[TowerTalk] [Bulk] FW: Coax lengths for phasing a stack

[Grant Saviers]

Thanks to all for the comments.

I did model pattern change for feedpoint phase differences and found 
that at even with one feed offset +/- 45 degrees the lobe angle is 
unchanged at 9 deg for my 20m stack (127' + 67') and the peak gain 
deteriorates by about 0.6db.   (17.75 to 17.15dbi).     For 20m the edge 
to edge wavelength delta is 0.025wl (9 degrees) which will have 
negligible affect on gain and pattern if I used the 1wl offset in 
feedline lengths, one half that if I used 14.175Mhz as 1 wl.  A 4.5 
degree band edge feed phase error models as 0.00 db change in Pro4.   
For 15m and 10m (28 >29Mhz) there are single digit degree wl change.  
The antennas are identical OWAs so swr variation is not a concern.  btw 
60' of ldf4 is 0.12db loss at 14Mhz.  So it seems either "same length" 
or 360 deg phase shifts both work, with more loss with added coax to 
make equal lengths.

Each antenna is set up on its own ring rotator and a transformer/relay 
box is at the tower base for azimuth diversity / selection of BIP/T/B.   
I have a 1200' spool of LDF4, but adding a phasing length and coiling it 
up at the tower base when it isn't really needed seems like a waste of 
db's, $, and makes clutter.

73,
Grant KZ1W
"a few tenths of a db here and a few tenths there and pretty soon there 
is -3db"

On 2/8/2015 7:56 PM, Matt wrote:
> Spot on John.
>
> As the phase angle between a single stacked pair increases, the high angle
> lobe starts to grow, and the low angle lobe starts to deteriorate... but not
> at all in a linear manner - quite negligibly for all practical purposes even
> at +/-45 degrees phase angle.   At 90 deg phase angle the degradation
> between the main and second lobe is only down by about -3db.   I will note,
> however, that the deep nulls between the main lobes do start to fill in.
> At about 150 degrees phase angle or so, the high angle BOP lobe really
> starts to really take over.
>
> Matt
> KM5VI
>
> -----Original Message-----
> From: TowerTalk [mailto:towertalk-bounces at contesting.com] On Behalf Of
> john at kk9a.com
> Sent: Sunday, February 08, 2015 9:12 PM
> To: towertalk at contesting.com
> Subject: Re: [TowerTalk] Coax lengths for phasing a stack
>
> Phase in antenna stacking is not that critical, you can be out 20 degrees
> and not see much difference in the pattern. Using equal lengths of feedline
> is so much easier and in most systems it is not cost prohibitive. According
> to my modeling software even if the feedline to one antenna is 100 WL longer
> the pattern does not change. Does this seem correct?
>
> John KK9A
>
>
>
>
> To:	towertalk at contesting.com
> Subject:	Re: [TowerTalk] Coax lengths for phasing a stack
> From:	Jim Brown <jim at audiosystemsgroup.com>
> Reply-to:	jim at audiosystemsgroup.com
> Date:	Sun, 08 Feb 2015 10:23:01 -0800
>
> On Sat,2/7/2015 6:48 PM, Grant Saviers wrote:
> I'm ruminating about what lengths I need for 5L over 5L stacks to equalize
> the antenna feedpoint phase. The literature seems thin and the common answer
> is "make the feed-lines the same length", which I am certain will work.
>
> Yes.
>
> However, when the stack separations are sufficient, there may be more than 1
> wavelength of coax needed to equalize the lengths. e.g. 60' separation on
> 20m at 0.89 vf (LDF5) is just about 1 wl. It seems to me there is little
> downside with ham radio modulation systems to phase at 0 (equal lengths) or
> +/- n * 360 degrees (where n is 1 or 2). Perhaps some cw waveshaped
> attack/decay envelopes will change a tiny bit, but that seems like a slight
> downside.
>
> It's important to understand that PHASE is NOT a two-valued function, it is
> a CONTINUOUS function, measured in degrees, and it is a function of
> frequency and time. That 1 WL line will only be 1 WL at the frequency for
> which it is cut; either side of that, the drive to the two Yagis will be
> increasingly out of phase. The difference may or may not be enough to
> matter. The effect would be to tilt the vertical pattern, and would be most
> visible in the nulls. Equal length lines has the advantage of making them in
> phase across the band.
>
> 73, Jim K9YC
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>