[Prev][Next][Index][Thread]
break-rate/power tests over weekend
All,
Over the weekend, I decided to tackle the input power question again
for various break-rates. My questions were; is my wattmeter accurate
at high break-rates, and how efficient are the high break rates? To
find the answer, I first "calibrated" a thermocouple ammeter using a
resistive load instead of a Tesla coil.
I used a household toaster for the resistive load, and applied 1080
watts as measured with my wattmeter. I also noted that a regular
ammeter read 4.4 amps, (240 volts). The thermocouple ammeter
(TCA) read 55. This number 55 means nothing in itself, it's simply
a reading on the meter. The thermocouple meter is also non-linear.
The thermocouple meter can read any kind of current waveform;
distorted, high frequency, etc.
Next I ran the Tesla coil at a break-rate of 400 BPS (approx) and
turned up the power until the TCA read 55, this means the
true input current is 4.4 amps (based on chart below showing the
toaster tests). The regular ammeter read 4.3 amps, so it was very
close. Next I read the wattmeter which read 1000 watts. This too
was very close to the reading with the toaster.
The conclusion is that the regular ammeter seems to read OK, and
the wattmeter seems to read OK at 400 BPS in this Tesla coil,
based on comparisons with a resistive load. If the power factor was
bad, (which might confuse the issue), the wattmeter would pick it up.
But it seems the power factor is pretty good. I figure that this
method of cross checking the meters, and calibrating first with
a resistive load should give a reasonable assurance of the general
accuracy of the measurements. I'm not looking for 1% accuracy,
I'm satisfied with 10%. My original concern was that the wattmeter
might be reading with a 50% or 100% error at the high break-rates,
but that doesn't seem to be the case.
I also did a number of other runs at different break-rates, different
calibrations with the toaster load, etc., as shown in the following
chart:
BPS Watts Amps TCA ballast set toroid spark inches
(est.)
120 600 3 25 3 10" 28
120 600 4 60 5 10 28
400 1000 4.3 55 5 10 30
800 1000 5 70-80 5 10 15
T 600 2.5 20
T 1000 4 48
T 740 2.8 25
T 1080 4.4 55
T 1220 5.1 70
120 620 3 20"
42"
400 1000 5 20
44" (?)
Notes: BPS at 400 and 800 are approx., watts read on wattmeter,
amps read on regular ammeter, TCA stands for thermocouple
ammeter. Prior tests showed that the wattmeter is accurate at 120
BPS based on TCA comparisons, T indicates toaster load (no TC).
Most likely the wattmeter is reasonably accurate at 800BPS also,
and is showing a poorer power factor.
I used my 42" spark TC, but I used a 3" by 10" spun smooth toroid
instead of the usual 5" by 20" dryer duct toroid, because I did some
other tests of spark coalescence etc. The power was deliberately
held back in the tests discussed in this posting.
It appears that my TC is actually drawing high power levels at the
high break-rates, without any real increase in spark length*. In other
tests, I ran the coil at higher power levels. At 600 watts at 120
BPS, sparks were 32" long, and multiple streamers formed. At
800 watts at 400 BPS, sparks were still 32" long, but were much
brighter. At 1000 watts at 800 BPS, the sparks coalesced into one
very bright streamer, of the same length. I could tell that the
capacitors were charging to a much lower voltage at the higher
break-rates, because the variac had to be turned up higher to
make the gaps fire. However, at 120 BPS, at very low power,
multiple streamers were still produced, which suggests that the
coalescence resulted more from the high break-rate than from the
lower capacitor voltage. *It should be noted that the average spark
length was shorter at 120 BPS, than at the higher break-rates, but
the sparks occasionally reached out and hit 32" despite the much
lower input power, and multiple streamers. So there is more
variability to the spark length at low break-rates; the sparks
occasionally grow to long lengths. Also, the brightness of the high
break-rate sparks makes them appear longer than they are, in a
sense. Both low break-rate and high break-rate sparks struck the
ground with equal frequency, despite the longer average length of
the high break-rate sparks. Of course the high break-rate sparks
required a lot more input power.
It is possible that the losses in the ballast or transformer are much
higher at the high break-rates although the 1500 watt rated potential
transformer should not have a great problem charging the small
.007uF tank cap. It has been suggested to me that larger transformers
will show lower relative losses at the higher break-rates. Thus, larger
systems should better tolerate the higher break-rates, but for small
systems, up to two kilowatts or so perhaps, I don't see how high
break-rates can offer good "efficiency". It seems to me, that in a
small system using traditional power supplies, longest sparks using
minimum input power will always be achieved at low break rates. I
don't know what will happen for larger coils, but I wouldn't be surprised
if low break-rates were more efficient also.
But if the brightest possible sparks are desired (especially with a
small toroid), then a high break-rateshould be used, but this will
demand more power. Most TC's will give longer sparks of course,
if the break-rate is increased, but the power input will rise greatly.
To obtain the best results from a low break-rate system, a larger
toroid may be needed than a comparable high break-rate system.
This will help keep the sparks brighter and longer despite the low
power input.
Although a high break-rate may create a bright, coalescent streamer
using a small toroid, the sparks will still be a lot shorter than when
using a larger properly sized toroid. At 120 BPS, I get a 32" spark
at 800 watts with the small 10" toroid, but with a 20" toroid, I get 42"
sparks at only 620 watts at the same break-rate, and the sparks are
much brighter. At 400 BPS, I get a 32" spark at 1000 watts with the
10" toroid, but with a 20" toroid, I get 44" sparks at 1000 watts or so.
It is interesting to note that using the small toroid, the longest spark
length did not increase at high break rates, but with the larger toroid,
the spark length did increase by a couple of inches at high break-
rates. Maybe the small toroid does not store enough energy to
effectively feed the streamer even at high break-rates, whereas
a large toroid provides enough stored energy to the streamer to allow
it to benefit from a high break-rate, although it only increases by a
couple of inches, despite the much higher input power. If this is
true, then larger coils with large toroids might benefit more from
higher break-rates than smaller coils with smaller toroids. This
might be the explanation for the good performance of some large
coils at high break-rates rather than the theory of lower transformer
losses in large coils. This is getting speculative here though. More
tests are needed.
The general principles mentioned in this post should apply
to mostly all coils using standard power supplies, although some
variation might occur at higher powers, or especially with different
capacitor sizes relative to the input power.
If anyone wishes to modify their TC for 120 BPS operation, it's
possible that a larger tank capacitor and toroid, and different ballast
setting will be required for best results. The components must be
carefully "matched for synergy" at 120 BPS, because there will no
longer be the option to use changes in break-rate to find the coil's
"sweet spot". Rather than adjusting break-rate to match the coil,
the coil will need to be matched to the break-rate at 120 BPS.
This may explain the reports of poor results with 120 BPS operation
that are sometimes heard.
Comments are welcome,
John Freau