EMRFD Message Archive 3249
Message Date From Subject 3249 2009-06-28 06:58:17 bkopski SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
At first it appeared the differences were because of core materials in the toroidal output transformers. Wes had used a FB43-2401 and I used a FT37-43. But was this really the difference?
The transformers were the same 3.2:1 ratios realized with 16/5 windings. Their purpose was to transform 470 ohms to a nominal 50 ohms on the 5T outputs. A casual comment in an e-mail caused me to wonder more about the transformer windings than the core materials so an experiment was carried out.
I wound two new cores, one of each type, and both with 16T primaries but two 5T secondaries. The latter were wound in 2 places: one over the primary, and one "opposite" the primary. I also wound a 16/5T BN-43-202 binocular core out of curiosity. Two photos of these transformers can be found among the five related photos in the K3NHI folder.
Return loss swept measurements were made on the three transformer samples using all 5T windings. Return Loss `scope trace photo results are shown in the same folder.
It is clear that the two toroidal transformers performed similarly when the same winding structures were compared. However, significant differences occurred when the differently located 5T secondaries were tested. It seems clear that the physical location of the secondary winding with respect to the primary winding contributes heavily to the transformer performance. Further, it is clear the binocular core did an overall better job with the impedance transformation than did either toroid.
All this pretty much undermines my historic understanding of and belief in toroidal transformer behavior. Now I suspect there may be even more subtleties present like maybe with "how spread out" or "how tightly" wound these cores are, and perhaps even with what wire sizes are used. But those experiments are for another day. For today, my windings will all be overlaid!
Bob, K3NHI3250 2009-06-28 07:55:17 Phil Sittner Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
Your results are very interesting and quite informative. I have had some poor results with some toroidal transformers and dismissed them as poor designs on my part-which it appears they were-but did not have a solution to the problem. I think I'll do a bit of experimenting as well. Thanks for sharing your experiments.
Phil,
KD6RM
----- Original Message -----
3251 2009-06-28 07:58:35 leon Heller Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
3252 2009-06-28 08:09:18 Graham / KE9H Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
You didn't mention what frequency and bandwidth that you are working
at, because that has a big impact, also, but broad band transformers
using ferrites are not simple "turns-ratio" transformers, even though they
might appear that way at first inspection.
Wire size (and insulation thickness) has a noticable effect. Bifilar and
trifilar
winding techniques can bring big improvements. The twisted wire
techniques don't work well above three wires wound together.
Parasitics, both inductance and capacitance can affect performance severely.
Towards the upper end of the frequency range in a broad band
transformer, you are in a transmission line mode, and you want
the native RF impedance of the wire in the windings to be about the
arithmetic mean of the input and output impedances. With just
simple "scatter" winding, the impedance is uncontrolled, and
seeming small changes in winding techniques have big unexpected
changes in performance. Twisted wire techniques can present a much
more controlled wire impedance effect to the transformer.
Chris Trask's article on broad band transformers in March/April 2005 QEX
taught me a lot. (Thanks, Chris). But I did need to read it several times
before what he was explaining sank into my head.
And if you are building a single frequency narrow band transformer, you
are probably better off using powdered iron in a resonant configuration,
and not using ferrites.
3253 2009-06-28 08:31:45 chuck adams Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
two inductors and a coefficient of coupling factor
k.
L1 1 2 0.5mH
L2 4 3 0.5mH
KXFRMR L1 L2 0.67
where 0.67 is coefficient of coupling.
The mutual inductance is M=k*\sqrt{L_p*L_s}.
There is a way to measure the coefficient of
coupling. Measure the impedance of one winding
with the other open. Then measure again with
the other winding shorted. Then calculate the
factor with a magic formula.
Unfortunately, I am in the process of getting
ready for a one week trip and a move, so my
derivation of the formula and my notes are
in a pile of papers somewhere. I will dig them
out next weekend when I get back. I apologize
for not having the info stored in my gray matter
at this time.
I was going to dig this up anyway as I was
going to do some experiments on coupling
using coils wound of 35mm plastic film
canisters (collected from one-hour photo
places that were going to throw them away,
so next time you are near one of these places
ask them for a few for boy scout projects)
in a study of coupling effects on Paul
Harden's (NA5N) regen receiver located on the
web.
Also critical to transformers is the impedance
load and there are tons of experiments to be
done on frequency effects.
So, maybe while I'm away, someone can do the
derivation, or find it in some esoteric electrical
engineering book before I get back. I'll even
see if I can derive it again while on the trip
to play in the WSoP in LV NV.
dit dit
chuck3255 2009-06-28 08:56:58 joop_l Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS 3291 2009-07-07 08:14:33 timshoppa Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
Maybe there is a good reason why so many popular and succesful and reproducible designs (radios, baluns, etc.) use bifiliar and trifiliar windings instead of separate windings.
I read the books about how bifiliar windings are called "transmission line transformers" and still don't really have an intuitive feel for why they can work so much better at 80M or 160M wavelengths. The spacing and winding lengths are all tiny compared to a wavelength, whether it's bifiliar or separate windings on a core.
Tim N3QE3292 2009-07-07 08:31:34 Chris Trask Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
>I read the books about how bifiliar windings are called "transmission
>line transformers" and still don't really have an intuitive feel for
>why they can work so much better at 80M or 160M wavelengths. The
>spacing and winding lengths are all tiny compared to a wavelength,
>whether it's bifiliar or separate windings on a core.
>
Lefferson, P., "Twisted Magnet Wire Transmission Line," IEEE
Transactions on Parts, Hybrids, and Packaging, Vol. PHP-7,
No. 4, Dec 1971, pp. 148-154.
The transformer core adds what is known as a "magnetizing inductance" to each of the conductors, which serves to lower the low-cutoff frequency of the transmission line. I discuss this aspect of wideband transformers in:
http://www.home.earthlink.net/~christrask/TraskTLTTutorial.pdf
as well as:
"Transmission Line Transformers: Theory, Design and Applications -
Part 1," High Frequency Electronics, December 2005, pp. 46-53.
"Transmission Line Transformers: Theory, Design and Applications -
Part 2," High Frequency Electronics, January 2006, pp. 26-33.
Chris
,----------------------. High Performance Mixers and
/ What's all this \ Amplifiers for RF Communications
/ extinct stuff, anyhow? /
\ _______,--------------' Chris Trask / N7ZWY
_3295 2009-07-07 12:27:41 Gary Johnson Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS 3296 2009-07-08 03:33:19 leon Heller Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
3297 2009-07-08 03:46:19 Ashhar Farhan Re: SOME OBSERVATIONS ON TOROIDAL TRANSFORMERS
close to each other and thereby ensuring not only better mutual
inductance but also almost similar inductance.
- farhan