Hi David

Okay so you can afford to use a relatively high gain system. You also
need to have the antennas high and clear as you can assume that 900MHz
is line of sight. You could model the exact environment if you have non
LOS paths. That will indicate whether it will work or not. Changes in
antenna gain for an omni probably wont help much if the link is marginal
when modelled. You could also establish whether you could (say) do the
whole system with simple 1/4 wave whips! (A 6dB increase in power or
combined antenna gain over an LOS path gives you twice the range)

If you are scaling an antenna yes you need to scale everything. I didnt
suggest you scale the Cebik antenna, just wanted to show you the
possibilities. It might however be more use to build an adjustable one
and have some method of recording/measuring the field strength and
changing dimensions for a maximum. I wouldnt call the super J pole an
especially critical design for dimensions though. You could for example
use 1/8" brazing rod for elements and make the length marginally shorter
by the amount indicated in end effect tables. The more elements you have
the more critical the dimensions and adjustment will become.

I would say it isnt normal to build at 150 for 900MHz. Too many
variables to account for.

What is on 900MHZ? Is this a data network like 2.4GHz 802.11?? Will
multipathing be an issue? (High data rates usually need directive
antennas not only for a satisfactory S/N with gain, but also to reject
any possble multipathing problem of a data bit arriving later than it
should.)

Cheers Bob

David wrote:
If I were to scale up the dimensions of the Super J-Pole shown on the
site, would I need to scale the element diameter too ? This would make
the element diameter less than 1mm at 915 MHz ?

Secondly, is it normal to make the 150MHz version and test it then scale
it up based on optimized dimensions from the lower frequency test ?