[Prev][Next][Index][Thread]
Re: Light Bulb Experiment (ala Brent Turner)
[The following text is in the "iso-8859-1" character set]
[Your display is set for the "US-ASCII" character set]
[Some characters may be displayed incorrectly]
All:
Thanks to everyone who responded re: the Light Bulb Experiment. Some
specific feedback:
>From Brent Turner:
> First off, what you are actually doing is using the light bulb as a
> sort of power indicator, ala a hot-wire ammeter or watt-meter. Ironically,
> I have seen a 100-watt bulb work better than a 40-watt bulb due to the
> lowered filament resistance (snip>. A low wattage lamp has a higher filament
> resistance, hence for a given.amount of power, will develop more voltage across it.
<SNIP>
I wasn’t able to get a 100 Watt bulb to light, but I easily burned out a
15 Watt bulb. I agree that given a relatively fixed amount of current
flowing through the bulb to the corona, the power dissipated in the lamp
will be a function of R*I^2. A smaller wattage light will develop more
voltage across the filament, lighting it brighter. In my trials, large
wattage bulbs were consistently dimmer.
>From Richard Hull:
<SNIP>
> The impedance of a Tesla coil's terminal is very high. The impeadance of
> pure air corona arc is even higher. Thus the voltage differential
> available across the bulb is high and the current is at a maximum for
> this condition. (It lights to a brillant glow) When an arc strikes
> ground, the current is maximum as the coil goes to a very low impedance,
> but the voltage across the coil and lamp are at a minimum (dim or
> invisible glow). To light a bulb, you need both current and voltage.
> you will note that the arc current is brilliant on a ground hit. Most of
> the energy is being fed to the air arc and not the lamp.
<SNIP>
Based upon an inductance of 73400 uH and an estimated self capacitance
of
about 15.8 pF, my secondary coil’s surge impedance should be about
SQRT(L/C)=42k Ohms. Based upon previous measurements of secondary Q
with an
air-only discharge, the impedance of the air-only corona looks to be
larger by at least an order of magnitude or more (between 500k and 1
Megohm). However, once the discharge to ground happens, I would expect
the arc impedance to drop, and the instantaneous surge current to
increase greatly. If this current were flowing through the filament, I’d
expect the bulb to increase in brightness unless the overall duty cycle
decreased enough to more than offset the increased current flow.
However. I now think these surge currents may be bypassing the filament
due to unseen arc-overs in the lampbase.
I didn’t quite follow the part about the coil going to a very low
impedance during an arc strike to ground... Could you go into that a
little bit more?? Thanks!
>From Robert Stephens:
<SNIP>
> As a possible attempt at explaining your phenomenon of the light
> dimming as power to make a bright arc is drawn through it to a ground target,
> I wonder if under these conditions, since there is more RF current being passed
> through the bulb circuit (we would think?) perhaps the inductance
> of the filament is causing the energy to seek a more direct path and
> there is some bypass arcing occuring in the base of the lamp or in
> the socket.
<SNIP>
And From Dave Huffman:
<SNIP>
> Bert can you tell if your socket/bulb is arcing over at these higher
> levels? The power you are putting in has to go someplace.
<SNIP
Bingo!!
Robert and Dave, after running a couple more experiments, I think I
agree with both of you. First, instead of using an incandescent lamp, I
substituted a 10” long fluorescent light, with one end hooked to the
toroid, and the other pointing straight up, with a gap of about 48” to
grounded wires above. This eliminated any inductive loops, and removed
the non-linear voltage/illumination relationship associated with the
tungsten filament. This time, the fluorescent light glowed MORE
brightly when passing the heavy ground surges.
I think I understand at least part of what is going on. Once most of the
primary energy has transferred to the secondary, the toroid and
distributed coil capacitance charge to a high voltage relative to
ground., At the voltage peak, all the energy in the system will be
concentrated in the electrostatic field around the toroid and top
portion of the coil. In the case of my coil, I estimate the toroid
capacitance to be about 26pF versus about 16pF for the coil. If we
suddenly discharge this stored energy to ground, we should get a high
current damped-wave oscillatory discharge (on a smaller scale, but
similar to lightning). The frequency will be a function of the
inductance of the arc and the ground path, and the damping a function of
the losses (primarily thermal losses of the arc). A “back of the
envelope” calculation indicates that the frequency of the discharge
oscillations should be in the 10 - 20 Mhz range. I made some actual
measurements, via a storage scope connected to a small antenna plate,
and set up the coil to do single shot 18” discharges to a grounded rod.
This revealed the first half of the energy transfer “beat” followed by
complex waveforms with components in the 10-20 Mhz range. However, once
the ground-fault happens, virtually all of the energy in the toroid is
dissipated very rapidly (microseconds).
By the way, these ground discharge currents look like they will be quite
heavy, since the fully-charged toroid fault current should not be
limited by the surge impedance of the secondary coil. The toroid would
tend to look more like a charged cloud during a cloud-to-ground
lightning stroke, and the surge current would be limited primarily by
the dynamic resistance of the ionized arc-path through the air, and by
the combined inductances of the arc-path and the groundpath. The arc
has a negative resistance characteristic (i.e., more current flowing
reduces the resistance of the arc), making it difficult to calculate the
current with any degree of precision. However, these current surges
probably far exceed 100 amperes with duty cycles of a fraction of a
percent. These heavy RF currents will easily “jump around” lumped
inductances, such as the looped leads going to the tungsten filament.
Since the main current flow now bypasses the filament, it dims. Heavy
current pulses may also explain the “jumping filaments” under this type
of excitation.
>From Edward Phillips:
<Snip>
> As far as discharge from the secondary goes, that both provides
> a shunt path and reduces the secondary Q, which in turn reduces the
> coupling (which is proportional to k times square root of primary
> and secondary Q's). Think the latter effect is what is happening.
<SNIP>
Ed, this may be why when I tried putting a light in the secondary base
groundpath it dimmed during the ground surges. I would not expect large
fault currents from the charged toroid to flow through the secondary
base path. However, if the discharge does not begin until much of the
energy has been transferred to the secondary, we’ve already gone through
the period where the secondary ground currents have max’ed out. Hmmm....
Again, thanks everybody for the various ideas and suggestions! Any
further thoughts or experimental results that shed more light on this
are welcomed.
As always, flames, stones, and brickbats are also welcomed! :^)
-- Bert --