induce a voltage on another. OK
> so far, but the machine always discharged the plane being tested, before
> proceeding to test another. But now here's the difficult thing, the other
> power planes retained their charge. The consequence was that when that
> charged plane came to be tested it blew the electronics away. Even a Spice
> model showed a charge build up with no rectification and yet all that was
> happening was another part of the board was first charged up then
> discharged.
How much of their charge did the adjacent tracks retain? I would
imagine that the time constants involved in charge leaking away would
be considerable and the lower the voltage, the longer it would take
in some circumstances.
> The consequence of all this was that after discussing this experience with
> my friend I now know less than I thought I did.
>
> Here's a few more questions some kind sole might like to put my thoughts
> back on track.
>
> 1/ When the gap fires the current in the gap will be from the transformer
> and from the capacitor. What current is hardest to quench? For the period
> of one half cycle the transformer current could be considered like DC but
> the
> capacitor current will be AC. Is one harder to quench than another.
The oscillating tank energy is by far the hardest to stop in my
experience. Peak tank currents dwarf the charging current. I see
active quenching as doing more to disconnect the supply than opening
up the tank. I have done it in what would appear to be ideal
circumstances in sshot tests but the sad reality is that in order to
do it, primary losses increased enormously. In fact, achieving a
first notch quench with a powerful jet of air, the gap discharge was
blown into an arc more than 3" long and secondary output voltage
dropped by about 50%.
> 2/ I have read descriptions like ... "When the Gap is quenched the energy
> goes into the secondary". Is it really breaking the tank circuit and thus
> removing the lossy gap from the circuit, increasing Q and allowing the
> secondary to ring true that gets the sparks going?
Again, my experiments have convinced me that you must considered the
system coupled during spark production. I have yet to see a
convincing demonstration that one can quench before getting an output
streamer.
> 3/ Are the oscillations, when the gap fires, like a carrier with two
> sidebands until the gap quenches and then become a single carrier.
I think this is an issue that has been unnecessarily confused by the
introduction of spectrum analysers. One must realize that in using
such an instrument one is throwing away all time-related information
about what the system is doing. If you look on the o'scope, you see
that with respect to time, the secondary rings up at its natural
frequency and the primary rings down at its natural frequency.
Decomposing that AM waveform into its frequency components tells you
nothing about the relative timing and phasing of the components. We
are surely interested about what the coil is doing in the time domain
i.e. Vtop wrt time and the scope shows that quite accurately IMHO.
> 4/ Why does the conventional TC have to be loosely coupled. Why can't we
> get more bang by increasing coupling if the insulation could stand it.
In my experience, it IS solely a matter of clearances. Commercial
resonant transformers used in particle accelerators at k's around 0.6
are run in inert atmospheres to allow much closer proximity of prim
to sec. The diagrams I've seen depict the secondary as a helix and
the primary as a cone which almost runs the same height as the
secondary.
> > With Brain Ache
>
> Viv Watts UK.
These are just my opinions based on experiment. Others may disagree.
Malcolm