Based on predefined first and last speed steps, which are picked
through trial and error, the spreadsheet calcuates CV values for the
26 remaininig speed steps. These values are based on a logrithmic
curve that, when set-up in the speed table, result in a gradual
increase in speed from low to high end.
To obtain a copy of the file, see Files section on Yahoo's Digitrax
NG -
Bob
--- In jmriusers@y..., "Alex Shepherd" <ashepherd@w...> wrote:
Perhaps I'm the only one that does this, but I use a spread sheet
to
calculate my speed steps then input those calculated values into
each of the twenty-eight speed table CV's. For example; the CV's
of
one of the Lenz Decoders where calculated as 22, 24, 26, 28, 30,
33,
36, 38, 42, 45, 49, 53, 57, 62, 67, 73, 79, 85, 93, 100, 109, 118,
127, 138, 150, 162, 175 and 190. Setting these values was
difficult
as I found it necessary to toggle back and forth between the Speed
Table and CV window in order to set each value. I tried inputting
the data directly into the CV table (CV 67 to 94) but found the
desired values where not transmitted to the decoder and
subsequently
all speed step values were set to that of CV 67, in this case 22.
What does the spreadsheet do? I wonder if we could copy the values
to the
Windows Clipboard in such a way that you could do a Paste into
Excel and
vice-versa.
Some time ago, I read about John Kabat's Racetrack program:
I looks to me (having
never run
it) that it lets you program the speed tables in multiple locos so
that they
match up and can consist together without problems. I also read
about others
calibrating their locos so the 0-100% number the DT3/400 throttle
indicates
scale MPH.
It seems to me that if you had a loop (circle) of track with some
sensors
(at least one) equally spaced around it for detecting the loco's
position
and you know the scale distance between the sensors and can fairly
accurately time how long it takes the loco to travel between
sensors, then
you should be able to operate the loco over the 128 speed steps and
then
work out what to set the speed table to, so that it indicates scale
MPH or
matches some other standard speed profile you use.
Maybe the sensors are an IR LED and IR detector connected to the
Serial/Parallel port on the PC. The trick/challenge would be to get
Java to
be able to receive the location pule in a timely fashion but it
shouldn't be
too difficult to get some sort of cheap position feedback back into
JMRI for
this be be possible. Depending on how big the loop is you may be
able to use
LocoIO and listen for a particilar sensor input message if the
message
latency does not add significant error.
Does this sound feasible/useful?
Cheers
Alex
