[Elecraft] OT: Effect of Compression and ExpansionontheInductanceof Toroids?

[W3FPR - Don Wilhelm]

Ron,

I have the same inductance meter, and it does a good job because it does
measure using the LR time constant rather than frequency, but 'physics is
physics', and with a true toroidal inductor, there is still no dependency on
the turns spacing (the math says so).

Now, for the practical side of things, I do believe that your results (and
others with similar findings) deviate from the 'classic physics' treatment
of the ideal (ignoring practical behavior) - and when the toroid turns are
irregular (not evenly spaced around the circumference of the core), there is
some part of that coil behaving as a solenoidal inductor where turns spacing
IS a factor.

Remember that a toroid is simply a solenoidal coil formed into a circle with
the ends of the coil meeting.  This ideal toroid has equal turns spacing all
the way around.

So my current conclusion is -- IF the toroid is equally spaced around the
core, the inductance is not dependent on the spacing of the turns, BUT, if
the turns are compressed anywhere around the core, additional factors rear
their ugly heads because the inductor is a combination of a solenoidal coil
and a toroid coil, and the math becomes quite complex - just how much of
each effect depends on just how much deviation from an ideal toroid shape
exists in the configuration at hand.

In a frequency dependent tuned circuit, how much of the frequency shift is
due to the inter-turn capacitance or a change in inductance is (to me) only
a matter of curiosity - the plain fact is that some change in the resonant
frequency (or inductance) can be achieved by changing the turns spacing, but
we all know that the adjustment range is small - the really big determining
factor is simply the number of turns.

As an example, I often improve the 10/12 meter Low Pass Filter
characteristics of a KPA100 by squeezing the toroid turns "just the right
way", but I determine what is the 'right way' by monitoring the impedance
with my MFJ259B as I do it.  Sometimes it is 'this way', while other times
it is 'that way' - the batch to batch change in the permeability of the
cores likely accounts for more variation than the turns spacing itself.
BTW, this change does not really help the KPA100 output, but it improves the
base K2 10 meter efficiency at 10 watts or lower when the KPA100 is
installed.

The overall inductance change that I have experienced is about 10%, so that
is within the normal design tolerances using 10% resistors and capacitors -
so except for satisfaction of the curiosity factor, I would say just to wind
the toroids with the proper number of turns, and 'tweak' them in-circuit as
required and as close as is permitted by your measurement capability.

73,
Don W3FPR

> -----Original Message-----

>
> Eric, KE6US wrote:
>
> I don't have your experience or background, Ron, but the L-meter is
> measuring inductance indirectly and ignoring the presence of distributed
> capacitance. Maybe it isn't as negligible at the frequency of the meter as
> you think.
>
> But in the end, it doesn't really matter. The real test is how a
> particular
> toroid reacts in the circuit it was intended for. If adjusting the turns
> spacing changes circuit resonance, then...it changes circuit resonance.
> That's the result we would have been looking for. Hi.
>
> --------------------------
>
> I'm fond of saying that the only dumb question is the question
> not asked, so
> your question is an excellent one, Eric. As for lots of experience, for me
> that's too often just an opportunity to remember more things incorrectly.
>
> About your question, my L-meter evaluates inductance using a low-frequency
> square wave at about 60 kHz that goes to a differentiator consisting of a
> 200 ohm resistor in series with the unknown inductance. The
> waveform at the
> junction of the resistor and inductor is a series of spikes at
> the frequency
> of the input square wave. The spikes decay at a rate proportional to the
> time constant of the resistor and inductor. This decay rate is converted
> into a voltage in a simple logic circuit.
>
> The ARRL Handbook has featured this circuit in a number of
> editions. It's a
> simple and surprisingly accurate meter, depending upon the quality of the
> inductors used to calibrate it.
>
> The way the circuit works means that any significant capacity in parallel
> with the inductor would tend to cancel the effect of the
> inductor, and make
> the inductance value displayed read lower than it really was, not higher.
>
> Again, at the bench I tested that conclusion by adding capacitors in
> parallel with the inductor. Sure 'nuf, adding capacitance decreased the
> reading. Recall that squeezing the turns together raised the test inductor
> value read on the meter from 3.1 to 3.3 uH. Now I added fixed capacitors
> across the inductor to simulate added distributed capacitance caused by
> squeezing them together. Adding 10 pf of capacitance across the
> inductor had
> zero effect on the reading. Adding 33 pf across the inductor *lowered* the
> reading by 0.1 uH. We're dealing with a very small toroid in this
> test whose
> inter-turn capacitance isn't going to be as much as 2 or 3 pf
> when squeezed
> together.
>
> So I'm confident that the added capacitance by squeezing the
> turns together
> is not what is causing my L-meter to show increased inductance.
> If anything,
> the capacitance would tend to cause the L-meter to show lower inductance.
>
> Ron AC7AC
>
>
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.344 / Virus Database: 267.10.16/83 - Release Date: 8/26/2005