Phasing Harness

[Raccon]

Assume two antennas that have to be connected together:

Am I right to assume that a phasing harness avoids the 3dB loss as compared to a power splitter? What's the typical insertion loss of a phasing harness (at around 400MHz)?

 

[N_Jay]

Assume two antennas that have to be connected together:

Am I right to assume that a phasing harness avoids the 3dB loss as compared to a power splitter? What's the typical insertion loss of a phasing harness (at around 400MHz)?

3db Gain!

Of coursed the antennas have to be the same and ther gain is only for waves in phase, an dthe loss will be 60 to 100 dB on signals that perfectly cancel.

 

[Al42]

IOW they'll become a directional array and you'll receive signals in 2 of the 4 quadrants, losing signals in the other 2.

 

[N_Jay]

IOW they'll become a directional array and you'll receive signals in 2 of the 4 quadrants, losing signals in the other 2.

Of course you started with the UNSTATED ASSUMPTION :evil: that the antennas were omnidirectional, vertical and spaced a multiple of 1/4 wave length apart.

Lots of other things can happen if you use different assumptions!

 

[kb2vxa]

Hi all,

What isn't stated are the details, that is the intended purpose and the antennas involved. Shall we wait until they're provided before trying to shoot a black cat in a coal mine at midnight? BTW, another term for "phasing harness" is power splitter, the phase relationships are understood. Go into a ham radio store and ask for a phasing harness and see the puzzled look on the counterman's face when you use a CB term. If true directional phasing of omnidirectional verticals is the goal the term is coaxial delay line(s) and it's more than a "harness".

 

[Raccon]

Thanks for all the responses. Just to clarify: the antennas connected would be directional (yagi or panel), not omnis. Intent is to cover two areas (e.g. along a road) with maximum gain in both directions from one transmitter.
The receive antennas would be of the same type, requiring some form of combining before feeding the signal into the receiver (no duplexers used).

BTW, another term for "phasing harness" is power splitter, the phase relationships are understood.

I only know power splitters as a small device, typically without any cables attached. I have also seen power splitters that can divide the power unequally.
A phasing harness however, based on my understanding, consists of various coaxial lengths with varying impedances, thus it seems to be different to a power splitter.

 

[N_Jay]

Thanks for all the responses. Just to clarify: the antennas connected would be directional (yagi or panel), not omnis. Intent is to cover two areas (e.g. along a road) with maximum gain in both directions from one transmitter.
The receive antennas would be of the same type, requiring some form of combining before feeding the signal into the receiver (no duplexers used).

BTW, another term for "phasing harness" is power splitter, the phase relationships are understood.

I only know power splitters as a small device, typically without any cables attached. I have also seen power splitters that can divide the power unequally.
A phasing harness however, based on my understanding, consists of various coaxial lengths with varying impedances, thus it seems to be different to a power splitter.

If you are pointing the directional antannas in two different directions than you don't need a phasing harness, only a power splitter or multicoupler (TX vs. RX). (A Phasing harness will also work.)

You will see about a 3 to 3.5 dB loss in gain on each antenna.

You may get some peaks and nuls of the gain in teh antennas have overlapping patterens.

 

[kb2vxa]

Hi Raccoon and all,

"I only know power splitters as a small device, typically without any cables attached. I have also seen power splitters that can divide the power unequally."

OK, now I see what you mean, the sort of power splitter that's used in commercial (usually SHF) applications. I've seen them in catalogs, rather pricey little buggers.


"A phasing harness however, based on my understanding, consists of various coaxial lengths with varying impedances, thus it seems to be different to a power splitter."

The sort of power splitters hams use for stacking Yagis are that sort. The length of the coax is dependent on the frequency band and the impedance selected so that it acts like a transformer matching the antenna impedances to a common feed point. For example if you simply paralell two 50 ohm antennas the feed point impedance will be 25 ohms so the coaxial transformer brings it back to 50 at the common point.

Since the latter is the cheaper of the two and easy to construct I suggest that's the way to go. If you don't know the math and it's a bit much to get into here the commercial unit is the obvious choice. Naturally you want one that splits equally without attenuation, then all you loose is 3dB plus insertion loss, the same as with coax. In any case each antenna should have sufficient gain to overcome the sum of the losses or you'll be operating at a deficit. I was going to write "total loss" but caught myself. (;->)

Just for information, first of all get rid of that CB term "phasing harness" to avoid confusion. Coaxial delay line(s) is the proper term, let me explain. If you feed two omnidirectional antennas (to keep things simple) in phase, that is with two equal lengths of coax and they're 1/4 wave apart the result will be an end fire array, a figure 8 pattern off the ends. If you add a half wave of coax to one you now have them 180 degrees out of phase (one delayed by 180 degrees) and it fires a figure 8 off the sides. A quarter wave of coax delays the phase by 90 degrees. Varying the phase angle electrically or physically (spacing) changes the radiation pattern accordingly.

 

[Al42]

The sort of power splitters hams use for stacking Yagis are that sort. The length of the coax is dependent on the frequency band and the impedance selected so that it acts like a transformer matching the antenna impedances to a common feed point. For example if you simply paralell two 50 ohm antennas the feed point impedance will be 25 ohms so the coaxial transformer brings it back to 50 at the common point.

Although it's usually done the other way - the phasing sections transform the antenna impedance to 100 ohms - paralleling two of them brings the impedance back to 50 ohms.

 

[DaveH]

Although it's usually done the other way - the phasing sections transform the antenna impedance to 100 ohms - paralleling two of them brings the impedance back to 50 ohms.

Correct, using 1/4 wavelength of 70.7 ohm coax (73 is close enough) from each 50 ohm antenna will transform it to 100 ohms at the common point, which in parallel as pointed out will be back to 50 ohms.

If you wanted to transform 50 ohms to 25 and put the two in series, it would take 35.3 ohm coax which is not common. I'm not sure the series thing would work, I've never heard of it done that way; but anything's possible.

Dave

 

[N_Jay]

Although it's usually done the other way - the phasing sections transform the antenna impedance to 100 ohms - paralleling two of them brings the impedance back to 50 ohms.

Correct, using 1/4 wavelength of 70.7 ohm coax (73 is close enough) from each 50 ohm antenna will transform it to 100 ohms at the common point, which in parallel as pointed out will be back to 50 ohms.

If you wanted to transform 50 ohms to 25 and put the two in series, it would take 35.3 ohm coax which is not common. I'm not sure the series thing would work, I've never heard of it done that way; but anything's possible.

Dave

You can build the phasing harness (delay lines)(Coaxial Transformer) out of any impedance cable as long as you take the cable characteristics into account.

 

[kb2vxa]

Hi again,

VXA:
"For example if you simply paralell two 50 ohm antennas the feed point impedance will be 25 ohms so the coaxial transformer brings it back to 50 at the common point."

Al42:
"Although it's usually done the other way - the phasing sections transform the antenna impedance to 100 ohms - paralleling two of them brings the impedance back to 50 ohms."

That's basically what I was trying to say, thanks. If it were all half waves of 50 ohm coax (no impedance transformation) the common point impedance would be 25 ohms but by making the antenna look like 100 ohms the common point is 50.

Cushcraft (?) makes an unusual lazy H 6M beam (2el double driven) using an entirely different principle. Each element is gamma fed but without the series capacitor, the result is a 100 ohm load. One element is fed directly while a rather odd bit feeds the other so the paralell impedance is 50 ohms. I say "odd bit" because it's not coax, rather two 14ga insulated conductors inside shrink tubing resembling Romex electrical cable without the protective ground wire. Gee, maybe we should stop using coax and use Romex from the local supply house? (;->)

 

[Al42]

Cushcraft (?) makes an unusual lazy H 6M beam (2el double driven) using an entirely different principle. Each element is gamma fed but without the series capacitor, the result is a 100 ohm load. One element is fed directly while a rather odd bit feeds the other so the paralell impedance is 50 ohms. I say "odd bit" because it's not coax, rather two 14ga insulated conductors inside shrink tubing resembling Romex electrical cable without the protective ground wire.

Otherwise known as a parallel feeder, twin-lead or a number of other names.

 

[DaveH]

You can build the phasing harness (delay lines)(Coaxial Transformer) out of any impedance cable as long as you take the cable characteristics into account.

Not exactly sure what you meant by this one N_Jay.

To properly transform 50 ohms to 100 ohms using a quarter-wave length, the coax impedence should be the "geometric mean" of the two you want to match; in this case multiply 50*100 and take the square root (70.7 ohms). Similarly for 50 ohms to 25 ohms, take the square root of 50*25 (~35.3 ohms). That's where the numbers come from.

You could use different lengths and/or different impedence, but it would complicate things. You also want to transform 50 ohms to 100 ohms resistive (no reactive component). You could end up with 100 ohms or 25 ohms with reactive component which would have to be tuned out with some other technique.

Dave

 

[N_Jay]

You can build the phasing harness (delay lines)(Coaxial Transformer) out of any impedance cable as long as you take the cable characteristics into account.

Not exactly sure what you meant by this one N_Jay.
. . .

You could use different lengths and/or different impedence, but it would complicate things. . . .


Dave

That is exactly what I meant.

Don't fear reactive loads (as long as you understand them.)