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

Re: Voltage profile on the secondary

Hi Bob,

I measured the profiles for may bare 10.25 inch dia. 30 inch long secondary
(75.4mH  1000.5 turns) at:

and the top load case at:

with the raw data at:

I took a tiny voltage probe feed to a scope to measure true RMS voltage.  I
tuned a constant voltage source (low-Z sine wave) and adjusted the
frequency slightly to account for the slight loading of the probe and ran
it along the length of the coil taking the readings carefully at each
point.  I had no primary, quench, or anything else involved.  Just the
secondary coil, sine generator, and scope..

I compare the secondary voltages from those readings to a sine distribution
and an l^e distribution in the graph at:

I use these measured distributions to come up with the voltage distribution
functions in E-Tesla5 which I now go in and average at one point to get
greater accuracy.  The resulting distribution in that program for a
terminated coil looks like:

One will notice that the distributions I got are not sine functions!  They
are power functions.  I used l^e but it may really be l^2 (or a torque
summation).  I think the voltage rise is much more like a bent ruler as
Malcolm suggests and not like a sine wave.  However, the harmonics are much
more sine like in their distribution but they may also actually be a power

A bent ruler is a summation of the masses and gravitational forces along
it's length and the resulting torques and the spring constant.  Very
similar to a sine type of oscillation.  However, if one compares the
displacement along the wire, it is a bending function not a sine function.

Although the secondary coil in E-Tesla5 is affected by the surrounding
primary, walls, top terminal, etc.,  they are all set to zero volts and
assumed to be grounded.  E-Tesla does not assume any frequency , quench,
arc / streamer loading, or mutual coupling effects.  I just set the
secondary voltage distribution to about what my measurements showed and let
it crunch away.  My new correction step in version 5.50 seems to do wonders
smoothing the "rough edges" and errors in my guesses and pushes the
accuracy to high levels.

So, I don't think the voltage distribution is a sine wave.  I think it
looks more like a steel ruler bending under gravity in the static case or
in the case of it springing back and forth as shown at:

I would submit that your work is independently finding the same voltage
profile I have found in my studies as has Malcolm.  Not a sine wave at all
but much more like "Malcolm's Ruler".  A steel ruler, clamped in a vise at
the base and pushed at the top with one's finger, appears to flex with the
same profile that the voltage distributes along the secondary.  Of course,
I think you are the first one to figure all this out using 'real' circuit
theory, T-line, lumped, etc. stuff instead of equation matching
measurements (like I do).

I will be very interested to see what all you have come up with.  The fact
that you are seeing "significantly less than a 1/4 sine wave at the start"
only confirms what I have 'suspected' is the case.  Perhaps that can now be
proven and much better equations found!!

I was looking at trying to use equations that predict a long steel spring
wire bending under gravity to better refine the profile setup in E-Tesla5
but I have not gotten very far at all yet.  The "C" versions and all that
took priority.  However, that voltage distribution is the only "loose
screw" in the program.  If that could be found exactly, the program would
then be theoretically perfect...



At 07:34 PM 6/10/00 -0400, you wrote:
>Hi all and Terry in particular,
>In the process of various attempts to extract intrinsic C from various
>equation for F of the secondary a related problem was giving me trouble.
>ie the voltage profile on the secondary.
>The problem is that it significantly less than a 1/4 sine wave at the start.
>Then I realized that the profile  is dependent on whether quench has
>and in the none quench case its also dependent on the current in the
>So my question to Terry is during which part of the operating cycle  did
>you determine the profile ie pre quench or post quench or can you confirm or
>deny the profile changes.
>I suspect you measured the peak voltage with no break out corresponding to
>the first primary notch but the quench may have been at the next notch.
>Regards Bob  (Robert Jones)