Phasing harness measurement

[Scott N7SS]

I need to build some phasing cables but am not clear exactly how to
measure them. Dealing with velocity factor and wavelength is simple
enough but I'm not clear where to measure since the connector changes
the length of the cable. I need to build both N and PL-259 cables. I
plan on using crimp connectors (they will be indoors).

How exactly do I measure the length of these cables?

Thanks,

Scott N7SS

[Jim Garver]

Having done this a few times to make 75 ohm phasing sections for twin
stacked 50 ohm antennas, I can say this:
It is never easy, gets way harder at 432 Mhz and above, and you can't trust
the book velocity factor values!

However if you are going to measure according to book values then use
tip-to-tip of the connectors, whatever they may be.

To more accurately check my harness length I have used:
- Grid dip meter method - small loop and shorted other end
- Time Domain Reflectometry equipment!
- very small 50 ohm dummy loads on the cable ends and adjust lengths for
best VSWR.

Of these methods #3 works best and is the only one to use at 1296 Mhz and
above.
I made a nice 1296 harness made of 1/4" 75 ohm superflex this way. Cut and
try all day!

Make sure each antenna matches flat at 50 ohms then try your harness. If
the match is somewhat close then adjust ONLY ONE antenna to get the final
match.

Good Luck!
Jim K7YO

Scott N7SS

[Jim Klitzing]

Hi Scott,

If these cables are all to be the same length, just cut them that way
and all ahould be well. The connectors will add the same amount to
each one, and coax type/velocity factor will be irrelevant.

If you need them different lengths at multiples of 1/2wl, for example,
you'll have to have some instrumentation that can measure phase shift
(like a scalar analayzer or a slotted line).

Jim
W6PQL

[TdM Labs]

If you don't need different lengths to feed your antennas from a non
central point, make them all the same length, otherwise odd multiples
half lambda pieces. An inch or so error will have little noticeable
effect on 144 Mcs. If you can, keep everything very short by doing it
all behind your antennas...

GL phasing! Jeremy w7eme

[kb7...@donobi.net]

>End of the center pin, regardless of connector type. Deduct the dimension
from center of the tee fitting... as this will be part of the phasing
assembly. This only gets really critical if you are shifting 90 degrees
to create circular polarization from a pair of yagis...
If building 75 ohm harnesses to impedance match two antennas to one 50 ohm
feedpoint IN PHASE... as long as the two 75 ohm pieces are the SAME
length, any length error will only result in slight mismatch rather than
change in phase. Not really critical unless working with 432 and above.
I have found out the useable bandwidth of phasing harnesses longer than
1/4 wavelength
gets narrower, so the lengths become more critical.

Prepare the cables, shove the crimp fittings in place. Measure overall
length, THEN if all OK, crimp together. This way adjustments can be made
"before it's too late"...

Eric
KB7DQH

[Jim Garver]

KB7DQH> I have found out the useable bandwidth of phasing harnesses longer

than 1/4 wavelength gets narrower, so the lengths become more critical.

So true, and maybe much of the problem cutting accurate length matching
lines. A stacked pair of yagis must be at least 1/2 Lamda apart for good
horizon gain. The velocity factor will shorten the lines so I usually end
up with five to seven 1/4 wavelengths each line. Any slight error in length
is multiplied by the number of odd 1/4 wavelengths.
Here I am describing using odd 1/4 wavelength 75 ohm matching lines to stack
a pair of 50 ohm antennas in phase. The 1.5:1 SWR bandwidth of this kind of
matching network seems to be about 200 Khz? I did check it using 50 ohm
dummy loads but do not recall. Under 500 Khz anyway.

K7YO

[KJ6KO]

Why not just make two 1/4 wave lines from the "T" then connect the rest of
the way to the antennas with equal lengths of 50 ohm coax? There is no need
to run 75 ohm line all the way to the antenna. There are commercially made
harnesses done this way. The only problem is an extra set of connectors.
I really don't believe connectors loose as much as people think they do.
I once put a string of various adapters together over 19" long on an HP
network analyzer and couldn't get 1/2 db of loss at 440 MHz thru the mess.

73 de Greg KJ6KO

[Jim Garver]

> Why not just make two 1/4 wave lines from the "T" then connect the rest of
the way to the antennas with equal lengths of 50 ohm coax? There is no need
to run 75 ohm line all the way to the antenna.

Sorry for the confusion, I do not run 75 ohms all the way (usually*). I use
75 ohms only after the split to the twin antennas. The goal is to match two
50 ohm antennas effectively in parallel (25 ohms) to the 50 ohm transmission
line. It is different for other arrays combinations such as four antennas
where you use 50 ohms throughout.
There is also a way to match a pair of 50 ohm antennas using only 50 ohm
coax:
http://amfone.net/Amforum/index.php?topic=5199.0;wap2

WB7RSG told me they measured N connectors with network analyzers in the lab
at Triquent and found about .1db loss. Other sources indicate as high as
.75db loss. W7EME just solders the coax or hardline together without
connectors but he is better than me :-)

* sometimes I use CATV 75 ohm hardline. I adjust the antenna to match 75
ohms, then install a 62 ohm 1/4 wave match section at the radio end to get
50 ohms (very freq dependent).

K7YO

[Larry]

Over the years I've done it just about every way you can think of.

First off, I've measured, in a lab with calibrated equipment, hundreds
of N and other connectors over the years. It's a difficult measurement
to make because the "loss" in a properly installed N connector is less
than the resolution of a HP/Agilent network analyzer. Usually if it
measures more than 0.05 dB at 2 GHz, I throw it in the trash bucket. I
don't recall ever messing with them above 3 GHz or so, maybe it's worse
than I think higher up.

Even though the connector losses are no more than the same length of
cable it's installed on, I try to avoid using them if at all possible
due to cost. I scratch-built the array described below, including the
tower, az-el drive systems, stack frame and phasing harness, yagis and
preamp for less than $500. Using connectors would have nearly doubled
that cost.

When I built my 16 yagi 2 meter array in AZ, I had only one connector in
the entire feed system located at the center. The phasing lines were
multiple 1/4 wave length pieces of CATV 75 ohm hard line (1/2 and 1 inch
sizes) arranged into 2 way splitters. I.E. 16 into 8 into 4 into 2 into
1. I think I had some 9/4, some 7/4 and some 5/4 in the mix. I built a
144.1 signal source, modulated it with 1 Khz, built a return loss bridge
with better than 30 dB directivity and used that with a HP415B to cut
the lines to exact 1/4 wave odd multiples.

The lines were directly connected to the "T" match used on those yagis
and to each other by bits of hobby brass tubing drilled and formed to
fit. The whole connection was held together mechanically with small
aluminum straps and hose clamps then gunked up with PVC pipe cement and
sprayed with black Krylon. When the whole thing was assembled, I had a
measured loss in the phasing lines of less than 1/2 dB. My measurement
resolution was not very good so I couldn't claim better than that evne
though it was. The 1.5:1 SWR bandwidth was much less than that of each
individual yagi. If I remember right, it was on the order of 300 kHz.

For the present 8 yagi array, I used 1/2 wave multiples of LMR 400 for
the phasing lines because yagis have 50 ohm direct feed. The coax was
cut and measured with my HP network analyzer. A ferrite tube is placed
over the LMR400 to create a current choke and the LMR is attached
directly to the driven element. I have two 4 into 1 1/4 wave dividers
followed by a single 2 into 1 half wave divider, all with N connectors.
You can see most of these pieces in the photos on my web page.

Eliminating the connectors eliminates some of the guess work involved in
cutting the phasing lines because the reference plane is not always
where you think it is on some of these connectors. On 144 it really
doesn't matter much but on 1296 and above it becomes a problem quickly.
I have not yet managed to wire the coax up without connectors on 1926
phase critical systems. So I use connectors on 1296 and higher and
struggle with lengths. Maybe I obsess over accuracy too much, because I
see lots of systems thrown together with little concern for phase
accuracy and they all seem to work fine.....

The answer is to just use a very long single yagi and not have to worry
about phasing lines.

Larry

Larry - W7IUV
DN07dg
http://w7iuv.com

[KJ6KO]

Exactly what I was trying to say, you do not need to run 75 ohm all the way
from the split to the antennas. Only 1/4 wavelength in the direction of
each antenna from the split, then you can run ANY length of 50 ohm the rest
of the way to each antenna, as long as they are the same length. The 50 ohm
portion doesn't need to be any specific length, just equal. An example
would be if you have two 2m antennas 12' apart. Run your 50 ohm coax from
the rig to the "T", then two runs of 17" of 75 ohm coax in each direction
from the "T", then any two equal lengths of 50 ohm coax the rest of the way
to the antennas from the ends of the 75 ohm phasing lines. You only need
1/4 wavelength of 75 ohm coax in each direction to connect any two 50 ohm
antennas, regardless of how far apart they are, therefore eliminating the
bandwidth problem of trying to have multiple odd 1/4 wave sections all the
way to the antennas.
Maybe that is what you said in the first place and I misunderstood??

73 de Greg

----- Original Message -----
From: "Jim Garver" <k7...@centurytel.net>
To: <PNW...@googlegroups.com>
Sent: Thursday, August 14, 2008 9:13 AM
Subject: [PNWVHFS] Re: Phasing harness measurement

>

[KJ6KO]

[Jim Garver]

> You only need 1/4 wavelength of 75 ohm coax in each direction to connect
any two 50 ohm antennas, regardless of how far apart they are, therefore
eliminating the bandwidth problem of trying to have multiple odd 1/4 wave
sections all the way to the antennas.

Yes, you can do that, and some other ways too, like the commercial T-block
with only one 1/4 wave match line in it.
I don't see any advantages, and I don't see how the bandwidth changes.
For me the full length 75 ohm lines make the most sense; minimum connections
and splices, slight advantage of lower loss 75 ohm coax, and I have a bunch
of various 75 ohm coax.
I have also used 92 ohm RG-62 to match stacked 75 ohm antennas into 75 ohm
CATV line.

I sometimes have weather problems with most any connection. A properly
installed N-connector with silicon grease at the rubber seal and liquid
electrical tape where the cable enters seems to be most reliable for me.
Keep fingers crossed and severely limit outdoor connectors. A slight amount
of moisture in there will not be noticeable at 50 Mhz but will kill you at
432 and above.

Over and out,
K7YO

[frank bechdoldt]

I found that the bandwidth narrows as the legnth of the 75 ohm cable increases.  I once had a stacked pair of cushcraft antenas with 2, 5/4 wavelength 75 ohm coaxes fed to a T.  I changed it out to a M2 divider and 2, "exact factory cut (cable experts)" 10 foot pieces of lmr 400 ultraflexes and the bandwidth increased from about 600 khz to 2 mhz below 1.5 swr. 
 
I abandoned that set up some time ago and I find that the divider/ combo still beats another 3/4 9913( that number might be wrong) 75 ohm coaxes to a T for band widthy.  However I still use the T because its shorter and who needs fm on a pair of loops? Unless I put them out of phase and work the sats.
 
k3uhf

 


> From: k7...@centurytel.net
> To: PNW...@googlegroups.com
> Subject: [PNWVHFS] Re: Phasing harness measurement

[Jim Garver]

 

K3UHF > I found that the bandwidth narrows as the legnth of the 75 ohm cable increases.  I once had a stacked pair of cushcraft antenas with 2, 5/4 wavelength 75 ohm coaxes fed to a T.  I changed it out to a M2 divider and 2, "exact factory cut (cable experts)" 10 foot pieces of lmr 400 ultraflexes and the bandwidth increased from about 600 khz to 2 mhz below 1.5 swr. 
 
Oh, then no wonder I have had trouble cutting exact lengths on my multi-1/4 wavelength 75 ohm cables!

I guess the ¼ wave filter effect must be like cavity filters in series?  Duh…

K7YO