Blog

In the layout room, nobody can hear you scream

June 2017

Never a problem on the Mighty MKT

Dont run trains like slot cars2

I sure hope the dispatcher doesn't have OCD.

IMG_4345

IMG_3489

Recap - June 10 Operation Session

19144041_10211920257655410_2041797274860141626_o
The Springfield/Bella Vista group, hard at work on the Mighty MKT. East Rhineland is in the foreground.

We approach perfection asymptotically.

I’ve been in a operation state of mind lately. I recently hosted the nice folks from Springfield and Bella Vista (Missouri and Arkansas). After all the work it takes to prepare for an operations session (including the track and wheel cleaning and the preparation of session snacks), I decided, upon the generally successful completion of the session, to call another in two weeks. The train room was clean, the crew lounge was clean, the track was clean – so I wanted to strike again while the iron was hot (soldering iron?).

A crazy thing happened prior to the first session. I was at peace. Sure, I had a several items on my to-do list, but I diligently worked on them and crossed them off one by one. I was ready. The layout was ready.

The session had its usual number of glitches. The DCC short beepers went off way too often for my satisfaction (I use the CVP Products ZoneShare products, and have the layout divided into eight zones, each using a separate tone to indicate a shorting condition). Most of those shorts were caused by train engineers running into turnouts thrown against them. Operators paying more attention will cure that.

A couple of signal aspects were found to be backwards (red for green, and vice versa), and those were easily corrected in the hour after the session, thanks to my detailed wiring and logic notebooks. It was a quick fix in JMRI to correct swap red and green aspects. I will watch the signals closely next session to make sure that they are correct.

I had a couple of logic glitches on the dispatcher panel, the most obvious being that the St. Charles autoramp tunout was being thrown when the East Boonville turnout was set for the siding!! This was an obvious programming error on my part, since they are in no way connected. After the session, it was quick work to pull up those turnout configuration and, sure enough, I had somehow inadvertently told JMRI to throw the autoramp turnout along with the siding turnout. Go figure! It was a quick fix.

The most frustrating aspect of the session, was the period “glitches” experienced by the dispatching panel that would blink the display, then lose all block occupancy for an instant, then correct. However, it was just enough to allow the block tracking variable to lose their train identification. The behavior was observed at a previous session, and I had planned for it by building a separate screen/panel, that could quickly restore the trains whenever the dispatching panel “glitched”.

The glitching behavior occurred several times during the session, and I determined to locate the cause and fix it before the next session. I recalled that it had not originally glitched and that these errors were a recent phenomenon. My Consultant for all things electrical (Joe from Kasper Electronics) suggested that it may be either a bad USB/485 dongle or bad RS422 chips on the SUSIC. I swapped out the SUSIC with a spare I have, and the glitching behavior went away! I now have several replacement RS422 chips on order from Digikey.

Another item that had bothered me from the session was that we couldn’t see the status of the east St. Charles turnout on the screen. The turnout is wired to that the block show occupied. Several times during the session my Ace Dispatcher (Joe) commented “Why is that block still showing occupied? The train left long ago!” It was a thrown turnout at east St. Charles. We couldn’t see that the turnout was thrown on the screen. So I wired an input line to the spare set of contacts on the torti (the other set of contacts being used to power the frogs) and, with a little bit of coding, we now can see the status of the turnout on the dispatcher’s screen. It’s the CMRI difference, baby!

The point is that, little by little, I am whittling down the irksome and bothersome items on the layout. And as the irksome items become less and less, my anxiety level also drops.

Who knows, maybe someday I will be able to actually enjoy a session on my layout!

IMG_4341
That's what I'd call a full torti. The outer two wires (pins 1 and 8) drive the throwbar. Pins 2, 3, and 4 power-route the DCC signal to the turnout frog. Pins 5 and 6 are fed back to the CMRI to toggle an input line that is read by JMRI to animate the East St. Charles Industries turnout on the dispatcher's screen.


IMG_4343
Poor quality image of the new animation of the East St. Charles Industries turnout

Anybody know anything about these?

MKT Trucks

email me if you can provide some background info.