As I posted a few weeks ago, I'm setting up a speed matching loop.? Using 8 BD-20 NCE detectors, 1 AIU, PH-Pro, RS-232 to computer running JMRI (Windows 11 OS).
?
Using the Ultimate Speed Matching (USM) scripts as a basis, modifications seem to be working well with the following exceptions:
?
Speed measurements can be all over the place due to "bounce" on the detectors.? The JMRI Speedometer Tool seems to not see this as much (doing exit to exit on the tool).? So will adding de-bounce to the 8 sensors help?? Do I use active bounce, inactive bounce, both?? Does this change the factors for calculating speeds since in the script, it's looking for sensors changed?
?
Is there a way for the USM script to open up the speedometer windows and provide an output back to the USM script.? In other words, is there a way in the USM script to call the speedometer and get the result back to use rather than the timers used in the USM script?
?
Another couple anomalies:
--? After the script is complete, I then run the loco at the speed steps and measure the expected speed and it seems to run a couple mph slower than during the script.
--? Has anyone figured out how to speed match with an ESU decoder where SS28 is always 255 and the top of the curve is adjusted with CV5 and the rest of CV67-94 are proportional?
--? Using the 3rd lowest measured speed seems to be a short cut.? Has anyone taken the measured speeds, throw out the top two and bottom two and return the average speed?
?
I have a current methodology to manually set the speed table by running the loco at SS 5, 10, 15, 20, 24, 28 and entering the measured speed into a spreadsheet that then tells me what setting to set each of those to in order to get the desired speeds (my standard is 10, 20, 30, 40, 55, 70 at those SS).? After I get more of the bugs worked out of the modified USM script, I'll take a turn at trying this method.
?
?
Phil Hartung?
Idaho Falls, ID
?
