Are we having fun yet Tom?


Jul 07 08:38 04, Tom Walker wrote to Paul Westell:

 TW> If the "dish" isin't perfect you will loose a great deal
of Gain for the antenna. Althouhg for closer distances almost anything that
will reflect the 
 TW> signal in the direction wanted can work. I saw one project using a
Chineese Vegtable Spoon.

yes, but the type of injury may have left it in otherwise usable condition.
a sharp blow with some force may have done only localized damage, but if
the dish was dropped with only enough force to bend the edge then its now a
birdbath. drop a brick and a basketball on sufficiently thin ice, the brick
will go through leaving a brick shaped hole, while the basketball crazes
the surface without breaking through

 TW> Actauly it is a very precision thing. The formulas for the Wave
guide antenna take into account all of the dimensions of the Can. The
placement of the 
 TW> Connector with the stub antenna is alos very inportant. As is the
length of the Stub antenna. Whose length varies a little with the
dimensions of the can 
 TW> and the length should be correct down to 1/100th of an inch. As an
example the dinension for the Anntenna for the Nalleys can antenna was 1.21
Inches. 
 TW> One cuts a little long and fine tunes with a small file using
actual signal strength readings..

I am well aware of the precision required for anything of this nature, and
was not really complaining about the results i am getting, rather resenting
the apparent success of far less effort. i just don't have the facilities
to work any closer than say 1/4mm at best (and that may be pushing it) on
any one component, and you can't depend on errors to cancel

while studying the theoretical design, one chart detailed upper and lower
dimensions of waveguides for the 2.4 gHz frequency, diameters and lengths,
and compared these to a variety of cans in common use. all of the cans fell
outside reccomended dimensions, and i was left with the impression that
performance would deteriorate rapidly. 

the greatest fault of using cans seemed to me to be their length, they are
far too short by any standard i've seen so far. most designs recomend the
tube be 3/4 of the resonant wavelength of the tube, however the more
detailed analysis reccomended minimum lengths of 2 1/2 times the resonant
wavelength, and extrapolating from that, it would seem acceptable tube
lengths are incremented at 1/4 multiples of the resonance. i am at a loss
as to why the 3/4 length was ever recommended except to speculate that at
some point someone included the 1/4 wave reflected portion twice to
calculate the required length for the tube.

i was most surprised to realise the dimensions of the tube seem to have
little (obviously direct) relationship to the broadcast frequency, but its
capacity for wave propogation seems to be related to the volume of the
tube. i calculated the dimesions required to bring the length of the tube
into some sort of correspondence with the broadcast wavelength, and found
it fell just outside the recomended dimension. it never occurred to me to
calculate a correspondence for diameter.

 TW>  Can size is quite important and the Nalley's can "Size"
gives the highest gain according to the info I found on the Internet. Do a
Google search for 
 TW> "WiFi Wave guide Antenna", Some have played with a
Pringles can and while they work they have the Lowest gain because the size
is a little small. But 
 TW> when you are out War Chalking they don't attract as much attention :-)  :-)

As i recall that site, he was only comparing different cans, and while the
nalleys can may have performed better than the pringles, my research shows
it too falls outside desired limits for diameter, at the other end of the
scale. i have yet to see anyone actually construct to theoretical standards
and compare that to a can.

despite the undeniable entertainment such chalking would afford, i like it
much better where i am. the neighbours are very thin by city standards, its
5-10 degrees cooler here than town and i can hear the beach from where i
sit on all but the calmest days.


It could be worse ...
Paul

--- Msged/LNX 6.1.1