[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [TCML] QCW sparks
Hi Steve,
Very interesting and thought provoking indeed. Some comments follow.
Steve Ward wrote:
<snip>
http://www.flickr.com/photos/kickermagnet/
Im not 100% sure what to make of it yet. The explanation i like to give is
that its a "self-stretching arc". If you consider the arc length you can
stretch from a 15kV transformer (many feet if there is enough current... say
2A), then the fact that this tesla coil functions at relatively low voltage,
yet still producing long sparks, makes sense. The spark just takes a
loooooong time to get to that length.
Some general rambling (mostly for the benefit of other TCML members
since you already know this):
By using a relatively slow-rising secondary envelope, the leader tip
E-field can slowly lengthen the leader over the extended ramp. However,
the peak E-field is kept below the point where leader tip instability
occurs. Limiting the rate at which the peak E-field grows (the envelope
of the RF) prevents the leader from splitting, and thus prevents current
from being diverted by branching. Faster ramp rates will cause greater
splitting, eventually leading to "frantically" branching discharges.
Leaders are basically arcs - the only difference is that they are
powered by displacement currents (electrical charges being exchanged
to/from the air in the regions around the discharges) rather than
between electrodes.
Vertically-directing the leaders also allows heat from existing channel
to rise. Hot air is less dense, lowering the breakdown voltage of the
air immediately above the leader tip. This helps to lengthen the
resulting leaders, similar to the way an arc will be significantly
longer when oriented vertically.
Finally, the divergence of the E-field is minimal in the flat region
immediately above the toroid. It is known that discharge branching
increases with increasing background E-field. Moving the discharge point
to the side of the toroid should cause increased branching and overall
shorter swords (everything else being the same).
Another interesting point is that the top voltage is a lot lower than V = IZ
would suggest for a lumped model of my secondary coil. Basically, the
secondary is ~25mH, and with 3Apk current through 25mH at 325khz, id expect
a peak voltage of 153kV. The only way i can account for this large
discrepancy in apparent impedance of the coil is that there must be
significant capacitance from the secondary to primary, so the base current
looks much larger compared to say, the toroid current. I'd like to next
work on measuring the current between secondary and toroid and see if it is
consistent with my voltage measurement claims. The fact that the 50kV probe
hasnt shown any signs of stress makes me think that the voltage really isnt
150kV, but really is more like the 56kV i claim.
This sounds reasonable after looking at the geometry of your system.
Measuring the secondary-to-toroid current and (better yet)
toroid-to-breakout point current would be even better to better
understand the dynamics and limiting factors for pseudo-steady-state
discharges such as your sword sparks.
For grins, I estimated the capacitance of a conductive stick-like
vertical discharge. I assumed that your breakout point was 36" above
ground, and used a formula from Terman's Radio Engineer's Handbook (page
119). The results were:
Spark Estimated
Length Capacitance
(feet) (pF)
1 2.35
2 4.31
3 6.18
4 7.97
The capacitive reactance of a single vertical discharge (simplistically
neglecting plasma series resistance) at 325 kHz is about 61.5k ohms.
Your secondary's unloaded impedance is about 51k ohms. This implies
that, with a 4 foot discharge, your secondary/toroid system is being
significantly loaded, and pretty closely matched, by your spark load.
For this case, the spark-loaded output voltage should be a little over
half of the unloaded case - roughly in the ballpark. With this model,
the peak leader current would be about an amp. Actual loading will be
less when plasma path resistance is taken into account. Potentially
future measurements can help to better quantify this.
Apparently, once breakout occurs, spark loading begins to steal an
increasing portion of the power that's being driven into the resonator
from the primary as we ramp up applied voltage. The result appears to be
"soft" clamping of the output voltage to a level that's perhaps only
10-25% above the initial breakout voltage. Sort of like the HV Zener
clamping mentioned in a number of earlier TCML posts.
The next step will be to measure the ramp voltage and work on a spice model
to determine the streamer impedance.
Thoughts?
Great work, Steve! Have you seen Todd Johnson's (a mutual friend)
handheld digital scope? Something like this might just be the ticket for
capturing floating current measurements via a Pearson or Ion Physics CT
atop the toroid.
You may wish to try slowly increasing ON times (using the same linear
ramp rate) to see if leader length is directly proportional to ramp-up
time. Can the system be driven with even longer rampup times? Also, what
is the initial ramp-up rate for the front portion of the waveform? And,
have you measured the coupling coefficient for your system? Looks much
higher than most systems.
Steve
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla
Again, great work - thanks for sharing the results!
You may wish to consider publishing an article in one of the IEEE
journals, like IEEE Transactions on Plasma Science or on Dielectrics and
Electrical Insulation, since this research has not appeared in the
existing literature.
Bert
--
Bert Hickman
Stoneridge Engineering
http://www.capturedlightning.com
***********************************************************************
World's source for "Captured Lightning" Lichtenberg Figure sculptures,
magnetically "shrunken" coins, and scarce/out of print technical books
***********************************************************************
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla