Any tricks to winding high ratio, high impedance RF transformers

Here's the shortest array of, Four 25ft antenna wire's on the ground, each antenna has the amp on it. A little description may help, the top has a the feed lines labeled X and Y, going to Ls and Cs that phase two antennas. And then the same for the other two antennas. Then the two signals are phased together in another set of Ls and Cs. Then it goes to the radio.

Sorry it's a download of a pdf from Dropbox. Maybe there is a way to get the pdf as a direct link, but I didn't find it.

Btw, I took one of the very large inductance *CMCs, I removed some turns on one side and used it as a transformer. I adjusted the turns on the primary so it measures about R = 400? and XL= 70? while feeding the amp. I fed it with a 380? source and measured the output. I have two separate circuits, antenna, matching transformer, Feed Line, Radio and antenna, stepup transformer , amp, Feed Line, radio. I can easily switch from one circuit to the other by applying power to the relays. I adjusted the turns on the CMC/transformer so the gain, over the Matching transformer, is 17db at 200kHz and drops down to 14db at 1.7MHz. I then connected it to the antenna. I can report what people here and what I have read for 25 years, an amp is no use in the AM band, unless you have a very low output antenna. It is totally worthless in regard to signal to noise, on the other hand I'm glad to see no negative affects, no mixing products. I'll listen tonight when I have a lot of bigger signals.

Mikek

Phasing together just 2-4 omnidirectional antennas, you can get a single deep null in the radiation pattern or some wide lobes, if the antennas are quite close together. Moving the antennas several wavelengths from each other, you will get a large number of narrow side lobes like "fingers".

Might it be better to use a combination of an L-match (adjustable L and C) or a T- or pi-match, and a lower-impedance 1:4 or 1:9 bifilar transformer? If you can limit the transformer impedances to relatively low ones, I'd think that the parasitic/distributed capacitances of the windings would be less of an issue.

See my next post and the paper I referenced. If the paper is correct, the pattern is much better than you describe. Never mind, here's the link,

The need some distance from each other but not wavelengths.

Mikek

I would have even less idea of what I'm doing with L-match, t or Pi, than doing what I have already done. I tried tuning out the 70? of XL, I measured looking into the CMC/transformer, it reversed the direction of the gain slope. So I just left it as it was.

The scan I ran, showed the antenna has 10 ? to 80 ? capacitive reactance 550kHz to 1700kHz. R = 300? to 360? over the same freq range. That's not much reactance, probably not worth monkeying with. Watch the frequency scale, 200kHz to 4MHz.

Mikek

Mike, you should get a book on matching circuits. My recommendation is the ARRL Handbook. In my copy (year 2014) the most relevent chapters are Chapter 2, Electrical Fundamentals and Chapter 5, RF Techniques.

Hi Tauno, I'm sure I could learn something, But is it worth it to do more than just a simple matching transformer when the antenna looks like this? Is it really possible to flatten the inductive line, or is it just moving it up, so that it swings above and below the 0 ohm line

That brings up a thought, I need to look at the antenna after my matching transformer and see what that graph looks like.

Mike

Check "Antennas" by John D Kraus, it contains a whole chapter "Arrays of Point Sources" which explains what you can do by phasing individual antenna elements.

You are aiming to such impedance and frequency levels that a wide-band transformation cannot be done with realizable components.

Please get and read the basics, you'll save plenty of dead ends later.

I haven't got out to the antenna, but I ran my antenna analyzer on the feed line that goes into the radio. So the system is, antenna, 330 ohm to 100 ohm matching transformer, 100 ohm feedline, Common mode suppression circuit and a 100 ohm to 50 ohm matching transformer. (nothing to do with the amp) Not to bad, but has a problem at 563kHz and down.

Mikek

I agree that high impedances at RF* are a problem (I just learned this, probably relearned) I'm not sure why the CMC, converted to a transformer worked as well as it did. I was looking for 8 to 10 ratio and got 5 to 7, I'm happy with that.

Ya, I need a some math courses.

Mikek

• Even hard to do at audio frequencies, but I do have an audio matching transformer That is 15,000 ohms to 1,000,000 ohms. 100Hz to 3000Hz + of - 1db. (UTC-15)

I see the file has been downloaded several times, I have a 20 other antennas he built, The first 4 wires on the ground was with 95ft wires. Has some with Deltas also. If any one is interested, in any of these, my email is good.

Mikek

That jump is rather extreme.

I'll second that opinion. The coupling is more "loose" if not bifilar.

That's why I posted it, I don't really know why it happened. I verified it several times. So it wasn't a reading error, or some other operator error.

Mikek

Not just a theory; WWII German aircraft communications used Wollenweber antenna technology (a bunch of operators with switches tried combinations of elements by connecting taps into a maze of dipoles-with-preamplifiers until they could talk to the fighter pilot).

whit3rd wrote in news: snipped-for-privacy@googlegroups.com:

Weird diversity.

Then there are side-looking radars, aperture synthesis radars and radiotelescopes, and so on. There are interesting techniques (rather like a 2D version of the Golomb ruler) that let you get a fully dense set of Fourier components from a very sparsely filled aperture, if you get to pick the receiver locations. It does cost SNR, but saves beaucoup hardware.

Cheers

Phil Hobbs

And, this technique is how phased-array radar systems such as PAVE PAWS work. With the right combination of phase shifts (horizontal and vertical) a system like this can aim or "sweep" a directional beam across a wide arc.

It's 2020 for the AM BCB why not just use op-amps and a ADC to sample the whole AM band and store a week's worth of data at a time to a large SSD and then extract whatever you like with DSP. The "front end" could just be a video diff amp like the LT1194 with an input bandpass network sort of tuned to broadly cover the AM band, one input to an indoor longwire antenna the other to earth ground. the only L you would need is maybe for antenna loading coil.

Like a radio telescope for mariachi music and right-wing blah blah shows.