[Prev][Next][Index][Thread]
NEW break-rate/power tests
All,
I did some tests at a higher power level. First I installed a .0148uF
cap with the 120 bps sync gap, then compared with the 400 bps
rotary gap. At 1370 watts, the spark reached 54" at 120 bps, but
at 400 bps, the spark could not reach 54" even at 2000 watts. But
it came close. However, the spark was brighter and more
flamelike at the high break-rate, and wanted to go upwards towards
the ceiling. The ceiling is 36" above the top of the toroid.
The sparks looked completely different at 120 bps.
At this low break rate, the sparks were not as bright, but
they tended to surge and grow, and emitted from different spots
around the toroid. The sparks floated around slowly. The high
break-rate sparks tended to be all about the same length and,
seemed to want to remain together like a giant
blow-torch or flame, and reminded me a little of tube coil sparks
(high powered tube coil sparks of course). So all of this was the
same as at the lower power level, but on a larger scale. I tried a
higher break rate of about 600 bps, but the sparks got shorter,
due to sagging voltage in the power supply. Of course,
in any case, the voltage on the caps was always lower at the high
break-rates, because the system was unable to benefit from
resonant charging voltage build up.
Next I replaced the 1500 watt 14.4kV potential transformer with (2)
1500 watt 7.2kV potential transformers in series for a "stiffer" power
supply. With this set-up I got 54" sparks at 1000 watts at 120 bps.
At 400 bps, the sparks were very bright, but never reached 54" but
they hit the ceiling a lot and seemed more powerful than with the
one transformer.
Then I replaced the 20" toroid with a 6" by 26" dryer duct toroid,
This arrangement gave 58" sparks at 1050 watts, this I believe is
a record for efficiency for my coils, based on my formula:
spark length (inches) = 1.7*sqrt input power, wallplug (watts).
The formula predicts 55", but this coil gave 58" sparks. I attribute
the greater efficiency to the use of two transformers, which reduces
the losses.
Note; those with NST powered systems are cautioned to never use
the NST power rating written on the NST in my formula above.
Many NST systems use resonant charging and actually draw double
or more the NST rated power input. The only way to know the input
power of any Tesla coil is to measure it with a suitable wattmeter.
If a wattmeter is not available, at least use an ammeter, and multiply
by the input voltage to obtain input VA (volt-amps). This will not
correct for the power factor error, but will give a ballpark
measurement. In my coils, input current distortions and/or high
frequency components did not seem to affect the meter accuracy
by much. My formula above uses true input wallplug power. I
measure the power before any variacs, ballasts, etc.
I then tried using 400 bps, but the sparks jumped to the ceiling too
often and would not reach 58", so this made it impossible to
properly measure them accurately, although I did get a general
idea. At 600 bps, the sparks did not get shorter, due to the use
of the stiffer power supply. The sparks seemed bright and strong
at these higher break-rates.
I used the same wattmeter as my tests from last weekend, which
should give reasonably accurate results.
Table of results:
Cap (uF) bps watts toroid (inches) spark (inches)
.007 120 620 5 x 20 42
.007 400 1000 5 x 20 43
.0148 120 1370 5 x 20 54
.0148 400 2200 5 x 20 less than 54
.0148 600 2200 5 x 20 weaker
.0148 400 3000 5 x 20 59 one time
.0148 120 1000 (2) xfrmers 5 x 20 54
.0148 120 1050 (2) xfrmers 6 x 26 58
.0148 400 2000 (2) xfmers 6 x 26 went to ceiling
but approached 58
.0148 600 2000 (2) xfrmers 6 x 26 not weaker
.0148 400 1000 (2) xfrmers 6 x 26 36 (low cap Volts)
.0148 400 1050 (2) xfrmers 5 x 20 41
.007 400 1000 (1) xfrmer 3 x 10 30
In the second to last test above, I of course had to keep the variac
turned down somewhat to hold the power level at 1050 watts, so
the bang size was probably the same as in the last test, so it's
really the larger toroid that is increasing the spark length in the
second to last test. It's hard to know how much voltage was on
the caps in any of these tests.
For a given input power, the low break-rates seem to outperform
the high break-rates by a significant margin in all these tests, as
far as input power vs. longest spark length is concerned.
I found that at high break-rates, I had to use a much smaller
ballast inductance than at 120 bps. At high break-rates, I used
about 2mH to 10mH. At 120 bps, I used about 15mH to 25mH.
The ballast was adjusted for best TC efficiency, and smoothest
operation in all cases.
High break-rate systems have greater losses due to:
1. Gap firing shorts transformer more times per second.
2. Each gap firing wastes more power due to transformer shorting
due to ballast adjustment, and line voltage at gap firing times.
3. Cap voltage is lower due to less resonant build-up which
causes greater relative tank resistive heating losses.
Also, the energy stored in toroid (per bang) is lower.
For a high break-rate system to outperform, it has to make up for
all the above losses and then some more by using ion channel
spark-growth----can it do it? Obviously, not in coils of the size I'm
testing (up to 2000 watts). I do not know how much of the high
break-rate under-performance results from losses, and how much
may be inherent to high break-rates (if any). I would not be
surprised if the losses at high break-rates are much lower
(relatively speaking) in large coils compared to small ones.
It is possible that by using 120 bps, and scaling up a TC in size,
that the spark lengths might outperform the predicted lengths
given by my formula. Let's see....around .03uF should be
good for a 100" spark, at 120 bps, at 2800 watts, using a 5kVA
rated 14.4kV pole pig for low loss. This is assuming the spark
length can outperform my formula. A 7.5" x 35" toroid will be
needed. This might be too optimistic but would be interesting to
try.
John Freau