Re: Computer data
Sent: Saturday, August 30, 1997 12:05 PM
To: Tesla List
Subject: Re: Computer data
To: Greg Leyh
Even without CW the tank current will dip as noted on the primary side
ammeters. We did this experiment in Calif. in 1981 with Bill Wysock
running one of
his large coils -- I believe it was a model 10. The coil output was around
32 ft. long spark. The power supply was set up with a continuously
variable series inductor and a standard parallel variac stack. As we ran
the power up thru the voltage variac in small increments we made small
adjustments to the series inductor. At almost any setting above 50% power
level (the range we are most interested in) as one increased the variac
(voltage) the current would begin increasing as well. At some point as the
power factor begins to correct back to unity (phase meter across input
line) the primary 240 vac side of the power supply will indicate a 5-10%
drop in the primary current to the power transformer. This is what I refer
to as the most efficient point of operation. The current actually "dips"
just like a plate milliammeter in a ham radio transmitter. It's not as
dramatic, but the effect is quite evident. From this point the voltage
variac is increased more and the primary current once again begins to
climb. The longest sparks do occur around 10-20% above the point at which
the primary current dip point however this is a region of what I call
"overdrive". This point trades off efficiency for more power into the sec
corona load and hence longer sparks. The spark gap however sounds the
"happiest" at the point of max. efficiency, ie, the dipped current area.
We use these techniques in running our smaller 7 kva systems as well.
After all our initial experiments are done with a variable inductor then we
carefully measure the correct value of inductance for most efficient
operation and wind future inductors to this value. On larger systems we
sometimes add a few taps (plus or minus 20%) to allow adjustment as the
voltage across the country varies from 208 VAC on east coast to 240 VAC
midwest, 208 VAC in south, and 220-240 VAC again in the west. I might
suggest you incorporate several taps into your limiting inductor system and
you will be able to "tune" your power supply as well as your resonance
points. A continuously variable inductor demonstrates these principles in
dramatic effect but this would be a lot of work to construct for your very
large system. A phase shift indicator illustrating where unity is while
your system is running is a very valuable addition to your instrumentation.
Regarding the data I posted .... does it agree with any of your measured
values of current input vs spark length output for your old system with the
0.18 MFD capacitor???
The algorithm we use is a series of plotted points from various coils
sytems we have run in the past. At long spark lengths (max was 78 ft) we
recorded 1.9 kva required per ft of spark length. At lengths of 30-40 ft
the power level was approx 1.1 to 1.2 kva per ft of sec spark.
Hope this data helps out -- please advise on the data I submitted vs your
> From: Tesla List <tesla-at-pupman-dot-com>
> To: 'Tesla List' <tesla-at-poodle.pupman-dot-com>
> Subject: RE: Computer data
> Date: Friday,August 29,1997 1:36 AM
> From: Greg Leyh[SMTP:lod-at-pacbell-dot-net]
> Sent: Thursday, August 28, 1997 5:19 PM
> To: Tesla List
> Subject: Re: Computer data
> DR.RESONANCE wrote:
> > Hi Greg:
> > Ran some of your previous coil design data thru our computer design
> > program, "Resonance Master", and it generated the following data:
> > assuming a pri cap size of 0.18 MFD, then:
> > at 90 amps input x = 21 ft. long sec. spark this would be min
> > value for this size cap
> > at 108 amps input x = 25.2 ft. long sec. spark this would be most
> > efficient operation point, ie, tank plate current dipped
> > at 130 amps input x = 30.3 ft. long sec. spark overdriving the
> > slightly for max spark output
> This coil is not CW, and therefore does not have a tank plate current to
> What are the ampere ratings shown above? Is that the wallplug current?
> So far in the low power tests (shown in the JPEG's on Chip's site), here
> what I have actually measured:
> At 45 amps input (480V 3-Phase) x = 25 to 32 ft. long sec. spark
> > The above values assume a large top load and all is in tune. Power
> > reactance matches cap size, pri in correct tune with sec system, proper
> > coeff. coupling (0.18-0.22), and good spark gap quenching.
> > I was curious if these values are in agreement with your observed
> > parameters. If in agreement I would be happy to do a runup of your new
> > system to see if your new numbers agree with our program output.
> Yes, I would indeed! The full power (120kW) test will occur in about 8
> Cpri will be 0.27uF
> Zpri will be 15 ohms
> k will be 0.22
> How does your algorithm relate output power to spark length?