PDA

View Full Version : Home brew r.f. chokes



K9STH
05-25-2011, 09:31 PM
While winding the r.f. choke for my latest amplifier project I "discovered" a very handy thing, 26 gauge enameled wire close winds at as close to 50 turns per inch that you could hope for! I did have a ceramic form 1 inch in diameter which I used. However, it would certainly be possible to use 1 inch PVC (or a different diameter) on which to wind the r.f. choke.

You can calculate the effective inductance of a single winding choke at the following URL:

http://www.daycounter.com/Calculators/Air-Core-Inductor-Calculator.phtml

Copying a couple of designs on the Internet for an amplifier covering 160-meters through 10-meters the designs showed a 195 mH r.f. choke in the B+ line to the final amplifier tubes. Other designs, using various tubes, show different values. Although I was trying for 195 mH I did go ahead and calculate the inductance for every quarter-inch from 1 inch long coil to 4 inches long. Those figures are in a table which accompanies this post.

I used the mentioned calculator when I was designing the final tank coil for my 160-meter amplifier using a pair of GI-7bT tubes. That coil worked very well.

Also attached is a photo of the r.f. choke that I am going to use in my new amplifier.

Since winding r.f. chokes applies to all sorts of home brew equipment I am starting this thread in the home brew forum instead of the amplifier forum. Also, the cited inductance calculator is not just for r.f. chokes. It can be used to calculate the diameter, number of turns, length, etc., of coils used in transmitters, receivers, converters, traps, etc.

Glen, K9STH

65357 65358

W0BTU
05-25-2011, 09:38 PM
Any idea at what freqs this is choke is self-resonant ?

K9STH
05-25-2011, 10:53 PM
I would have to "dip it out"!

Basically, I started out copying a design that was "proven"! :rolleyes:

If I get a chance, I'll "dig out" my GDO and see.

Glen, K9STH

K9STH
05-25-2011, 11:16 PM
I found this site:

http://www.smeter.net/feeding/transmission-line-choke-coils.php

Which allows you to calculate the self resonant frequency of a single layer coil. Putting in all sorts of differing factors (lead length, etc.), all of the answers keep coming out right at 12.6 MHz.

I'll try some of the r.f. chokes that are described in some of the amplifier articles on the Internet and see how they compare! Since I haven't even installed this r.f. choke I need to do some comparison with the self resonant frequency that others are using.

Glen, K9STH

W8JI
05-25-2011, 11:30 PM
I found this site:

http://www.smeter.net/feeding/transmission-line-choke-coils.php

The program can't possibly be accurate because it ignores dielectrics which have a profound effect on the self resonant frequency of a choke like that.

I can move series resonances about 20% or more in a large RF choke simply by changing the form from a solid materials to hollow thin form.


Better people measure it.

Here is how:

http://www.w8ji.com/rf_plate_choke.htm

K9STH
05-25-2011, 11:57 PM
I just checked the construction details of the r.f. choke that was supposed to be 195 mH. Using the inductance calculator that I posted in the first entry it calculates out to right at 80 mH! Big difference!

Looking at other "proven" designs I find the range from around 50 mH to around 130 mH for 160-meter through 10-meter operation! The "plot thickens"!

Unfortunately, your method of measuring the self resonant point requires an "opened up" transmitter. However, I think it would be possible to accomplish the same thing using a signal generator and an r.f. probe and/or an oscilloscope.

Since I don't have an "opened" transmitter but do have several signal generators, r.f. probes, and oscilloscopes I am going to see if that works!

Glen, K9STH

W8JI
05-26-2011, 03:53 AM
I just checked the construction details of the r.f. choke that was supposed to be 195 mH. Using the inductance calculator that I posted in the first entry it calculates out to right at 80 mH! Big difference!

Looking at other "proven" designs I find the range from around 50 mH to around 130 mH for 160-meter through 10-meter operation! The "plot thickens"!

Unfortunately, your method of measuring the self resonant point requires an "opened up" transmitter. However, I think it would be possible to accomplish the same thing using a signal generator and an r.f. probe and/or an oscilloscope.

Since I don't have an "opened" transmitter but do have several signal generators, r.f. probes, and oscilloscopes I am going to see if that works!

Glen, K9STH

A signal generator and sensitive diode detector can work.

K9STH
05-27-2011, 04:41 AM
Checked for self resonance of the single pie r.f. choke using my old TS-497B (military Measurements Model 80) and my Motorola TEK-7A r.f. volt meter. Checked frequency with a counter.

Got 3 frequencies:

16.104 MHz Great!

23.125 MHz Pretty good!

29.075 MHz Whoops!


Removed some turns and made it into a 4 section choke.

Got 1 frequency (below 35 MHz, didn't check higher)

20.391 MHz Pretty good!

However, I am having problems trying to determine the inductance using the grid dip meter and known capacitor method in a parallel circuit. No matter what value capacitor I use the circuit "dips" at pretty close to 40 MHz. Tried capacitors from 27 pf to 470 pf and get basically the same dip with my Millen GDO. The inductance then calculates anywhere between about 33 mH and well over 550 mH!

I'll try my Heath GDO in the morning "just in case" there is something "screwy" with the Millen. However, have never had problems with the Millen on any other tuned circuit.

Also, will try a different way of coupling rather than the "usual" putting the GDO coil right next to the choke because it may be reading the individual section and not the entire choke.

Glen, K9STH

K9STH
05-27-2011, 03:30 PM
I gave up on trying to measure the entire choke using the grid dip meter / capacitor method. Therefore, I calculated the inductance of each section using the on-line calculator at

http://www.daycounter.com/Calculators/Air-Core-Inductor-Calculator.phtml

The form is 1 inch diameter ceramic tube and the windings are 26 gauge wire. The sections consist of, from top to bottom, 16 turns, 21, turns, 27 turns, and 53 turns. The calculated inductance of each section is

16 turns = 8.31 mH
21 turns = 12.67 mH
27 turns = 18.41 mH
53 turns = 46.50 mH

Total = 85.89 mH or, for all practical purposes, 86 mH.

Using the same calculator and treating the coil as a single coil I get a calculated inductance of 86.64 mH. Pretty darn close!

Using the calculator at

http://www.smeter.net/feeding/transmission-line-choke-coils.php

The self resonant frequency comes out to be just over 18.900 MHz and I get 20.391 MHz by measurement.

Glen, K9STH

KL7AJ
05-27-2011, 04:28 PM
Back when I had access to an HP network analyzer, just for jollies I used to measure simple chokes like these up to 3GHZ. It's AMAZING how many resonances you find!....series, parallel....both at the same time...etc! The only inductor that didn't have multiple resonances was a piece of straight wire....as long as it was short enough to resonate above 3ghz. :)

Eric

W8JI
05-27-2011, 05:25 PM
Hi Glen,

You should be able to dip the choke regardless of multiple sections if you dip low enough. You would probably need to measure around the broadcast band or slightly higher (maybe up to 80 meters). The capacitance is going to be very low value!!

However, using that method to determine inductance is meaningless except down around 160 and 80 meters. If you sweep a large choke you will find all sorts of spirals on a Smyth Chart display, so the choke impedance is really all over the place except way down at the low end.

Over much of upper HF they are capacitive, not even inductive.

Look at the chart near the bottom of this page:

http://www.w8ji.com/rf_plate_choke.htm

That table gives actual measurements of the later choke in the AL80A and B. If you read the chart, you will see the inductance changes between 160 and 80 and that above 80 most readings are capacitive!!!

If you try to measure inductance with a capacitor above 5 MHz, good luck!!! Because the choke is for the most part on the capacitive side of the Smyth Chart.

F MHz L or C Xs Rs Xp Rp Q
1.8 255 ÁH 2550 22 2550 295590 116
3.5 275 ÁH 6060 85 6061 432127 71
7 .36 pF -315000 3851 -327203 26764200 82
10.15 2.3 pf -6900 79 -6901 602737 87
12.1 0 0 504 0 504 0
14 .6 pF -19200 380 -19952 1008110 51
15.7 0
0 550 0 550 0
18.2 .6 pF -14300 220 -14357 933205 65
21 1 pF -7800 126 -7846 486452 62
24.8 1.25 pF -5240 87 -5368 467250 60
28 1.37 pF -4130 69 -4229 291096 60

Rp = Rs + Xs2 / Rs
Xp = Xs + Rp2 / Xs

WB2WIK
05-27-2011, 05:38 PM
I just checked the construction details of the r.f. choke that was supposed to be 195 mH. Using the inductance calculator that I posted in the first entry it calculates out to right at 80 mH! Big difference!

Looking at other "proven" designs I find the range from around 50 mH to around 130 mH for 160-meter through 10-meter operation! The "plot thickens"!

Unfortunately, your method of measuring the self resonant point requires an "opened up" transmitter. However, I think it would be possible to accomplish the same thing using a signal generator and an r.f. probe and/or an oscilloscope.

Since I don't have an "opened" transmitter but do have several signal generators, r.f. probes, and oscilloscopes I am going to see if that works!

Glen, K9STH

Shouldn't all those units be "uH" and not "mH?" I don't see how you'd get up into the "mH" region with chokes like this...

K9STH
05-27-2011, 08:29 PM
WIK:

True! My right index finger keeps wanting to go down and not up on the QWERTY keyboard. Therefore "m" instead of "u"! Well, that's my story and I am sticking to it! :rolleyes:

Glen, K9STH

K9STH
05-27-2011, 08:38 PM
JI:

I gave up trying to determine the inductance using a grid dip meter! I do have a couple of low frequency coils for the Millen but they are a "pain" to use. I figure the calculations of the uH using the on-line calculator are pretty accurate.

The fact that the series resonance is far enough outside of any amateur band (15 meters is the closest) I am going to try the r.f. choke.

Attached is a photo of the choke. I used several coats of clear Krylon to hold the windings in place.

Glen, K9STH

65481

WB2WIK
05-27-2011, 08:57 PM
Looks good. I've used the P/N RFC-3 from R.F. Parts for this (225uH, 1.5A, 4kV, three windings on a 1" diameter ceramic form) and they always seem to work. Pictorial below.

W8JI
05-27-2011, 11:16 PM
Looks good. I've used the P/N RFC-3 from R.F. Parts for this (225uH, 1.5A, 4kV, three windings on a 1" diameter ceramic form) and they always seem to work. Pictorial below.


That's a choke I designed and tested. :-)

WB2WIK
05-27-2011, 11:21 PM
That's a choke I designed and tested. :-)

Well, there ya go! That's the choke I'm using in my homebrew 4-1000A for HF. I wound my own a few times, always had some resonance near a ham band, and said the hell with it and bought theirs. It worked.

K9STH
05-28-2011, 05:59 PM
After spraying the choke with clear Krylon (several layers) the self resonant frequency moved up into the 15-meter band. Therefore, I removed several turns from the bottom (most turns). There are now 2 "dips" below 35 MHz and above 1.5 MHz. The first is at 22.503 MHz and a MUCH greater dip at 31.240 MHz.

Glen, K9STH

W8JI
05-29-2011, 04:09 PM
After spraying the choke with clear Krylon (several layers) the self resonant frequency moved up into the 15-meter band. Therefore, I removed several turns from the bottom (most turns). There are now 2 "dips" below 35 MHz and above 1.5 MHz. The first is at 22.503 MHz and a MUCH greater dip at 31.240 MHz.

Glen, K9STH

I hope you mean series resonant frequency and not self resonant fequency. :-)

My web page on chokes shows how to test or probe for the best area to remove turns.

You see now why I say the on line calculator isn't so good. Anything that changes the dielectric changes the series resonant frequency all over the place.

This includes the chassis we mount the choke in, and everything near the choke.

K9STH
05-29-2011, 04:48 PM
True! The "series" resonant frequency. Even without the Krylon the on-line calculator was well over a MHz away from the actual measured frequency.

My old TS-497B signal generator definitely has enough output, when the attenuator calibration is run to the maximum, to get very good indications on my old Motorola TEK-7A r.f. volt meter. Neither of my service monitors have enough output to get any meaningful readings on the TEK-7A. Also, it would take hours of stepping through the frequency range 100 Hz at a time! MUCH easier just to turn the analog dial and watch the meter! When the meter dips, read the frequency with the counter.

Glen, K9STH

KM1H
05-30-2011, 09:07 PM
Have you tried a ceramic or hard Teflon tube with a ferrite rod?

Carl

ad: AC6LA-1