On Fri, 09 May 2014 12:02:50 +0100, Mike Perkins
wrote:

On 09/05/2014 04:24, Jeff Liebermann wrote:
On Thu, 8 May 2014 21:31:45 -0400, "Tom Miller"
wrote:

At 200 m (down the street, why he is doing this) the FSPL is 71 dB. The Tx
antenna will have lots of loss buried in the car body. The Genie opener just
has a short (15 inches) length of wire hanging down for its antenna. No real
400 MHz ground plane so I doubt it will be 0 dB.

Agreed. 0 dB losses and 0 dBm gains were mostly so I didn't have to
reach across the desk for my calculator. I made no effort to be
accurate or even realistic. I simply wanted to demonstrate the effect
of putting a passive repeater in the middle of the RF path. I can
change the numbers to more closely reflect reality, but the added
losses due to the passive repeater will not change things very much.

A 200 meter outside path will make a passive repeater even less
desirable. If I leave the inside path at 10 meters, which I think is
realistic for across a large garage, and use 200 meters for the
distance between the car and the garage the path loss will be:

Without passive repeater. 210 meter path at 433 MHz:
http://www.qsl.net/pa2ohh/jsffield.htm
the loss is 72 dB.

With 200 meters plus the passive repeater at 10 meters, the loss is:
71.2 dB + 45.2 = 116 dB
With this arrangement, the RF signal level at the receiver will be
-126 dBm, which is well below the receiver operating level.

That's a difference of 44 between using a passive repeater and a
direct shot. Throwing in realistic antenna gains will make things
even worse.


Quiz: Starting with a 1/4 wave whip antenna over a large ground, how
short can this antenna be cut before the gain drops substantially?
http://802.11junk.com/jeffl/antennas/Monopole/index.html
The numbers in the filenames are the antenna lengths in wavelengths.
For example, monopole_0_125 is 0.125 wavelengths or 1/8th wavelength.
I only ran 3 applicable lengths, but the gain figures should offer a
clue that as long as the feed impedances are reasonably well matched,
you can probably cut an antenna down to below 1/20th wavelength, and
still have a fair amount of gain:
wavelengths gain(dbm)
0.250 5.19
0.125 4.85
0.050 4.75
The VSWR is atrocious, but the gain is still mostly there.

What is the software model you've used here?

When the physical dimensions of an antenna are reduced, the antenna
effective capture area and hence the obtainable receiver power drops
much slower. Of course, there must be a proper match between the
(reactive) antenna and the receiver.