Some thoughts about homebrew inductors
 

A number of recent posts have discussed inductors used in homebrew
equipment. At least, in states-side, it appears that toroids are the
defacto inductor material for anything from audio to lower VHF. and
why not, they are easy to wind, cheap (relatively) and available from
well known sources.

i live in india, and toroids are hard to come by. the only time i did
use amidon toroids was when a friend coerced me into assembling a
NorCal40 for him (and then he 'gifted' it to me and took home my
homebrew 40M rig instead). as a result of being deprived of easy
supply of toroids, I had to look at almost everything as an coil
former (from TV balun cores to ... well, lets see...).

i discovered that getting a particular inductance on an air wound coil
is a bit of a black art. The same coil will report varying
inductances, if measured at different frequencies! for measuring my
coils, i use a test oscillator with an equivalent capacitance of
165pf. I plug in a test coil, measure the frequency on the frequency
counter. I punch these frequencies back into the program given below
to get the inductance.

here is the small program that i wrote (apologies to the non-coders):

#include stdio.h
#include stdlib.h

void main()
{
char input[100];
float uHpF;

while (1)
{

puts("Enter the freq in KHz:");
gets(input);

uHpF = (1000000/(6.28 * atoi(input))) * (1000000/(6.28 *
atoi(input)));
printf ("uH-pf required: %f\n", uHpF);

printf ("at 165pf, coil inductace is %f\n", uHpF/165);

puts("enter inductance/capacitance:");
gets(input);
printf("cap required is %f pf\n***\n", (uHpF/atof(input)));

}
}

recently, i started a VHF NBFM transceiver project on a dare. To my
surprise i found out that the inductances that were reported by this
method were not applicable at VHF (the equivalent inductance at VHF
was much higher for the same coil). To measure the inductances at VHF
accurately, I rigged up a different test oscillator with an equivalent
capacitance of 22pf. This worked better.

My explanation for this is two fold. First, except for space (and
practically also air), every core type has a different frequency
response. this is similar to an electrolytic capacitor (which will not
handle HF frequencies). Second issue is that the inter-winding
capacitances tend to become more prominent at higher frequencies
leading to the troublesome but frequent self-resonance of coils at
VHF.

Of all un-conventional materials that I have tried, I have finaly
discovered something that works for me for almost all of HF and VHF.
Although it does not work for broad-band (due to lower inductance per
turn).

From my local hardware store, I purchases a bag full of nylon tap
washers. These are about 15mm across, 5mm thick and have a hole of 4mm
in the center. I purchased them at an equivalent of 1 cent each. When
I use them as toroids, I have calculated that they give 1.5 nH / turn
squared. That is, wind 20 turns, I get 20 * 20 * 1.5 = 600 nH = 0.6uH
coil. At higher HF and VHF, they give 1.6 nH / turn squared.

Now I use them in almost everything that I build. From receiver
front-ends to VHF power strips. There are only two limitations to
them. The tempurature drift can be quite high because nylon is
elastic. The easy fix is to dunk the finished coil in some kind of an
epoxy glue. The other limitation is that quite a high number of turns
are required to hit required inductances. This makes them unusable for
broad band work (i am only hypothesising, i havent build a broadband
amplifier using the washers and found it unfit, someone please do this
and tell me).

Nylon has pretty good performance well into upper VHF. I dont have the
capability of checking things above 200MHz. but what i can see upto
200MHz is pretty impressive.

I made a hartley oscillator for 10.7 MHz using 45 turns with a tap at
10 turns using the nylon washer as a toroid. Without the epoxy glue, i
found a rapid upward drift (powering on the oscillator immediately
after soldering in the coil). After 15 minutes, it slowed down (having
drifted about 50 KHz) to a crawl of under 1 KHz/min. After applying
the epoxy glue and sticking it to the copper clad board (it is made
ugly style) the oscillator drifts less then 100 Hz/min after 10
minutes warm-up on an open bench without any shielding. This is good
enough for me. I was intending to use it as a modulated NBFM carrier
generator to be mixed back to 144MHz.

- farhan

At VHF/UHF most inductance formulas lose accuracy because the thickness of
the wire,
used to make the inductor, can no longer be ignored.

"Ashhar Farhan" wrote in message
om...
A number of recent posts have discussed inductors used in homebrew
equipment. At least, in states-side, it appears that toroids are the
defacto inductor material for anything from audio to lower VHF. and
why not, they are easy to wind, cheap (relatively) and available from
well known sources.

i live in india, and toroids are hard to come by. the only time i did
use amidon toroids was when a friend coerced me into assembling a
NorCal40 for him (and then he 'gifted' it to me and took home my
homebrew 40M rig instead). as a result of being deprived of easy
supply of toroids, I had to look at almost everything as an coil
former (from TV balun cores to ... well, lets see...).

i discovered that getting a particular inductance on an air wound coil
is a bit of a black art. The same coil will report varying
inductances, if measured at different frequencies! for measuring my
coils, i use a test oscillator with an equivalent capacitance of
165pf. I plug in a test coil, measure the frequency on the frequency
counter. I punch these frequencies back into the program given below
to get the inductance.

here is the small program that i wrote (apologies to the non-coders):

#include stdio.h
#include stdlib.h

void main()
{
char input[100];
float uHpF;

while (1)
{

puts("Enter the freq in KHz:");
gets(input);

uHpF = (1000000/(6.28 * atoi(input))) * (1000000/(6.28 *
atoi(input)));
printf ("uH-pf required: %f\n", uHpF);

printf ("at 165pf, coil inductace is %f\n", uHpF/165);

puts("enter inductance/capacitance:");
gets(input);
printf("cap required is %f pf\n***\n", (uHpF/atof(input)));

}
}

recently, i started a VHF NBFM transceiver project on a dare. To my
surprise i found out that the inductances that were reported by this
method were not applicable at VHF (the equivalent inductance at VHF
was much higher for the same coil). To measure the inductances at VHF
accurately, I rigged up a different test oscillator with an equivalent
capacitance of 22pf. This worked better.

My explanation for this is two fold. First, except for space (and
practically also air), every core type has a different frequency
response. this is similar to an electrolytic capacitor (which will not
handle HF frequencies). Second issue is that the inter-winding
capacitances tend to become more prominent at higher frequencies
leading to the troublesome but frequent self-resonance of coils at
VHF.

Of all un-conventional materials that I have tried, I have finaly
discovered something that works for me for almost all of HF and VHF.
Although it does not work for broad-band (due to lower inductance per
turn).

From my local hardware store, I purchases a bag full of nylon tap
washers. These are about 15mm across, 5mm thick and have a hole of 4mm
in the center. I purchased them at an equivalent of 1 cent each. When
I use them as toroids, I have calculated that they give 1.5 nH / turn
squared. That is, wind 20 turns, I get 20 * 20 * 1.5 = 600 nH = 0.6uH
coil. At higher HF and VHF, they give 1.6 nH / turn squared.

Now I use them in almost everything that I build. From receiver
front-ends to VHF power strips. There are only two limitations to
them. The tempurature drift can be quite high because nylon is
elastic. The easy fix is to dunk the finished coil in some kind of an
epoxy glue. The other limitation is that quite a high number of turns
are required to hit required inductances. This makes them unusable for
broad band work (i am only hypothesising, i havent build a broadband
amplifier using the washers and found it unfit, someone please do this
and tell me).

Nylon has pretty good performance well into upper VHF. I dont have the
capability of checking things above 200MHz. but what i can see upto
200MHz is pretty impressive.

I made a hartley oscillator for 10.7 MHz using 45 turns with a tap at
10 turns using the nylon washer as a toroid. Without the epoxy glue, i
found a rapid upward drift (powering on the oscillator immediately
after soldering in the coil). After 15 minutes, it slowed down (having
drifted about 50 KHz) to a crawl of under 1 KHz/min. After applying
the epoxy glue and sticking it to the copper clad board (it is made
ugly style) the oscillator drifts less then 100 Hz/min after 10
minutes warm-up on an open bench without any shielding. This is good
enough for me. I was intending to use it as a modulated NBFM carrier
generator to be mixed back to 144MHz.

- farhan

At VHF/UHF most inductance formulas lose accuracy because the thickness of
the wire,
used to make the inductor, can no longer be ignored.

"Ashhar Farhan" wrote in message
om...
A number of recent posts have discussed inductors used in homebrew
equipment. At least, in states-side, it appears that toroids are the
defacto inductor material for anything from audio to lower VHF. and
why not, they are easy to wind, cheap (relatively) and available from
well known sources.

i live in india, and toroids are hard to come by. the only time i did
use amidon toroids was when a friend coerced me into assembling a
NorCal40 for him (and then he 'gifted' it to me and took home my
homebrew 40M rig instead). as a result of being deprived of easy
supply of toroids, I had to look at almost everything as an coil
former (from TV balun cores to ... well, lets see...).

i discovered that getting a particular inductance on an air wound coil
is a bit of a black art. The same coil will report varying
inductances, if measured at different frequencies! for measuring my
coils, i use a test oscillator with an equivalent capacitance of
165pf. I plug in a test coil, measure the frequency on the frequency
counter. I punch these frequencies back into the program given below
to get the inductance.

here is the small program that i wrote (apologies to the non-coders):

#include stdio.h
#include stdlib.h

void main()
{
char input[100];
float uHpF;

while (1)
{

puts("Enter the freq in KHz:");
gets(input);

uHpF = (1000000/(6.28 * atoi(input))) * (1000000/(6.28 *
atoi(input)));
printf ("uH-pf required: %f\n", uHpF);

printf ("at 165pf, coil inductace is %f\n", uHpF/165);

puts("enter inductance/capacitance:");
gets(input);
printf("cap required is %f pf\n***\n", (uHpF/atof(input)));

}
}

recently, i started a VHF NBFM transceiver project on a dare. To my
surprise i found out that the inductances that were reported by this
method were not applicable at VHF (the equivalent inductance at VHF
was much higher for the same coil). To measure the inductances at VHF
accurately, I rigged up a different test oscillator with an equivalent
capacitance of 22pf. This worked better.

My explanation for this is two fold. First, except for space (and
practically also air), every core type has a different frequency
response. this is similar to an electrolytic capacitor (which will not
handle HF frequencies). Second issue is that the inter-winding
capacitances tend to become more prominent at higher frequencies
leading to the troublesome but frequent self-resonance of coils at
VHF.

Of all un-conventional materials that I have tried, I have finaly
discovered something that works for me for almost all of HF and VHF.
Although it does not work for broad-band (due to lower inductance per
turn).

From my local hardware store, I purchases a bag full of nylon tap
washers. These are about 15mm across, 5mm thick and have a hole of 4mm
in the center. I purchased them at an equivalent of 1 cent each. When
I use them as toroids, I have calculated that they give 1.5 nH / turn
squared. That is, wind 20 turns, I get 20 * 20 * 1.5 = 600 nH = 0.6uH
coil. At higher HF and VHF, they give 1.6 nH / turn squared.

Now I use them in almost everything that I build. From receiver
front-ends to VHF power strips. There are only two limitations to
them. The tempurature drift can be quite high because nylon is
elastic. The easy fix is to dunk the finished coil in some kind of an
epoxy glue. The other limitation is that quite a high number of turns
are required to hit required inductances. This makes them unusable for
broad band work (i am only hypothesising, i havent build a broadband
amplifier using the washers and found it unfit, someone please do this
and tell me).

Nylon has pretty good performance well into upper VHF. I dont have the
capability of checking things above 200MHz. but what i can see upto
200MHz is pretty impressive.

I made a hartley oscillator for 10.7 MHz using 45 turns with a tap at
10 turns using the nylon washer as a toroid. Without the epoxy glue, i
found a rapid upward drift (powering on the oscillator immediately
after soldering in the coil). After 15 minutes, it slowed down (having
drifted about 50 KHz) to a crawl of under 1 KHz/min. After applying
the epoxy glue and sticking it to the copper clad board (it is made
ugly style) the oscillator drifts less then 100 Hz/min after 10
minutes warm-up on an open bench without any shielding. This is good
enough for me. I was intending to use it as a modulated NBFM carrier
generator to be mixed back to 144MHz.

- farhan