Up the Creek Column - 13
The i/o pin was being switched to input for about 9 uS.
There was a 4 uS + charge curve for the lines without their
push button pressed.
There was a small amount of electrical garbage on the line.
I sent 'scope photos to Duncan. He speculated the Dual 3-Way con-
troller was sampling the signal too soon after switching to input
mode, so he made a fewmicro-code tweaks, and mailed me a pair
of replacements.
I'm pleased to say the new controllers are working fine! I success-
fully tried themwith the same 15' of twisted pair wire that hadn't
worked originally. Then I torture tested them using 300' of twisted
pair wire. The LED's in the control panel didn't light – a matter of
too much resistance in the cable – but the controllers correctly
interpreted the push button commands and switched to red, yel-
low, or green aspect as desired! Go team!
Conclusions
It's neat to watch the semaphore blades go up and down. The
Dual 3-Way controller even has a bouncing blade effect.
If you're on a budget you might want to think twice about
automated semaphores – a Tomar train order semaphore, Dual
3-Way servo controller, pair of R/C aircraft servos, and control
panel wiring will run over $100 per train order installation.
A simpler alternative used by Tony Koester is moving the sema-
phore blades by hand. You could also install fixed semaphores
and use LEDs in the fascia to show aspect.
I was in a hurry when I installed the train order semaphore at
Oakhill and didn't enjoy the debugging. Many of the problems
were my fault – double checking parts before installation saves
time in the long run! Once I got Duncan the 'scope photos he
was quick to apply a fix and
send me replacement boards
but having him away on
vacation (and away from his
e-mail) didn't help my anxiety
levels.
Allow lots of time when install-
ing new electromechanical
devices – don't be like me and
attempt it with an op session
imminent! I ended up so fraz-
zled that I cancelled the ses-
sion. Grrrr...
Now South Jackson needs its
train order semaphore!
22: The waveform on the yellow
wire. After switching to input mode
over, 4 uS was needed to charge
the wire's capacitance. Reading this
pin during those 4 uS can result in
reading a '0' instead of a '1'.
MRH-Oct 2013
