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

Re: Method of measuring self C of secondary

Hi Bob,

At 01:45 AM 04/27/2000 -0400, you wrote:
>Hi Terry and all,   (combined reply)
>
>It hard to keep with Terry's prodigious output.  Thanks for the measurement
>and mods to T5 or is it T6 with an option of voltage weighted  self (vwsC or
>medC) or  self C (true or LC bridge C).  Does it  now output the self C
>(true)
>along the coil if so can I have a copy.  Should very useful.   How about a
>output for mathcad.  Does it now produce self C (true) of the topload. You
>should set the primary V to 0v because you dont know the phase relative to
>the secondary so zero will give the least error.

E-Tesla5 can be modified for many different situations.  I sort of
"kludged" the program to get these recent results so it's probably not
ready for prime time.  I just set the primary height to zero which
effectively removes it from the calculation (lays it on the ground plane).
The 1000 Hz or "true" capacitance of the coil is simply the trivial case of
setting the voltages all to 1.  I'll modify a version and send it to you.

>
>The measured self C (true) confirms a fiddle factor in Medhurst but Medhurst
>may be for an isolated coil.  In which case  the difference between Medhurst
>and self C (true) will be dependent on the separation from the ground plane.
>Does anybody know what Medhurst assumes and if he assumes a ground
>plain what was the separation.

I know he assumes a ground plane.  I forget about the separation.  I posted
the scans of his paper (all 8Mbytes of them) at:

http://users.better-dot-org/tfritz/site/other_papers/medhurst/

Starts on page 35 then jumps to page 80 plus some letters followups...

Perhaps these scans could be put into a much smaller *.pdf file ???

>
>The self C (true) or better the equation is required  for simulation and
>maths models.
>
>The use of a voltage weighted C in a lumped resonance calculation or any
>calculation is totally new to me. As is the concept of a current weighted L.

It gets back to the electrostatics and gauss's law.  If you place a give
charge on an object it will create E-Fields around the object.  If you then
chop the objects up, each part will have it's own capacitance.  In the case
of resonating secondaries, the voltage distribution from a given charge
varies wildly along the coil and drags the E-fields with it.  That current
weighted L stuff is still a frontier ;-))

>
>The secondary can be shorted across the ends during C measurements to
>decrease L to 1/4 and hence reduce errors.

Good idea!

Cheers,

	Terry

>
>Regards Bob
>
>-----Original Message-----
>From: Tesla List <tesla-at-pupman-dot-com>
>To: tesla-at-pupman-dot-com <tesla-at-pupman-dot-com>
>Date: 26 April 2000 19:55
>Subject: Re: Method of measuring self C of secondary
>
>
>>Original Poster: Terry Fritz <twftesla-at-uswest-dot-net>
>>
>>Hi Bob and Antonio,
>>
>>At 11:58 PM 04/25/2000 -0300, you wrote:
>>>Tesla List wrote:
>>>
>>>> Original Poster: "Robert Jones" <alwynj48-at-earthlink-dot-net>
>>>
>>>>...
>>>> Using your Variac controlled high voltage AC power supply  apply 2.5kV
>AC
>>>> between the coil and ground and measure the current flow to the coil.
>Repeat
>>>> this with 5kV. Wind up the volts slowly while checking the current does
>not
>>>> say go over 100uA.
>>>>...
>>>
>>>I see a problem with this measurement. At 60 Hz, the voltage will be
>>>the same along all the coil. In real operation the voltage is zero at
>>>the bottom of the coil and maximum at the top. The difference in the
>>>charge distribution leads to different values for the equivalent
>>>"self-capacitance" of the coil. It would be interesting to see how
>>>these measurements compare with the predictions using Medhurst's
>>>formula.
>>>
>>>Antonio Carlos M. de Queiroz
>>>
>>
>>I just simply connected my LCR meter between the base wire of my bare
>>secondary and ground.  The meter gives a capacitance of 26.4pF.  The
>>Medhurst (and actual value) is 9.28pF at resonance.  I think Antonio is
>>correct that the voltage distribution on the secondary is far different at
>>the meter's 1000 Hz than at resonance.  Antonio thinks it is constant, so I
>>made the simple modification to E-Tesla5 and ran it.
>>
>>Assuming the voltage on the secondary is a constant 1 volt at all points,
>>the capacitance value the program gave is 27.2pF (only 3% off from the
>>measured reading of 26.4pF.  Even I was shocked at how close ET5 came!
>>;-)).  The distance from the base of the coil to the foil ground plane has
>>a dramatic affect on the capacitance.  I had to be sure that actual coil
>>and the programmed values were exactly the same.  If I moved the coil 1/2
>>inch closer to the ground plane (it was 1 inch) the capacitance would go up
>>about 10pF.  It appears that the vast majority of the capacitance is
>>between the base of the coil and the ground plane.  The field plot is shown
>>below:
>>
>>http://users.better-dot-org/tfritz/site/misc/Bob30.gif
>>
>>Of course, in this case, the voltage at the top of the coil is no longer
>>200X as driven by the coil's Q at resonance.  The voltage at the top of the
>>coil and the bottom are just 1.  This makes the coil have a larger apparent
>>capacitance.
>>
>>I assume the 1000Hz LCR meter is far enough away from any resonances that
>>the voltage distribution on the coil is uniform as Antonio suggests.
>>
>>Cheers,
>>
>> Terry
>>
>>
>>
>
>
>