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

Re: More Coupling...



Original poster: "Kurt Schraner by way of Terry Fritz <twftesla-at-uswest-dot-net>" <k.schraner-at-datacomm.ch>

Hi Bart, all,

my (lower precision!) M and k measurements look about expected,
represented by acmi. The errors are in the same order of
magnitude, as by Mark's MandK. However, there is a sort of
nonrandom error distribution-difference between the 2 programs,
which I would yet like to resolve (not too dramatic...:-)).
Thanks for nice results! Other comments interspersed...

Tesla list wrote:
> 
> Original poster: "Barton B. Anderson by way of Terry Fritz
<twftesla-at-uswest-dot-net>" <tesla123-at-pacbell-dot-net>
> 
> Hi All,
> 
> I ran Kurt Schraner's large coil in acmi pulling his coupling
measurements from
> his Excel spreadsheet. Kurt's test identified coupling measurements at
various
> turns (1, 2, 3, 5, 7.2, and 9) at four different heights. Kurt, you may
notice
> the height position is 8.858 instead of -8.858 as in your spreadsheet
(this is
> just a height relationship changed to match acmi's way of thinking).

I'm aware of the different 'conventions' of what is + and - in
acmi and MandK.
> 
> Single turns are no good (lot's of reason's such as meter measuring a
very low
> inductance and I believe the program cannot produce the correct dimension
for a
> single turn because the turn grows around the primary in width and the
2nd turn
> is required to account for this growth via ID and OD - 2 or more turns
are just
> fine).

Agree completely, single turn is prone to high error, in
measurement as well as in simulation. However I observed with
MandK, the calculated mutual inductance M to be in much closer
agreement with the 'measured' M = U / ( 2pi . f . i ), than k,
taking the calculated values of Lp,Ls in k = M / sqrt(Lp*Ls).
When replacing Lp in the last formula by the measured Lp's, but
keeping the calculated M's and Ls, the k-values  were represented
much better. For Lp measurement I was using an LCR meter ELC131D
from Escort Instruments Corp. in Taiwan, which allowed the
compensation of the measuring leads L, before taking
measurements.

BTW, at the risk of talking obvious matter: In order to take the
coupling measurements by the 'best method', I'm using a
transformer "line-->low voltage" (230V-->12V in my case). The
advantages are:
1.) No safety risk with line voltage --> quick change of primary
tap.
2.) High current to TC primary, under good control, with small
adjustable resistor, giving secondary voltage readings, less
prone to noise errors.

> 
> I also put up Jon Rosentiel's data, Terry's, updated Marco's, my own, etc..
> Marco's however shows a larger error which I'm sure is due to dimensions
being
> as was the case with my own coil for a while.
> 
> Here's the data:
> 
>  <http://www.classictesla-dot-com/temp/k.html>http://www.classictesla-dot-com/temp/
> k.html
> 
> Take care,
> Bart
<snip>

Thank you,
Kurt Schraner