[NLRS] PSK31 on 2m SSB

[Gerald]

On Mon, 2005-08-29 at 13:39 -0500, Dan Larson wrote:
> On Mon, 29 Aug 2005 11:49:02 -0500, Gerald wrote:
> 
> >
> >You only need ONE direction towards MSP. You don't need to spin the
> >antenna for that dedicated path.
> >
> 
> True, if thats the only direction I ever wanted to talk...
> 
> >You are running horizontal polarization on both ends of the path?
> 
> Of course....
> 
> 
> >You loose at least a couple dB from RG-8X with the 30' run. A
> >significant amount on a marginal path.
> 
> But the savings going to LMR400 might only be about 1dB.

2 dB on a threshold path can change the S/N by several dB.
> 
> >Easier than running coax in a
> >window, get a long coax feed through and mount it in a 5/8" hole
> >permanently through the trailer wall.
> 
> Not if you consider objections from the XYL!

She needs a ham license.

> 
> >11 DB is a lot of gain for 6' and 6 elements.
> 
> Yeah, I know the simulation is optimistic. I figure about 9dBi/7dBd realistically.
> 
> Although, S-meters being what they are, I have received reports of 2-3 
> "notches" higher than using my omni vertical. "They" say an S unit is 6dB and
> most FM radios seems to indicate 2 S units per bar, then I can't be doing 
> too bad, but I sure as heck don't believe that I am getting 18dB gain either. 
> And yes the comparison is the the vertical portion of my beam at the same height.

Few S-meters are as stingy as 6 db per S unit and most FM S-meters run
under 2 dB per S unit seems like and they saturate early from being
displays of the first limiter current that saturates easily.
> 
> (Its a Horizontal / Vertical cross beam with separate feeds)
> 
> >Is that over isotropic over earth? 
> 
> Yes. IIRC. 
> 
> Its interesting how gain can appear to go up when earth is added, but then
> looking at the plots you can see why because you start getting lobes in 
> elevation plot and if you're lucky the big one points at the horizon.

But if you compare the lobes in elevation to the elevation plot in free
space you find all those lobes fit IN the free space elevation plot.
Those lobes are just the diffraction grating effect of one source and
its image below ground. The granularity of the lobes is greater when the
antenna is higher above ground.

I claim that all the NEC and MININEC programs have a fundamental error
over earth working their gain over isotropic. That is that with earth
involved, they compare the antenna field strength to the field strength
of 1/2 isotropic since the isotropic reference is a ground level and
they don't include the power from that isotropic source that is into the
ground. It can't reflect because its half embedded in the earth. (even
if it is only one epsilon in diameter).
> 
> 
> >Did you take into account element length effects of end
> >effect and boom effect? 
> 
> The design SW claimed to have taken the end effects into account.
> 
> >The optimizer programs never include boom effect
> >and sometimes are out in space on end effect.
> 
> Well, I designed it with QuickYagi 4.0 and modeled it with 
> MMANA 0.72. I used MMANA more as a verification of the design rather than
> to optimize it. It sure was crazy trying to model the gamma feed. 
> Quick Yagi was about 2-3db more optimistic than MMANA for the same design.
> MMANA didn't fare well trying to optimize a design that was created with 
> QY4, but when modeled and then plotted it showed the same pattern but
> the gain figures were about 2-3db less. QY4 always seem to optimize a design
> with a cleaner pattern than anything "optimized" by MMANA. Thats why I 
> eventually just used MMANA to model / plot only.

Most of the yagi optimizers work with an antenna element coupling model
from W8JK that is good only when the elements are nearly a half wave
long and it doesn't include effects of more than the two elements being
considered. It can be misleading when misused.
> 
> ...Yes, I'm too cheap to pay for EZNEC. 
> 
> >The 3/8" elements do a
> >nice job of increasing bandwidth and allow getting all of the elements
> >for a 2m beam from 6' lengths of tubing without waste.
> 
> I used 8' tubing and have plenty of scrap pieces left over for a UHF beam!
> 
> >I used NBS data
> >to build a cross polarized yagi for my dad to use with Oscar satellites.
> >It worked well. I don't remember the dimensions.
> 
> I *wanted* to use my H/V cross yagi for satellite / terrestrial SSB/ FM.
> But when I looked at how many relays I would need to select phasing and 
> elements so I could use one feed for LHCP, RHCP, HP, and VP, I decided
> it would be cheaper to run to lengths of RG-8X down the pole and forget 
> about LHCP & RHCP. 3 coax relays would cost more than rest of the antenna
> and feed. And then there's the matter of an elevation rotator....
> 
> >
> >How do your element lengths compare to the NBS yagis in 80s vintage ARRL
> >publications?
> 
> Didn't look.
> 
> >Those yagis are within a half dB or a dB of the optimized
> >latest designs. If you get back to the NBS report (977 if I recall
> >correctly) you can see the adjustments needed for boom and end effects.
> >Many have built using the NBS designs and unless the builder ignored the
> >changes need for different sizes of booms and elements the gain came out
> >as claimed. At the same time when the element lengths for an insulated
> >boom were used with a metal boom the gain was way down.
> 
> I have a metal boom, but the elements are insulated from it. 

The only way a metal boom doesn't affect element length is when the
element is spaced a few boom diameters from the boom. A boom along side
the element but insulated from the boom requires an experimental
increase in element length on the order of 1/3 boom diameters. Going
through and insulated needs a correction close to 2/3 boom diameter not
far from that with the boom connected to the element going through it.
So the effect of the boom is to effectively shorten the elements and
that makes the peak gain of the yagi lower in frequency and the gain
falls rapidly on the high side of the peak gain. I saw the combined
effects of 1/4 vs 3/16" elements and insulated length vs a 3/4" square
boom take a 12 db NBS yagi down to 7 db at the 1984 CSVHF conference.

There is a reason why all serious VHF and UHF antennas use a T-match.
That is that the gamma has always resulted in a significant loss of gain.
It may to do with the skewed pattern of the gamma matched dipole couples 
poorly to the director array. That's not in the modeling program.

> 
> >
> >As others have said, 125 miles should be a reliable path for SSB. Long
> >ago, as Iowa Army MARS technical coordinator I tried to establish a 2m
> >SSB net to replace noisy and propagation clobbered 75 meter nets.
> >Unfortunately I never could convince any MARS members that weren't
> >already on 2m to get there. But at 100 miles the signals were generally
> >better than on 75 (at the minimum of a sunspot cycle) during the evening
> >hours. There were deep fades, probably a couple an hour but the rest of
> >the time signals were solid.
> >
> 
> Maybe that those trees are causing more harm than I expected, but there is
> nothing I can do about it. Also to consider, is the all of those lovely
> aluminum clad campers near by must raise the apparent ground level 
> somewhat. 
> 
> >During the UHF contest, I worked 200 mile paths on 222 and 432 cross
> >polarized from a discone here. And there were many complaints in the
> >soap box about the poor conditions that weekend.
> >
> >Bring your yagi to the next CSVHF or Aurora conference and see if the
> >computed gain matches the real world gain at the antenna measuring. Many
> >home built and commercial antennas have missed the claimed gains by
> >large amounts. Yet some do well. See how your's does. I know its been a
> >sobering experience to see how far theory is from real world works
> >sometimes. Yet some antenna makers haven't bothered to change either
> >their designs or their inflated gain claims despite low results at CSVHF
> >conferences.
> 
> I'll see if I can bring it to Aurora this year. I'm taking it down for the winter
> anyway.
> 
> Dan KC0LUY
> 

CSVHF conference will be around MSP next summer.

-- 
73, Jerry, K0CQ
Dr. Gerald N. Johnson, electrical engineer
All content copyright, Dr. Gerald N. Johnson