The following message on Bruce Winter's misterhouse mailing list (http://misterhouse.sourceforge.net) from:
Clay Jackson [email@example.com]
OK - here's how I solved the RF range issue (lack of range) with my MR26A.
My MR26 is located in a room with 5 computers and two ham radios (VHF and
HF) on the ground floor of our tri-level. Before any 'enhancements', the
range of a PalmPad controller to the MR26 was about 10' (through one
interior wall, although that didn't seem to make much difference). These
tests were conducted over the course of a few days, in a pretty 'informal'
fashion (I didn't have a tape measure, or any sort of RF 'range' set up, I
was just running around the house pushing buttons on the PalmPad, making
note of where I was and which buttons I pushed, and then reading the debug
logs from the computer running MH). Since my receiver is in such an 'RF
rich' environment, I suspect anyone else's mileage will vary quite a bit,
but perhaps this will help with a baseline. I did perform all the test with
fresh alkaline batteries in both the PalmPad and the various sensors.
First, I built the antenna described at
http://www.laser.com/dhouston/turnstil.htm (thanks Dave!).
Tests with the antenna wired to the end of the 'tail' coming out of the MR26
increased my reliable range to about 15'. I also discovered that the
antenna performed best when vertically oriented (by almost double the
range), so all subsequent tests were conducted at
Next, I got an 'F' type (cable TV) through-plate connector and mounted it on
top of the MR26 (I'll try to get some pictures if someone's really
interested). I soldered the ground on the connector to the large Using
25' RG58 (50 Ohm cable, NOT 75 Ohm RG56) between the antenna and MR26 the
increased the reliable range to about 30'.
Then, I had a Radio Shack 'InLine Cable Amplifier' (15-1170) in my junk box,
so I put that in the line. This amp delivers 10Db gain from 50-1170 MHz
(the X10 signals are around 319 MHz). I only got about 10' more range with
this setup. Since I have one HawkEye that's almost 100' away, that still
wasn't good enough (be careful to get the one that's 50-1170 MHz, and NOT
the 20db one that's only good from 450-2000 MHz).
So, I went out and bought a separately powered amp (again from Radio Shack,
I don't have the part number handy) that delivers VARIABLE gain (there's a
pot on the side), up to 20Db. With that installed, and the gain control
cranked all the way up, I get about 150' reliable range, BUT, I also see a
bunch of 'garbage' (basically bad values, an occasional stray signal that
still gets decoded correctly, but doesn't match one of my remotes). If I
back the gain down to about 3/4 (assuming it's linear, that would probably
be somewhere between 12 and 15Db gain), the garbage goes away, and I get
reliable reception of the PalmPad, a couple of the SlimeFire switches, and
several HawkEye and EagleEye motion detectors, some of which are as much as
100' away from the receiver.
One of the reasons I was so interested in getting this working was that I
was expecting to see a much more rapid response to the RF 'event' (since the
path is now Sensor->MR26->MH->CM11->X10 Device, as opposed to Sensor->RF
Receiver->CM11->MH->CM11->X10 Device). In fact, I did get much faster
response (I haven't measured, but the delay between sensor activation and
X10 response is almost imperceptible now).
Hope this helps someone else....