Thursday, May 14, 2020

May 2020 Journal

Trackwork


5/14/2020 - I finally came up with a design for the industrial/switching district in section L5. After ordering a few more turnouts, I laid the track and I'm happy with the overall design.


This photo shows where the industries can fit into this track arrangement.


Based on a suggestion from my friend Brad, the siding at the top will actually enter the building that's going into that space, plus there will be room on the outside for a few more car spots. The bottom track towards the left is primarily space for pulling cars out of the siding to the right, but once I get some power on the track, we'll see how many cars I can reasonably pull at once without having to move the cars to the left.

5/18/2020 - After the glue dried, I finished the section by adding feeders into all the track segments and between the various turnouts. I hadn't soldered in a while and the first joints showed, but I got the hang of it after a few bad connections. Once all the wiring was done and tested, I grabbed some cars from the main yard and put them into place. It looks like I'll be able to fit 19-20 cars in the district without losing the maneuverability to hit all the tracks. Here are some photos of the finished area.




Wiring/DCC


5/19/2020 - After getting the industrial district done, I went back to the helix room to diagnose my detection issue. The helix will have a total of 12 detection blocks, 6 in each direction, starting at 300W/300E and going up the helix to 305W/305E. When I last worked on this, I was getting what seemed like false readings on 302 that wouldn't go away. I went down some rabbit holes about whether the coils were too close, the wires were crossed, etc. After disconnecting the bus wires from the terminal strip, I had an engine continue running even when its bus wire was disconnected. I then proceeded to remove the other side's bus wires one by one and found that I had mixed up one pair of feeders. After making that fix, the detection started working properly again. 

I managed to get both sides of blocks 301, 302, 303 wired up with detection coils and got them connected to the Watchman board. In addition, I happened upon the Indicator Track feature in Panel Pro and replaced the somewhat awkward icons/lights with indicator tracks. The panel looks so much better doing this. I also got a reminder of JMRI's web server feature, which is also accessible from the mobile application. This made it a lot easier to test the detection since I didn't have to keep running out to the computer.

Unfortunately, I ran out of red 20 gauge wire and I'm running low on 3M suitcase connectors, so I'll have to wait a few days to restock my supplies. That's the first spool of 500 feet that I've run through, not sure if I'll need another one after the second spool, but we'll see. I get all my wire from Del City since you can order a variety of colors individually. 

The next steps are:
  • Split block 301 into two blocks -- block 300 will cover the track from the entry to the helix room around to the back wall of the helix.
  • Add feeders to block 300W and 300E and connect up the detectors to the Watchman board, which will finish out the 8 connections on that board. 
  • Program another Watchman board to accommodate blocks 304-305W/E and future blocks that will go into the peninsula. This board will be installed in the upper deck section prior to entering the helix, since that puts it in easy reach of peninsula blocks when I get those built. 
5/25/2020 - I spent today working on wiring for the lower deck. First, I extended the RR-Cirkits Simple Serial Bus (SSB) from the boards under the main yard around the corner to the staging yards on the north wall. I then extended the Tortoise power bus, since the RR-Cirkits Motorman boards require secondary power from a power supply. I added 22 gauge wires to each of the six Tortoise drives at my work bench, since that's far simpler than trying to solder under the deck.

After that, I was able to get the lower deck's Motorman board programmed and throwing the turnout that was previously installed under the first turnout in the throat. I installed a few more Tortoise drives and got those connected to the Motorman board. I still haven't figured out a good way to install Tortoise drives with only one pair of hands, but being able to throw the turnouts via the Engine Driver app is far easier than running back and forth to the computer.

In JMRI, I added those turnouts to the panel and after swapping the two power wires on the Tortoise drives, got them in sync with the panel. Since I've got three of the 7 turnouts wired now, I could try out the Routes feature of JMRI. This lets me create a route, say IR105 for staging track 5, and trigger it from the control panel or an Engine Driver throttle. This is simpler than having to throw all the turnouts individually. I still need to figure out how to put a button of some sort on the panel to trigger the route, but the basic logic works nicely.

The next step here is to finish the remaining four turnouts and get them wired into the Motorman board. In JMRI, I'll add routes for all eight staging tracks. After that, I'll wire up the Tortoise drives on the upper deck, although I first need to rework the yard entry using the double crossover I found in Minnesota last year. 

5/27/2020 - I wired up two additional Tortoise drives in the lower staging yard this evening. I also created the other routes so that I have one route per track in the lower yard. For the upper yard, since there will be a double crossover at the entrance, I'll end up with two sets of routes: one set to reach all eight tracks from the inbound/westbound main and one set to reach the tracks from the outbound/eastbound main. 

5/28/2020 - I finished wiring the last Tortoise drives in the lower staging yard and tested out all the JMRI routes. It's nice to have that work done. 

5/30/2020 - I finished wiring up the remaining Tortoise drives at the "west" end of the main yard and got them all connected up to JMRI. I can now control the entire yard from the dispatcher panel. I am planning to do a "proof-of-concept" with the push buttons to control the crossovers locally, as well. The idea is that the yardmasters will be able to use these controls as needed, but they're also available for the dispatcher to use. 

The lower deck JMRI control panel now looks like this, for now:


5/31/2020 - I soldered feeder wires onto six detection coils to be installed in the helix. I also "discovered" that I had a staple gun in my toolbox and replaced all the unreliable duct tape with staples to secure all the feeder wires. This cleans up the wiring quite a bit and lets the detection coils stay securely in place. 

Operations


5/26/2020 - I've been working on software that ties into my train management/inventory system that was showing promise in terms of random train generation, but after thinking how much computer interaction there would be, the ongoing tracking of car numbers, etc. I decided to give that up and switch to the car order system that Dan Hinel, Hank Tenwolde, Mike Wolf, and others are using. In this system, you deal with the type of car and focus on the car spots on industry tracks. The system is self-repairing and simpler since you are only looking at car types, not car numbers. While I'll still have a database of all my cars and locomotives, the computer won't be needed for car/train management during a session. 

I went through my three industry districts and identified 55 car spots and chose various types of cars to fill those spots. After the car is picked up at the location after loading/unloading, those cars will be sent out on either eastbound or westbound manifest freight trains. This simulates the load leaving the layout going to some other location and helps rotate car usage. 

I then ran trains to all three districts to adjust the cars on each of the industry spots and brought the remaining cars back to the main yard. The next step is to rotate the cards -- setouts become pickups, holds remain holds for one "turn" -- and then run local trains to pick up the cars that need to be moved. After that, I'll run manifest trains in either direction to pick those cars up and then I'll have a corrected car cycle to continue with. It'll be more interesting as I build out the peninsula and the additional twelve switching districts I'm anticipating, but this system lets me start operating on my own without having to constantly refer back to the computer. It's ironic that as a software developer that I'm choosing a manual system over a computerized one, but I'm looking more at the ongoing maintenance and choosing a simpler system.