Whats all this impedance stuff anyhow?

[dnyberg2]

Okay, so my short experience in radio has taught me that the input and output of every RF system is designed around some certain impedance. Typically 50 ohms or 75 or 300... So you take a transmitter and hook it to a tuned antenna with some sort of feedline like coax.

The coax is the same impedance as the design of the output of the transmitter and the antenna, let's says this is all in a 50 ohm world for this example okay?

Now you decide you need to move the transmitter a little further away from the antenna feed point, so you go to your shack and grab a roll of coax that gets the job done. This new coax is 10 feet longer than the one you were using. Let's say for argument sake its RG8.

The new longer length has a little more loss to it than the shorter length, I get that much. But what else does this new longer coax have going on than a bit more loss and why?

Does the longer coax now phase shift the RF to some other phase angle than the shorter one?

I assumed all these years that a slight difference in length of coax does little but change the loss in this perfect 50 ohm matched system.

I know different lengths of coax are used as tuned lines when phasing two antennas together and such or making a filter even, but the length of coax in those cases are at some multiple of the wavelength or present a calculated impedance, delay or phase change right?

What would you say if I told you that I have an RF system that seems so impedance dependent from source to load that even a slight change in the length of transmission line (inches) seriously affects the overall performance of the entire system from RF source to antenna load?

Doesn't that smack of some kind of design flaw? If an RF system were designed to act that way on purpose, what does that say about the system?

As always, your comments are greatly appreciated.

 

[skeptic2]

The new longer length has a little more loss to it than the shorter length, I get that much. But what else does this new longer coax have going on than a bit more loss and why?

Yes, the added length will phase shift the output but unless the output is phased with another antenna, the change in phase has no effect.

What would you say if I told you that I have an RF system that seems so impedance dependent from source to load that even a slight change in the length of transmission line (inches) seriously affects the overall performance of the entire system from RF source to antenna load?

I would say this is not at all uncommon. It means that at least one of the source, cable or load is not matched to the others.

Doesn't that smack of some kind of design flaw? If an RF system were designed to act that way on purpose, what does that say about the system?

It is doubtful it was designed that way on purpose but it suggests that the system was designed by someone who didn't understand impedance matching.

 

[vk6kro]

If the load and the feedline are both 50 ohms, then the input to the feedline will also be 50 ohms and the length of the feedline will not matter.

However, this is a rare luxury.

More often an antenna system will have a tuned feedline.

For example, an antenna consisting of 1 wavelength loop of wire will have an impedance of 130 ohms or so.

If you fed this with 50 ohm cable there would be a bad swr, but also, the length of the feedline would determine the input impedance of the feedline.

We could feed the loop antenna with a quarter wavelength of 75 ohm coax and get an impedance of 43 ohms at the feedpoint.
This is a lot closer to 50 ohms than the 130 ohms we started with, so this would be a better match for the transmitter, but the length of the feedline will be very important. It must be a quarter wavelength long.

(Z of feedline)² = Z of load * Z of feedpoint.
so 75 * 75 = 130 * Zin
so Zin = 43 ohms

If we had some 80 ohm cable, we could get a perfect load for the transmitter, but 80 ohm cable is not commonly available.

 

[Jiggy-Ninja]

Characteristic impedance does NOT change with length, it is a constant for that specific type of cable.

That's what my textbook says, at least.

Changing the length will affect the ohmic losses, but that's usually very, very small compared to the impedance.

The length will affect the propagation delay through the cable, and therefore the phase of the signal, but as skeptic said, unless you're trying to synchronize two signals the phase doesn't make any difference at all.

 

[Bob S]

vk6ro points out the importance of coax length and quarter wave transformers. Read about quarter wave transformer here

http://en.wikipedia.org/wiki/Quarter-wave_impedance_transformer

or in Terman's EE book.

Bob S