Help: how to design air-core transformer?

Like a transmission-line transformer, which can even be made without ferrites, if you sacrifice the low end.

On 22 Nov 2004 15:08:36 -0800, snipped-for-privacy@tom.com (Jim) wroth:

It can't be done. Try another solution.

Jim

Experience. (With trying to create such a design that's viable, for most applications that I've tried anyway.) What's wrong with adding a ferrite into the mix?

Right. But the output of a transformer (two coupled inductors) ends up being controlled more by coupled capacitance than by magnetic field at those frequencies. Hmm, perhaps if one were to interpose a shield... with approriate slits... Hmm...

Exactly. A common practice is to wind the windings by twisting them together, bifilar or trifilar, etc. This increases the coupling. At several hundreds of MHz, the turns may be just a few or one turn. Here's a tutorial on how to wind one.

James Meyer posted:

I hope to design an air-core transformer which may operate at 100-500 MHz.

It can't be done. Try another solution.

I don't have time to get deep into this, but I don't see why it can't be done. Probably a couple turns tuned on one side. Why do you dismiss it as impossible?

Don

But that's with cores. If there's no core, and the length is made long enough for the desired response, you'll end up primarily with a coupling capacitor. Or a transmission line.

I observed two unwanted phenomenon:

1. direct capacitive coupling between primary/secondary coils. It interferes with the mutual inductance, and makes the transmissivity of the transformer change significantly with frequency.

1. Insertion loss is higher than 10dB.

I googled air core transformer before i posted my question. It is weird that the concept is mentioned in many places, but I can never find a real one which works in VHF/UFH band except transmission-line type.

My purpose is for impediance matching, from 50 ohm cable to a 0.5 ohm load.

To my understanding, air-core transmission-line transformers have fixed conversion ratio, e.g. 1:1, 4:1 or 1:4, is it right? What i need is a ratio of 10:1.

I saw some companies provide RF transformer with ferrite coil inside. The insertion loss is only 1-2dB, however, for UHF band, the maximum ratio is 4:1 or 6:1 only.

Thank you all again! And hope to hear from you again.

RF inductors are often wound without cores, especially at VHF and UHF.

Leon

What is your loss like WITH a core? Maybe the core is not your problem. What application/environment?

Google 'air-core transformer', 'air cored transformer', 'coreless transformer', 'contact-free transmission', 'contactless transmission', 'transcutaneous energy' etc.

RL

The original posts sounds like he want it to be broadband 100 - 500 MHz. Not obvious why you would want that, unless it was for some sort of instrument. Half of the FM band on one end, and 5 UHF TV channel on the other end.

Tam

Tam

You could simply put a 50-ohm resistor in series with your o.5-ohm load, to prevent reflections and insure a fixed known current at all operating frequencies.

As far as wideband transformers are concerned, the common 4:1 transmission-line transformers (get Sevick's books) can be made for 50 to 12.5 ohms and then 12.5 to 3.25 ohms. The transmission line in each stage must have an impedance that's the geometric mean between Zin and Zout, i.e. 25 ohms and 6.5 ohms. You'll have to make the transmission line yourself. I handmade a 16:1 wideband transformer this way, and it was a tricky task, taking several days. But it worked well.

In theory you can do the next step from 3.25 to 0.81 ohms, but it's nearly impossible to make 1.6-ohm transmission line.

Perhaps the spot for a resistive match is here, at the 3.25 to 0.5-ohm level. At least you will have quadrupled the delivered current into 0.5 ohms with your 16:1 transformer.

If you want to play a narrow-band matching game, no problem, have at it.

On 23 Nov 2004 05:24:48 GMT, snipped-for-privacy@aol.com (Dbowey) wroth:

Because of the 100-500 MHz specification. A bandwidth that wide precludes a simple air-cored transformer.

But then, you mention tuning. If you had started out with a more complete set of specifications, you would have gotten a more complete reply. Is it really necessary to cover 100 to 500 MHz? What is the step up or down ratio? How large or small does it need to be? What's the problem with loss? How lossless does it have to be?

Jim

On Tue, 23 Nov 2004 10:02:42 -0000, "Leon Heller" wroth:

Low loss wideband RF transformers are almost never wound without cores.

Jim

There is a very good Motorola applications note discussing wide band baluns, transformers, etc. both ferrite wound and air wound..

The treatment of the structure as a transmission line is appropriate. Actual coax and the bi-filar and tri-filar wires are used. So how wide a band width are you going to use?

Marc

RF power and components FA here

I was thinking of a narrow-band transformer within the range the specified range.

Leon

If I　simply need a narrow-band transformer, e.g. 300MHz with bandwidth>1MHz, what is the best method? I don't need continuous tuning, so I can make a set of such transformers to cover the range of

need

Looks to me like you need more lie a hundred to one.

You can do 9:1 or 16:1 with bifilar wound air core coils.

air-cored

very good

then increase

increase

I agree about the ferrite, but he needs a 100:1 impedance change so a