A few comments:
> The PiL was designed using the Pi-EL Designer by Jim
> Tonne, WB6BLD,
> and checked using Ian's, GW3SEK, PiL spreadsheet. L2 is
> made from
> two, T200-6 toroids with the taps on the #12 wire
> established as
> close as possible along one side of the coil to the
> calculated
Toroids in theory are self -shielding. This would mean all
flux is confined to the iron core magnetic path, and any
short would take the reactance of all other areas of the
winding to zero inductance. It also means anything not wound
directly around the core would have no effect on the
inductance. It means spreading or squeezing turns would have
no real effect on inductance once a reasonable amount below
the self-resonant frequency of the inductor.
In real life cores like the larger T-2 mix have very
noticable flux leakage. If you squeeze or spread turns,
inductance changes quite a bit (a 2:1 inductance change
ratio is common). You can sometimes in low Q circuits get
away with shorting part of the winding out without melting
something, and the inductor will still have inductance with
turns shorted. Lead lengths and what is around the core will
affect inductance.
What this all means is if you found the inductance by
looking at "taps" without the inductor being connected to
the actual switch, or if you measured inductance on 4 MHz
and are actually using the inductor on 14 MHz (or some other
frequency), the inductance will be nowhere like "measured".
That's because you didn't measure at the working frequency,
didn't measure in the final mounting position, and perhaps
didn't measure with taps connected (?).
The air inductor in the tank is the same way, except
coupling to the outside world is greatly increased. If I
measure inductance at 4 MHz it will not be remotely close to
the real world inductance when it is connected to a band
switch, or connected to variable capacitors through leads of
any significant length, or mounted in a cabinet, or mounted
near something else. The inductance on 28 MHz (or some other
high frequency) will be nothing like the inductance measured
on 4 MHz.
Finally, the Autek is not well-suited for measuring L and C
values without very careful and limited use. It does NOT
normalize out transmission lines used to connect from the
bridge to the device you are testing. It does not normalize
out test leads.
There are so many variables involved in a dry measurement of
components on one frequency outside of the box to actual
working reactances on another frequency inside the box I'd
be amazed if it did work as planned. I wouldn't bother
doing that with an analyzer that normalizes out the
transmission line, let alone one that is a rough instrument
that does not.
What I would do, assuming I wanted to be overly fussy about
setting the taps, is get some fixed silver micas that
represent the calculated plate (input) capacitance of the
network and the calculated loading capacitance.With
everything in the final position I would disconnect the
variables and substitute (with very short leads) the fixed
capacitors and adjust the taps for a close match when the
plate side is terminated in the calculated tube loading
resistance.
You can get more than one combination of L values in a Pi L
even with two fixed caps, so you would have to be sure the L
value was reasonably close when starting. This could be done
by measuring the L section as a series resonant circuit with
known fixed capacitors just to rough it in.
I'm not a big fan of tapping a toroid with a shorting
switch. IMO that is just begging for problems unless through
experience we know what we might run into. Everything
interacts, circulating currents can get very high, and it
totally ruins the flux distribution in the core. This is
especially true when the core is used on multiple bands over
a very wide frequency range. I think most people get away
with it through pure dumb luck, or they just don't care or
know what the result is. Air coils, because of the very high
flux leakage, are fine when tapped. Especially air coils
with a long form factor compared to diameter.
73 Tom
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|