Although I don't have the K6STI program, I have many other
inductor Q programs and none of them hit close. One way to test
the program is to raise the frequency and watch Q.
1.) Q is almost always maximum between 1/5th and 1/7th the self-
resonant frequency of a single layer inductor having a non-ferrous
core. Q decreases above and below that region.
2.) Q is almost always maximum when turns are spaced one turn
diameter, unless the wire is covered with a dielectric. If the
conductor is insulated, the Q peak generally peaks with slightly
increased turn spacing.
3.) Any dielectric or anything that increases turn-to-turn
capacitance causes a noticeable reduction in Q.
4.) Optimum form factor (L/D) ranges from around 1:1 to 5:1 or
more depending on application. The higher the reactance required
for a given frequency, the longer the optimum form factor becomes.
That's why long slender mobile loading coils are so good, and why
tank coils (with lower reactance) are short and almost a square
(1:1) form factor.
5.) The highest Q I have ever measured is in the high hundreds,
around 800 or so. That was with a moderate reactance silver plated
copper tubing inductor in a large shielded fixture. Many edge-
wound inductors are very good, in the Q 500-700 range. Yet most
inductor model programs I have tried can indicate Q's in the
thousands, while the actual inductor is in the hundreds or less!!!
Most modelling programs I have tried fail to follow these well-
established real-world measured criteria. The above criteria is a
good test to see how good a modeling program is.
> I ran a program written by K6STI, which computes coil losses and the
> results are interesting.
The results are also misleading, if anyone takes them as directly
related to loss change in a normal inductor. The biggest problem,
assuming the program is reliable, is you almost made a strip line
or transmission line rather than a good inductor. Transmission
lines, like linear loading system touted as "low loss" by antenna
manufacturers, make notoriously poor inductors and capacitors!
I measured the Q of a very similar inductor on 28Mhz, and I get
Q=105 with air dielectric, Q=100 with .05" thick teflon formed into a
form inside the coil, and Q=102 with Nylon .05" thick. All are within
tolerance of measurement error.
The less electric field in the dielectric, the less change the
dielectric will cause. For example suspending a length of 1" PVC
conduit inside the 3" coil caused no measurable change in Q. If the
PVC coated the turns, the change would be dramatic.
The amount of electric flux in the dielectric is a major factor.
If Q was important I'd follow the guidelines for selecting proper form
factors, and suspend the coil away from the form with narrow
spacers. I'd worry much more about HOW I built the inductor rather
than what type of material gave the mechanical support. Of course
I wouldn't use wood-coated wire or a grooved wood form that the
wire lays in for a number of reasons.
>
> Using 28 mhz, the highest frequency of the normal amp, I looked at a 4
> turn coil made of #14 wire with 4 turns on a 3" diameter form and 2"
> long.
>
> Form loss was calculated as:
>
> Dielectric Loss in %
> Air 0
> Air Dux .1
> Polystyrene .2
> Teflon .3
> Steatite 3.0
> Glass 3.8
> Polycarbonite 18.9
> PVC 17.0
> Porcelain 21.2
> Bakelite 24.7
> Nylon 28.2
> Douglas Fir 31.3
>
> QST in September, 1931 reports on the new National SW-3 receiver.
> James Millen notes that bakelite is often made with wood flour
> (perhaps Douglas Fir) and has very high losses - especially on 10 and
> 20 meters. He recommends the use of National R-39 material.
>
> Colin K7FM
>
>
>
>
> --
> FAQ on WWW: http://www.contesting.com/FAQ/amps
> Submissions: amps@contesting.com
> Administrative requests: amps-REQUEST@contesting.com
> Problems: owner-amps@contesting.com
>
73, Tom W8JI
W8JI@contesting.com
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|