Towertalk
[Top] [All Lists]

Re: [TowerTalk] 80m 4 square advice sought

To: bob.5b4agn@gmail.com, towertalk@contesting.com
Subject: Re: [TowerTalk] 80m 4 square advice sought
From: TexasRF--- via TowerTalk <towertalk@contesting.com>
Reply-to: TexasRF@aol.com
Date: Thu, 29 Oct 2015 07:56:16 -0400
List-post: <towertalk@contesting.com">mailto:towertalk@contesting.com>
Hi Bob, no experience here with your style antenna but do have a question:  
doesn't the physical spacing of the verticals have an effect on phase  
relationship between them?
 
It seems there could be an effect on feedpoint phase due to the change in  
feedpoint reactance of individual verticals from the lower frequency to the  
higher frequency as well.
 
Current forcing seems a good idea but achieving the deep nulls you are  
looking for also requires the proper phase relationship. We know that combining 
 two signals 180 degrees out of phase with equal amplitudes causes complete 
 cancellation. With equal amplitudes and a 10 degree phase error, the  
cancellation is reduced to 15 db using 20 log sin of phase angle. I forget if  
this is a 20 log or 10 log thing. If 10 log, then the number is worse, 7.5 dB 
 cancellation.
 
I guess I am wondering if you might be chasing a ghost here. Perhaps  some 
antenna modeling would be helpful. 
 
If I am just stating the obvious here I apologize.
 
73 and happy hunting!
Gerald K5GW
 
 
 
 
 
 
In a message dated 10/29/2015 5:08:09 A.M. Central Daylight Time,  
bob.5b4agn@gmail.com writes:

My  thanks to those who replied directly.  It has been suggested I  first
adjust the phasing lines for low band.  It has also been  suggested I leave
well alone and accumulate experience in behaviour of the  array before
making further adjustment, given perceived F/B and F/S are so  dependent on
signal direction and arrival angle.  Sadly I don't feel I  have the luxury
of the latter as rains might come anytime.  When they  do they tend to come
in the form of a deluge which can make the antenna  site an unworkable
quagmire for months.

In view of the likely  deterioration in the weather I have pressed on.  I
provide details of  my findings for those who may be interested and in the
hope they might  prompt further suggestions.

I felt doubt over the likely benefit of  extending the phasing lines.  As
the phase relationship between the  four verticals is determined by the
hybrid it feels like the relationship  will be preserved so long as feed
lines to the verticals are of equal  electrical length.  So far as I am
aware 1/4 wave lines are used is by  virtue of their current forcing
property which can to an extent compensate  for differences in
characteristics of the four verticals.  My four  verticals appear remarkably
similar both physically and electrically.   I did anyway extend the phasing
lines by 61cm to make them 1/4 wave at  3525kHz.  This adjustment is
equivalent to 3 degrees of arc or  3.3%.

I discerned no difference in F/B or F/S as a consequence of the  extended
phasing lines.  Percentage power dumped to the 50 ohm load  changed only
slightly as follows:

Frequency (kHz)

Original  phasing lines

+61cm per  line

3500

5%

5%

3580

3.5%

3.6%

3800

11.5%

13%

So  far as I can tell the effect of adjusting phasing line length from mid
band  to low is marginal.  A mid band length appears fine.

This morning  I have taken the hybrid back to my workshop for a more
detailed analysis to  see if I can spot anything which might explain my loss
of 10dB in F/B and  F/S.  Measurements were made using an N2PK VNA.  All
hybrid ports  were terminated in 50R.  The unit is configured such that  
when
controller power is absent the array fires NW.  All measurements  were made
with this setting.

Phase angle relative to  source.


Port

3500kHz

3650kHz

3800kHz

NW

-128.0

-130.5

-133.0

SW

-49.0

-50.7

-52.1

SE

62.6

60.5

58.6

NE

-47.5

-48.9

-50.3


Normalised  to SE (rear)  port


Port

3500kHz

3650kHz

3800kHz

NW

-190.6

-191.0

-191.6

SW

-111.6

-111.2

-110.7

SE

0

0

0

NE

-110.1

-109.4

-108.9


Port  amplitude relative to source  (dB)

Port

3500kHz

3650kHz

3800kHz

NW

-4.75

-5.03

-5.25

SW

-5.66

-5.82

-5.96

SE

-6.74

-6.95

-7.15

NE

-5.85

-6.15
-6.35

In theory the  hybrid coupler provides outputs at 0, -90 and -180 degrees
with equal  amplitude 1/4 of input or -6dB.  In practice mine isn't quite
like  that but port amplitudes are within +/- 0.25dB between 3500 kHz and
3800  kHz and phase relationships within +/- 0.6 degrees.  I see  nothing
here to clearly account for my loss of 10dB F/B and  F/S.

Perhaps it is simply the signals I have heard since I low banded  the
array?  This doesn't feel too likely as I found plenty of signals  to
demonstrate 20dB at the top of the band and none so far at the  bottom.  I
didn't have many ideas on the cause of my problem at the  start and I am
fast running out.

I have rigged a small battery  powered beacon TX which I will deposit in a
hedge approximately 1km SW of  the array.  This will at least give me a test
signal during daylight  hours when adjustments can be made.  Here in Cyprus
80m has nothing  but noise on the daylight side of grey line so on air
checks can only be  made when adjustments cannot.

If any reader spots something untoward in  my findings please let me know.
Additional suggestions on where I might  look next will also be appreciated.

Many thanks.

73 Bob,  5B4AGN

On 28 October 2015 at 08:14, Bob Henderson  <bob.5b4agn@gmail.com> wrote:

> In the weeks before CQWW SSB I  installed my first 80m 4SQ.  I am mostly a
> CW guy and want the  antenna to work well at the bottom of the band.
>
> The vertical  elements are Titanex V80S verticals which I bought having an
> 80m 4SQ  in mind some 15 years ago.  It takes me a while to get around to
>  things at times!  The V80S is set for mid EU band i.e. 3650kHz and  
without
> adding extra alloy or cutting some off there is no facility  for
> adjustment.  I started out with the idea I would rig the 4SQ  for 3650 and
> see how effective it turned out to be at 3500 and  3800.  The sides of the
> square are 20.54m.  The four 1/4  wave phasing lines have been cut for 
3650
> and the array is fed via a  home built hybrid coupler.
>
> 4SQ experts will likely have  already spotted my mistake. I overlooked the
> inevitable rise in  resonant frequency of the vertical elements due to
> mutual  coupling.  Anyway in some respects this was kind of lucky or  
would
> have been had we not had serious electrical storms last  weekend.
>
> First tests were very promising but with best results  at 3800 not
> 3650kHz.  F/B was a good 20dB and F/S not much short  of that.  Power 
dumped
> to the 50R load was around 2% but rising  to around 15% at 3500 where both
> F/B and F/S were significantly less  impressive at around 10-12dB.
>
> With CQWW SSB visible now only  in the rear view mirror I have turned my
> attention to shifting the  array down band.  I reasoned I would need to 
pull
> the individual  vertical resonance down by around 250kHZ to 3400kHz for
> optimum  operation in the CW band.  Taking down each vertical then adding
>  further alloy would be a far from easy option so I have instead added  a
> small amount of base loading.  In fact 6 turns of wire around  the bottom
> 30mm diameter insulator on each vertical did the job.   The effect of the
> adjustment is as follows:
>
> Power  dumped
>
> 3500 = 5.0%
> 3580 = 3.5% (minimum)
> 3800  = 11.5%
>
> The above looks acceptable but....
>
> F/B  and F/S are around 10dB across the band so far as I have been able to
>  determine.  So to my question..........
>
> What step do I  best take to improve the F/B and F/S in the CW part of the
> band.   My options appear limited to adjusting the L&C in the hybrid and
>  adding length to the 1/4 wave phasing lines.  My gut feel tells me to  
hit
> the hybrid first.  Does this make sense or is there something  else I am
> overlooking?
>
> Advice from the experienced  will be much appreciated.
>
> Thanks.
>
> 73 Bob,  5B4AGN
>
_______________________________________________



_______________________________________________
TowerTalk  mailing  list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

_______________________________________________



_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk

<Prev in Thread] Current Thread [Next in Thread>