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

Re: More Coupling...



Original poster: "by way of Terry Fritz <twftesla-at-uswest-dot-net>" <paul-at-abelian.demon.co.uk>

Bert Hickman <bert.hickman-at-aquila-dot-net> wrote:

> ACMI Results versus measured values:
> ===================================
> 
>                    ACMI:        Prev. Measured:          Error:
>                    ====         ==============           ======
> Pri Self Ind       168   uH     161.8 (-at- 15.625 Turns)      -
>                                 Est 170 uH at 16 Turns    -1.2%
> Sec Self Ind        74.0 mH      73.5 uH                  +0.6%
> Pri-Sec Mut Ind    708.8 uH     not directly measured       -
> Coupling Coeff.    0.201         0.209                    -3.8%
> 
> The results appear to be within the estimated accuracy of my
> measurements.

Thanks for the results Bert - they are comfortably within the 
tentative limits I might guess for acmi accuracy. The success with
primary windings is pleasing - I had in mind the possibility that
with the thick conductors typically used, I might have to integrate
source current filaments over the volume of the conductor. The
likelihood of this seems to be dwindling and we seem to be able to get
satisfactory predictions by assuming that all the primary current is
concentrated in a filament at the wire center. Interesting that your k
factor measured at RF agrees with the value calculated for low
frequency - it appears that neither the skin effect, nor the non-
uniform secondary current are interfering with the coupling enough to 
be troublesome at this precision. Could it be that your slightly
higher k factor at RF is due to the secondary current bias towards the
low end of the coil?

> Unfortunately, my RLC bridge "fried" since these were
> taken - too much RF or EMP... :^(

Too bad! I may have a requirement for someone handy with an LCR bridge
to take some delicate measurements on the base input impedance,
Zin = Rin + jXin, on a certain type of coil, specifically the slope
dXin/dw near resonance on a large, high h/d secondary, in order to
resolve a slight theory issue. More about this later on the tssp list.

John H. Couture <couturejh-at-worldnet.att-dot-net> wrote:

> I have entered Bart's coil parameters into the JHCTES/2.3 program
> with the following result.
> 
>             acmi           Bart             JHCTES/2.3
> 
> Pri         108.6 uh       107.13 uh         108.8  uh
> Sec 1000 T   80.3 mh        87.6 mh           80.21 mh
> 
> Mut Ind     439.8 uh       627.0 uh          482.63 uh
> K Factor    .15            .20               .16
> 
> Pri turns     ?            11.6              11.56

Thanks John. Since you ran this, Bart has admitted to having 1041
secondary turns, which brings his secondary inductance into line.
The extra turns have little impact on k. I guess JHCTES/2.3 is using a
standard formula for the flat spiral inductance - is your program
released with source? BTW, acmi needs integer turns so I used 16,
again little impact on k.

> As with other past Tesla Listings trying to get test data to agree
> with TC programs is always a struggle.

Yes, and always the struggle is worthwhile. Science takes place when
theory and measurements meet head on. Sometimes the theory needs to be
fixed, other times the experimental method needs to be revised, but 
regardless of whether or not the results match, there is always a
benefit one way or another - an increment in our knowledge, an
improvement in technique. Funny this, but dealing with a mismatch of
results is always ultimately more rewarding than the rather transient
warm glow of satisfaction from having got something right!

> It is unfortunate for TC programs that less than one percent of the
> TL posts are in regards to actual TC tests.

Indeed. And too often those results which are posted are not reviewed 
critically and are thus wasted. The very best we can do for any set of
results, software, or theory, is to take it apart and try to demolish
it, on the basis that if it is correct, it will survive any amount of
scrutiny. Only by doing so do we do justice to the author's efforts
and improve the science of tesla coiling.

> Of course there are plenty of random spark length tests but these
> have little value for TC engineering design.

Absolutely. 

In response to my questions, Barton B. Anderson <tesla123-at-pacbell-dot-net>
wrote:

>> You mentioned 1uF + 18k ohms in series with the secondary - can I
>> ask what these are for?

> This is a simple rc noise filter also used to swamp any resonance
> which can be significant at these levels as expressed in a post by
> Terry back in Nov. 97.

Good point, especially when using mains supply for the primary
current - I guess every spike will set the secondary ringing.

> Today, I removed the rc filter and measured only a slight variation
> of 0.002V, so the filter isn't causing the problem.

Seems to take care of that issue.

>> The DMM may suffer a 0.2V or 0.4V shortfall in its AC readings due
>> to forward drop across internal diodes
 
> Yes - today I checked it against a scope measurement at a low AC
> voltage and found 0.54V difference, but I trust the DMM more than I
> do the scope (won't go into details). I'll take my DMM into work
> tomorrow and check it with other equipment.

Good stuff - I take it the DMM was low wrt the scope. You can always
check the scope amplitude calibration against the clipping levels of
a diode or zener. 

>> Were you able to continuously monitor the primary current during
>> the measurement run?
 
> No. I believe the resistor heating up did cause a primary current
> change during the course of the measurements. Today I checked the
> coil this time using a toaster. To keep the toaster elements at a
> stable temp, I placed a fan to blow into the heating elements.
> Measured a K of 0.172 at 1.5" (p above s). Then back to the hair
> dryer (I'm finding that the hair dryer is giving the most stable
> current readings over the course of time). (I also used a vacuum
> cleaner and iron). The hair dryer is the best. With the hair dryer,
> K is 0.174 (slight change from yesterday and today's tests
> indicating most like cause is meter reading).

You seem to be getting some consistency on the individual readings,
and the shortfall in k remains about the same. Sounds like the hair
dryer or a fan heater is the best bet for a stable primary current.
I'll bet the resistor was responsible for the trend.

>> Other than the above suggestions, the possibility remains that some
>> inductively coupled conductor loop is present, which would disturb
>> the induced voltage on the secondary.

> Well, a 4-bay light fixture was just off to one side and above the
> secondary. Before anything today, I moved the coil assembly center
> to the garage with nothing near by. The above mentioned readings
> were taken from there, so the light fixture wasn't causing
> misreadings. There's nothing else around.

That seems to take care of that possibility - it would take quite a 
closely coupled eddy loop to throw the coupling out by this much,
something in the secondary supports perhaps.

> Paul, let me check the meter at work. Bert's measured values seem to
> agree with acmi and Johns program is also very close. I'm suspecting
> the secondary meter readings to be off by about 0.4V.

If the DMM were low by 0.2V or so, the allowance would provide just 
about the right increase in k, although I'll be surprised if such a
quality instrument does not contain a precision rectifier or A/D.
At work we sometimes have problems with AC V and I readings when
the instruments involved have a different response to the mains 3rd
harmonic - a form factor sensitivity to the calibration. Might be an
issue.

> (Is this acmi similar to Marks program MANDK?).

Not having heard of MANDK I checked with pupman archives to find a
single reference,

http://www.pupman-dot-com/listarchives/1998/June/msg00510.html

It was Mark who pointed me to Grover's tables, so I'm guessing Mark's
program works the same way as acmi. The post refers to a beta version
from 18 months ago. Wonder why we've not seen a release?  Mark?

> I think after I check a suspected meter reading error, I'll redo
> the tests again using the hair dryer and grab a second Fluke for
> concentric Amp Volt readings.

Your efforts are well worthwhile and much appreciated. It goes to show
that the simple statement 'measure this, measure that, and use the
formula...' glosses over a host of subtleties, and that, as always,
some considerable effort is required to obtain unequivocal precision
measurements.

Regards All,
--
Paul Nicholson,
Manchester, UK.
--