[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Current Limiting and Impedence/Phase angle of an 18khz argon discharge?



Original poster: Harvey Norris <harvich@xxxxxxxxx>


--- Tesla list <tesla@xxxxxxxxxx> wrote: > Original poster: "Gerald Reynolds"

> >Instead we must break down the Z's into R, and X
> >
> >Z1^2 = R1^2 + X1^2
> >Z2^2 = R2^2 + X2^2
> >
> >And must assign the correct sign for X in case both
> inductive and
> >capacitive reactance is present.
> >
> >Then
> >
> >Z^2 = (R1 + R2 + ...)^2 + (X1 + X2 + ...)^2
>
> YES, this is correct
>
>
> >I'm not sure this solves my problem, because in my
> case the R's were
> >near zero and the X's were inductive.  I will go
> back and recalculate.

Hi Gerald, I had a similar problem where two
impedances were added in series, and I was wondering
if the following data would provide enough information
to determine the acting phase angle of a gas
discharge, which in this case was a 2 ft argon tube.
First let me digress a bit about what I have seen with
gas disharges. We all know that a gas discharge must
be current limited or ballasted. But why is this
specifically true, shouldnt the gas discharge itself
have a high resistance? Apparently not initially upon
preliminary firing where ionization first occurs, it
may have to do with the so called negative resistance
portion of operation. I play around with AC
alternators a bit, and I have a pole pig. I decided to
put a 2 ft neon as a solitary secondary output on the
pole pig, and power it by an alternator. I put an
amperage meter on the pole pig primary and used 480 hz
available from the alternator to go through the pole
pig primary. What happened next showed the need for
ballasting. The neon breifly lit, and then everything
went dead. Inspection of the problem showed that the
primary must have experienced a large amperage surge,
as the fuse in the amperage meter had blown. It would
require at least 10 amps to do this. So next I limited
the neon current by putting it in series with about 2
nf available from a plexiglass plate capacity. This
doesnt sound like much but recall that at 480 hz a
capacity will have 8 times less reactance then it
would have at 60 hz. Having a variac control of the
alternator voltage output I gradually increased the
input voltage until the neon lit. Then I measured both
the voltage across the capacity, and the voltage
across the neon, but unfortunately I dont recall
measuring the voltage across both elements in series.
What I found was that at this lowest ionization level
where the bulb first fires, the voltage across both
elements were almost equal, with the neon disharge
having a slightly higher voltage. I think it will be
important to repeat this procedure and measure the
total voltage across both elements because of the
following thought. The neon contains impedance, but
what kind of impedance? I would think that it should
be inductive reactance, not capacitive reactance. And
since I was using capacitive reactance to limit the
current, it might seem possible that since intially
the voltages were almost equal, that perhaps the
reactances might be cancelling, as occurs in series
resonance. If this were the case the voltage across
both elements at the pole pig secondary would be less
then opposing voltages inside the circuit across each
element, and then we could even find a q ratio
according to that ratio of inside vs outside voltages.
Now what I did next was to increase the intensity of
the neon disharge by increasing the primary voltage of
the pole pig, since the DC field of the alternator was
regulated by means of variac control thus the AC
voltage output of the alternator connected to the pole
pig primary could be guadually increased. When this
was done the voltage across the limiting capacity went
way down, and then most of the voltage was across the
neon. What this seems to tell me is that if the neon
has a phase angle of operation, its phase angle must
change according to its volume of current, or
intensity of discharge.
    Now for something completely different which comes
to my problem. I have some small 4 inch neon bulbs
that fire at smaller voltages, so I could actually
scope out the neon discharge. I found that the solid
state NST  advertised as 20,000 hz output would fire
the bulb around 200 volts, where at 60 hz it would
fire at 260 volts. The scoping however showed that the
AC signal was not sinusoidal but triangular shaped,
but more importantly the frequency showed itself as
18,000 hz, not 20,000 hz. I do not know if this was
just an idiosyncrasy of the scopes internal capacity
reacting with measuring a high input frequency or
what, but for the purposes here I will use 18khz as
the input frequency. Incidentally the higher input
frequency has quite a bit to do with errors
encountered in measuring instruments, which are
probably designed to give accurate answers at the
conventional 60 hz, not something as high as 18 khz.
An analogue needle voltage meter will give a different
answer for the voltage across the 4 inch neon on each
different scale setting. It will only read ~ 50 volts
if set to the smallest scale, but it reads accurate at
the highest scale showing 200 volts, this is known
because at that scale both the scope and meter agree
on the voltage across the bulb. Another factor here in
the problem is that because of the higher freq, and
the triangular AC form showed with the signal, no
amperage meter was placed on the output. If I knew the
amperage changes involvolved with this problem, the
answer might be easily found, but because of these
extraneous matters, I wouldnt trust an amperage meter
on a gas discharge circuit at 18 khz, so I didnt use
one. So the question becomes can I determine what
phase angle exists on a 18 khz 2 ft argon bulbs
discharge, by adding a known inductive reactance in
series with the discharge?  For the known inductive
reactance I used a 20.5 mh solenoidal winding of 1500
ft of 14 gauge wire on a sonotube form, (thus R is so
miniscule in comparison to Z that we can call this a
predominantly inductive reactance, and thus R is
ignored) Here is the data and speculations I had made
from records, with further comments contained in
brackets[],

"Since we are given the information that 590 volts is
enabling the solitary argon discharge with a 120 ma
input on primary, but we cannot actually measure the
secondary amperage for fear of meter damage, or
totally off whack readings because of that high
frequency now shown to be 18,000 hz on secondary end,
we can still establish some comparisons by calculating
what the 20.5 mh coils impedance would be at 18 khz.
This would be 2317 ohms, by the inductive reactance
formula X(L) = 2 pi *F* L. Now we have no idea what
the
impedance of the argon discharge would be at this
frequency, but it is accepted that it consists of
primarily a phase angle of inductive reactance with a
small amount of resistance in that phase angle,
whereas the impedance of the 20.5 mh coil is primarily
all inductive reactance X(L) of 2317 ohms. We also
know that when the coil was placed in series with the
bulb, the voltage across the bulb went from 590 volts
to 515 volts, with an extra 140 volts appearing across
the coils inductive reactance, so the voltage dropped
to 87.3 % of the former level without anything except
the amperage changing on the primary side. Now the
primary amperage dropped from 120 ma to 90 ma
when more impedance was added to the secondary side.
It dropped to 3/4 of the former level. This is then
equivalent to stating that for the unknown impedance
of the bulb to have more impedance to be added
in series, if a linear primary/secondary amperage
relationship exists for the solid state transformer,
the addition of the known impedance of 2317 ohms
to the secondary side, for the amperage to drop to 3/4
of the former level, 4/3 more total impedance must be
present on the secondary side.

Now since we have speculated that 4/3 more additional
impedance has been added to the secondary side, than
what the bulb itself delivers, if both the bulb and
the coil were ~ all inductive reactance, this would
mean that the voltage across the bulb should
drop to about three quarters of its former level, or
75%. Then the distribution of series voltages would
also sum to equal the first case for a total of 590
volts. But this did not occur, the voltages
in series sum to a total higher value, and the voltage
drop across the bulb was instead 87.3% of its former
value, or about 7/8 ths of what it was in solitary
operation. [Note; this fact in of itself shows that
the argon discharge MUST NOT be all inductive
reactance and that it should have a phase angle when
driven at 18 khz.  Furthermore what we are trying to
determine here is that the gas discharge does HAVE a
real resistance coordinate, despite some
misunderstandings concerning what is called the
negative resistance portion of operation. It is my
speculation that this portion of operation only occurs
near the point of initial ionization, and that as the
intensity of the discharge increases this ALSO changes
its operational phase angle. A simple proof of that
would be made by driving the solid state NST at the
recomended 240 volts, instead of household 120 AC
volts as was done here, and then repeating the voltage
mesurements across each component to see if this same
ratio exists. If the ratio remained the same the phase
angle would not have changed, but I suspect that would
not occur.] Without knowing what the actual secondary
amperage was, this indeed poses a confusing problem to
find "what phase angle" the gas discharge itself
consists of..

Let us also speculate that with the coil and bulb in
series, the total impedance from the unknown phase
angle consists of 4 * 2317 ohms or 9268 ohms. The coil
contributes little or no resistance for the X
coodinant, but the gas discharge does. [proven by
voltage distributions  adding to a higher total then
what exists for the bulb alone] By the pythagorean
theorem we then speculate that
9268 = sq rt{ x^2 + y^2} The Y portion of the equation
consists of Y = { X(L) as coil= 2317 ohms + X(L) as
gas discharge = ?} R for the gas discharge as X
coordinant is also unknown, thus with two unknowns
here, the equation still presents a great problem.

Truly baffling problem here. Perhaps it is only
appearing that way because of how the problem itself
is being presented, and we are being deceived by a
sort of mirage. Let us form a phase angle so that
the higher impedance of the gas disharge WILL have 75%
of its Y coordinate value as the total inductive
reactance on the Y axis. Let us also use 100 volts for
this example instead of 590 volts, to make the math
simpler. We then know that the vector magnitude,
(determined by pythagorean theorem) will be 87.3, and
its y coordinate will be 75. The sin function is the y
coordinant divided by the vector magnitude or the sin
of this unknown angle will be 75/87.3 = .859 radians.
Dividing this by 2pi= 6.28 radians/360 degrees we
obtain a phase angle of 49.25 degrees, where the
cosine of this angle will be .653.

Now let us return to the original problem then where
590 volts are across the bulb alone. We theorize that
the resistance x coordinant will consist of 65.3 % of
the total impedance of the bulb, and its inductive
reactance will consist 85.9 % of the bulbs impedance,
with all of this based on the assumption that the
discharge itself has a 49.25 degree phase angle. Now
we will add 1/3 more inductive reactance added by
putting the coil in series with the bulb. This
will change the total resultant phase angle.

I think [perhaps] I have made some serious
miscalculations here, and have approached this problem
in a totally incorrect fashion. I think the whole
problem needs to be back engineered so that the total
RESULTANT phase angle would be that 49.25 degrees, and
then by then knowing that information, the actual
phase angle of the bulb discharge itself might be
known. So before this problem drives me nuts, I'm just
going to put it on the back burner for future
consideration.  If anyone can fathom an answer from
this information, please post it, as I am rusty
concerning phase angle mathemathematics.

The problem breifly reformulated is this. 590 volts
appears across a bulb, at a frequency of 18,000 hz. We
do not know the amperage consumption or the phase
angle of that impedance load. Then a known
purely inductive reactance of 2317 ohms made at 18,000
hz is added in series, whereby we then speculate that
the unknown amperage consumption on the secondary has
decreased to 3/4 of its former level, indicating the
addition of 4/3 more impedance on the secondary, with
a new unknown phase angle made by that addition in
series for all practical purposes, a purely inductive
reactance. These facts are predicted from observing
the primary amperage consumption differences of 120 ma
for bulb alone, and 90 ma for coil in series with bulb
on secondary, on the solid state 18 khz transformer
secondary, with primary amperage differences being
known, but not the secondaries. Now the voltage across
the bulb decreases to 7/8 of its former value, and not
the lesser 3/4 of its former value, which would occur
if both impedances were predominantly inductive. The
new voltage distribution from the former 590 volts
total now becomes 515 across the bulb and 140 volts
across the pure inductive reactance of the coil. What,
if any phase angle considerations can be gleaned from
the above information? A phase angle of 45 degrees
would mean that both the X(L) inductive reactance, and
R resistance would be equal, and for conventional
frequency determinations at 60 hz would actually
involve a fairly huge inductance. But here because the
frequency is already so high we can speculate that the
inductive reactance from the gas discharge itself
would be high, as that current going through the bulb
does also produce a magnetic field that causes a
higher inductive reactance to be present as a
component of its resultant phase angle."

Postnote;
I thought I would also mention the following use of
the solid state NST. They are very cheap in cost
compared to a signal generator, I think I only paid
20-25 dollars for the ones I purchased. One might
wonder why I would go about putting a coil in series
with a gas bulb in the first place, driven by a high
frequency solid state NST. Richard Hull notes the
following;
"If we place a quantity of electrical energy into the
coil and do it quickly enough, the coil will ring at
its natural resonant frequency, much like a bell.
Voltage nodes and peaks will appear along the coil.
If the coil is floating in free space, it will tend to
oscillate at its natural 1/2-wavelength resonant
frequency, and each end of the coil will exhibit a
voltage peak while a voltage nodal point will
exist in the exact center of the coil. If, however, we
ground the base of the coil, this is a forced nodal
point and the coil will oscillate at its natural
1/4-wave resonant frequency. "

The key here appears to be doing it "quickly enough".
The 20khz solid state NST can do this quickly enough
so that the scoping of the coil in series with the gas
discharge will show harmonics riding on the source
frequency, and then this becomes a method to scope out
the natural resonant frequency of a solenoid.

   In fact it would seem that when we scope out that
coil in series with the gas bulb, we can find higher
harmonics riding on the AC signal, and by finding the
higher frequency riding on the lower frequency of the
20khz solid state input, this is a signpost of that
coils natural resonant frequency, so we should be able
to use the solid state NST as an indicator of a coils
natural resonant frequency. In the 20.5 mh solenoidal
coil I used I applied Ed Harris' formula found at
<http://www.pupman.com/listarchives/1996/june/msg00227.html>
When I scoped out the cited coil in series with the
argon tube driven by the solid state NST the harmonics
gave about the same answer as Harris' formula gives.
This scoping and other solid state NST work is shown
at
http://groups.yahoo.com/group/teslafy/message/517
Further references on this same phase angle problem
http://groups.yahoo.com/group/teslafy/message/1000

I also tried this with my 26 gauge secondary having a
resonant freq ~ 330,000 hz and found that some of the
scope forms were uncohered or rapidly moving across
the screen. However by taking a quick digital picture
of the scope form it showed a small slice in time of
an AC oscillation that was identical in answer to the
resonant frequency of the secondary.
Sincerely HDN



Tesla Research Group; Pioneering the Applications of Interphasal Resonances http://groups.yahoo.com/group/teslafy/