Arc / spark length versus frequency
james
I have been wondering how much the frequency of an arc affects its
length in air? Would a sine wave of say 50kHz arc further than one of
50Hz?
Can anyone suggest search keywords / formulae / resources that could
point me in the right direction?
Thanks and regards,
James
William J. Beaty
I've never seen articles about this.
I'd imagine that at fairly low frequencies the arc would start
to cool off between pulses. With Tesla coils and plasma globes
this gives you a multi-branched tree rather than a single hot
flame-arc.
Here are articles on spark gaps from Jim Lux' site, but I don't see
mention of any frequency effects:
http://home.earthlink.net/~jimlux/hv/hvmain.htm
http://home.earthlink.net/~jimlux/hv/paschen.htm
http://home.earthlink.net/~jimlux/hv/spherev.htm
http://home.earthlink.net/~jimlux/hv/sgcorr.htm
(((((((((((((((((( ( ( ( ( (O) ) ) ) ) )))))))))))))))))))
AMASCI.COM is up for a 2003 Webby award! Please vote in the
"People's voice awards" section of http://www.webbyawards.com
William J. Beaty SCIENCE HOBBYIST website
bi...@eskimo.com http://amasci.com
EE/programmer/sci-exhibits amateur science, hobby projects, sci fair
Seattle, WA 206-789-0775 unusual phenomena, tesla coils, weird sci
Bert Hickman
James,
I suspect you're really asking about how far a _spark_ will jump versus
frequency, since an arc, once formed, can be increased almost
indefinitely as long as sufficient current can be provided. It's not
uncommon for air break switches in electrical power systems to generate
arcs that are 10's or even 100 feet long on a circuit operating at
relatively low (100-300 kV) voltage.
There's surprisingly little direct information in the literature, and
for the most part what is there seems to show that the breakdown voltage
in a fixed gap will decrease by about (20%) as you go from DC to 60 kHz,
but it staying constant from 60 kHz - 425 kHz (Cobine, "Gaseous
Conductors", pp 184-186).
More recent research (Bazelyan and Raizer, "Spark Discharge", CRC Press)
proposes that an appropriately shaped rising voltage can cause a spark
to propagate virtually indefinitely. An incredible demonstration of this
effect can be seen in a photo of a 325 foot(!) spark that originated
from the top of a 5 million impulse generator (Siberian Institute for
Power Engineering). The waveform that produced this record holding spark
was specifically tailored to enhance propagation. Pictures of this spark
and an MPEG of some long three phase power arcs from a 345 kV air break
gang switch opening "hot" can be seen here:
http://205.243.100.155/frames/longarc.htm
In the case of a spark discharge Tesla Coil, all three effects
apparently come into play. Once the top terminal voltage achieves
initial breakdown, the rising voltage envelope enhances further spark
propagation, injecting space charge into the region just beyond the tips
of the visible discharge (the leader and streamers). The next voltage
reversal (from the RF inside the rising voltage envelope) causes the
"effective" electric field to be higher at the leader's tips, further
aiding in propagation. Displacement currents (charging and discharging
of the capacitance of the conductive leader channel) help keep the
leader channel hot, making it behave much like an arc.
Finally, repetitive "bangs" cause preferential breakdown in the hotter
(less dense) air left from the previous leader, so that leaders tend to
reappear in the same place from bang to bang. Future sparks have an
easier time than their predecessors until an overall balance is reached
between power in and power dissipated. This appears to be modeled quite
well by an empirical formula (from Tesla Coil researcher John Freau):
L = 1.7 *Sqrt(P)
where:
L = distance in inches
P = input power in Watts
The above winning combination of events results in sparks that are
considerably longer than sparks that from a purely sinusoidal source
having the same peak voltage. These differences are very evident when
comparing the output from a CW versus pulsed operation in a solid state
or vacuum tube driven Tesla Coil. This is also why a spark discharge
Tesla Coil with a peak output of 250-300 kV can easily create sparks
that are 6-8 feet long. However, there's still much work to be done
before these mechanisms are fully modeled and predictable.
Best regards,
-- Bert --
--
-----------------------------------------------------------
Unique Offerings - Scarce Technical Books, Coins Crushed by
Ultrastrong Magnetic Fields, and Captured Lightning Bolts!
All from Stoneridge Engineering - http://www.teslamania.com
-----------------------------------------------------------
james
observed, more to do with the spark length as I had failed to make the
differentiation in my topic
thanks for the help guys
James
Bert Hickman <bert.h...@aquila.net> wrote in message news:<3EA68E2...@aquila.net>...
RIAA/MPAA's Worst Nightmare
could look similar to a tesla spark (all my ladders I've made up to
100kv/240 mA look like wavy white flames rather than long curly things). Any
approximation of how much current per phase was behind the 345kv (by the way
when they say 345kv, is that phase-to-phase or phase-to ground? Wouldn't
765kv be over 1.3 million between phases?)
Bert Hickman
> That was a cool video of the 3-phase switch, I never thought 60-Hz power
> could look similar to a tesla spark (all my ladders I've made up to
> 100kv/240 mA look like wavy white flames rather than long curly things). Any
> approximation of how much current per phase was behind the 345kv (by the way
> when they say 345kv, is that phase-to-phase or phase-to ground? Wouldn't
> 765kv be over 1.3 million between phases?)
> "Bert Hickman" <bert.h...@aquila.net> wrote in message
> news:3EA68E2...@aquila.net...
From what I was told, the gang switch was most likely only switching a
comparatively small amount of reactive current (unloaded transformer or
unloaded transmission line) so the current was likely only in the 10-20
ampere range. True high current or fault current switching must be done
by Oil Circuit Breakers (OCB) or gas breakers in order to rapidly quench
the arcs.
The voltage on each of the three phases is 345 kV RMS referenced to
ground. Each phase is +/-120 degrees from other two phases, and the
phase-to-phase voltage is ~1.732 (square root of 3) times the phase to
ground voltage, or about 598 kV. The peak phase-to-phase voltage is 1.41
times the RMS voltage, or about 843 kV.
-- Bert --
--
-------------------------------------------------
Out-of-Print Physics and Engineering Books and
coins shrunk by ultrastrong electromagnetic fields!
Stoneridge Engineering: http://www.teslamania.com
-------------------------------------------------
Dan Barlow
> From what I was told, the gang switch was most likely only switching a
> comparatively small amount of reactive current (unloaded transformer or
> unloaded transmission line) so the current was likely only in the 10-20
> ampere range. True high current or fault current switching must be done
> by Oil Circuit Breakers (OCB) or gas breakers in order to rapidly quench
> the arcs.
>
> The voltage on each of the three phases is 345 kV RMS referenced to
> ground. Each phase is +/-120 degrees from other two phases, and the
> phase-to-phase voltage is ~1.732 (square root of 3) times the phase to
> ground voltage, or about 598 kV. The peak phase-to-phase voltage is 1.41
> times the RMS voltage, or about 843 kV.
>
> -- Bert --
Wow. Hey Bert, what do you think about these?
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26237&item=2526610867
-Dan Barlow
Bert Hickman
Dan,
The caps in the above auction are designed for use as DC filter caps,
and would be superb in this application. While they're not really
designed for high current energy discharge applications, they may work
for low duty cycle gas discharge applications (flashtubes), where the
discharge is resistive/overdamped and the peak current is not extremely
high.
True low-inductance energy discharge caps don't have the projecting
"Frankenstein" insulators of the above caps. Instead, they use broad,
flat insulators to minimize inductance while still providing adequate
creepage distance. Compare the insulators on the above caps with the low
inductance pulse discharge caps in the seller's other auction:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26237&item=2526608313
BTW, I haven't had good success when using [misapplying!] similar
3OF1600 series, 54 uF, 15 kV GE HV caps in a high current discharge ~50
kA/cap magneforming application. The internal construction of these caps
included a hairpin loop in the bus connecting one end of all the
capacitor rolls to one of the HV bushings. During high current pulse
discharges, magnetic forces in the hairpin loop would start tearing the
bus away from ends of the capacitor rolls. After relatively few
discharges, this led to decreased capacitance and finally catastrophic
capacitor failure (complete with an internal explosion, case rupture,
and oil and capacitor innards dumped onto the floor). Your mileage may
vary depending upon your application... :^)
The high current energy discharge caps in the seller's other auction
would be much better suited for pulsed power applications. However, the
seller is probably optimistic in estimating the weight for these - it's
probably closer to 250-300 pounds/cap. And at >31 kj/cap, these are some
SERIOUS energy discharge caps.
Dan Barlow
> The caps in the above auction are designed for use as DC filter caps,
> and would be superb in this application. While they're not really
> designed for high current energy discharge applications, they may work
> for low duty cycle gas discharge applications (flashtubes), where the
> discharge is resistive/overdamped and the peak current is not extremely
> high.
OK, so chalk this one up for copper vapor laser or relatively high
inductance magnetic cannon. Might make a good primary supply cap
with smaller low ESR caps handling the repetitive discharge pulses?
> True low-inductance energy discharge caps don't have the projecting
> "Frankenstein" insulators of the above caps. Instead, they use broad,
> flat insulators to minimize inductance while still providing adequate
> creepage distance. Compare the insulators on the above caps with the low
> inductance pulse discharge caps in the seller's other auction:
> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=26237&item=2526608313
> The high current energy discharge caps in the seller's other auction
> would be much better suited for pulsed power applications. However, the
> seller is probably optimistic in estimating the weight for these - it's
> probably closer to 250-300 pounds/cap. And at >31 kj/cap, these are some
> SERIOUS energy discharge caps.
Thanks very much for your advice! :) I am slightly pondering one for
the "dolly mounted can shrinker" value. However the shipping is
probably more than the bid price. :P
-Dan Barlow