Welcome to SIGNALS - a site about electronics and model railroads

Signals at East Tolland, CO. Photo courtesy of Kevin Morgan

May 24, 2010: APB/CTC unified intermediate signal
Today I finally had the time to trace down an annoying bug on the APB/CTC unified intermediate signals. When in CTC mode and with fleeting enabled, the signal correctly cleared after a train had passed, by showing a Restricting aspect (flashing red). When fleeting was not enabled and the clearing command was issued manually, the signal did not clear, even though the situation was exactly the same. The error has now been ironed out, along with some other betterments, thus version 1.4a (from 1.4) has been posted on the module's page.

April 28: CTC panels
Maybe someone already noticed that I've updated the CTC panel page. The concept is now defined, schematic confirmed, and the PCB layout underway. Software can only be tested when the panels are fully assembled, and that should happen before end of May. So stay tuned!

March 24, 2010: Generic logic, specific configuration
I'm thinking about an interlocking safety logic that runs exclusively on configuration read at runtime. Ideally, it should not be necessary to modify the source code of an interlocking node, but only to load a generic firmware. The whole configuration, from physical connection of the hardware to the safety locks of the interlocking, should be developed with personal computer GUI tools, and loaded into the controller. While I'm pondering this around, any hint, help or idea is appreciated (e-mail). A page for this project has been created here.

March 16, 2010: Turnout controller goes I2C/TWI - POSTPONED
OK, this project has been postponed. For the signal installations mentioned in the February 25 update a new PCB concept has to be worked out. That's top priority now.
I'm currently working on a new version of the turnout controller, that is interfaced via the serial I2C (or TWI) bus. The advantage over commercial controllers is that debouncing of input pins, motor control, is already done in this "field device controller".

February 25, 2010: Movin', movin', movin'
Seems like some people in FREMO got hooked by signals! :-) Let's see if we can come up with a gauntlet, completely signaled by an automatic interlocking plant (like Gary, Indiana)!
Additionally, we will address the issue of CTC machine-style modular control panels, so stay tuned.

Frebruary 13, 2010: APB/CTC unified intermediate
Update: Two support tools have been posted: APB Configuration and APB Emulator tool. Please see the APB/CTC unified intermediate page.
The APB intermediate module has been present since the beginning of this site. A bit later came the CTC hold signal module, which allowed dispatchers to hold trains back from normal CTC points, being controllable block signals.
At the last FREMO meeting in Lauffen we found out that it would be nice if the behavior of block signals could easily be changed from straight APB to controllable Hold signal. Of course, having to reprogram the AVR chip is not as user-friendly as wanted; therefore, I can now present you the APB/CTC unified intermediate, incorporating both functions and switchable by writing EEPROM configuration values.
This module runs on the same controller as the standard APB signal and does not necessitate circuit modifications.
In the "Modules" sections of the Signaling compartment, the new APB/CTC unified intermediate has been posted on top, as it replaces the other two (APB and CTC hold signals). Those two are going to be kept for a month, and will be removed from the site afterwards. Have fun!

November 9, 2009: Cleanup and update

Update: The CTC Command Structure document on the CTC Control Point page was completed with the information about Safety Functions (link).

I just did a bit of cleanup on the website. Old and outdated software modules were deleted, the respective pages removed, and the left side menu simplified.
On the other hand, the following items got updated:

CTC control point software
APB intermediate signal

The software modifications warranting the update are as follows:

rework of the safety logic in the CTC control points;
introduction of the safety functions mentioned in the April 10 update.

The safety logic had some conceptual flaws where routes were not correctly identified when certain turnout positions occurred, causing inappropriate locking or removal of safety locks.
With the modification implemented, each signal has a defined number of routes, each being configurable as before (turnout positions, ... ) in the SafetyTables.h file. By remembering which route was set for a certain signal, it is finally possible to detect inappropriate turnout position changes, as well as easily impeding other turnouts' movement.
If you want to adapt a preceding CP installation, copy the CTCCmdProcesses, SafetyProcesses files from the archive (on the CTC CP site here). Then adapt your SafetyTables to the new structure as shown in the archive template. Define NUMBER_OF_SIGNALS and MAX_NUMBER_OF_ROUTES in your main file, as well as an unsigned char route_currently_set[NUMBER_OF_SIGNALS] array. All NUMBER_OF_EXITS get replaced with NUMBER_OF_SIGNALS.

The safety functions introduced allow to:

clear a current-sensor occupancy;
actively occupy a track circuit.

The packet architecture is described in the header of the main files.

The APB intermediate was integrated with those functions, necessitating the update. Note on the intermediate: the version with integrated CTC hold signal functionality is in the works; when finished and tested, it will replace both the current APB intermediate as well as the current Hold signal module.

November 3, 2009: First real operating session
From October 23 to 27, we were in Lauffen am Neckar in Southern Germany, attending a FREMO americaN meeting. The dates gave us nearly three full days of operations, and they were thoroughly used. This meeting, its layout and operation scheme were very special, as it was the first operating session where the whole CTC/APB combination could be really tested. My lap siding Hazeltine and Oliver Hampel's Fremont Branch junction (both equipped with CTC signals) delimited a stretch of line with one APB intermediate and one spur track, creating the first more-than-just-an-island CTC application in the americaN group. Additionally, this was the first occasion where all the specific features of APB (tumbledown, ...) and the APB interface could be seen in action, and more important, the dynamic behavior be monitored.
The west end of Hazeltine was next to the small classification yard, and Fremont Branch was the gateway to a quite trafficked branch line, creating a funnel with a need for capacity. Needless to say, CTC was able to provide it. Summa summarum, we were very delighted that the set-up worked flawlessly during all five operating sessions, and it was real fun to operate.
No first field test without betterments, though, so the following inputs from group members are in the works:

combine APB intermediates and Hold signals, so the intermediates can be controlled by the dispatcher if necessary, very easily switchable by writing a EEPROM configuration variable;
when a track circuit is occupied, the direction arrows disappear; it would be nice if the arrows remained active, in order to help the dispatcher keep trace of the individual trains;
logic update: the current logic of the CTC control points has some flaws not encountered before, and needs a revision; it's already done, actually, but needs testing before being posted.
Cheers!

October 11, 2009: Circuitron Tortoise turnout controller
A new item is present in the Signaling section: a turnout controller for Circuitron Tortoise slow-motion switch machines. The machine is directly run from the controller's output pins at 5 V, and position feedback is debounced, delayed and forwarded to the CTC equipment. Versions for two or one turnout are available. Go to the TOC page.

October 5, 2009: TDisp program suite version 1.2 released
The update of the TDisp dispatcher panel programs mentioned in the last post here is complete and available on this site. Please see the Signaling TDisp page for more information and downloads. Sources will come, too.
Update: Fixed an error in the TDisp-Editor, and posted version 1.2a.

September 27, 2009: PC-side program for USB interface
The USB interface for MRBus MRBus-CI2 does not send human-readable hex code to the PC, but raw MRBus packets encapsulated by the Serial Line Internet Protocol (SLIP). Therefore, a new PC interface program has been developed. Additionally, the TCP/IP side has been revised and now transmits complete MRBus packets, instead of just user-relevant data (source, destination, type, data). The program is available on the Tdisp-download and MRBus-CI2 site.
Update Oct. 3: Revision of TDisp software announced.

September 18, 2009: USB interface with DIL chip
The USB interface for MRBus presented some time ago had the flaw of using an ATmega128, available only in a TQFP package. Thankfully, the ATmega162 has also 2 USARTs and is available in a handy 40-pin DIL plastic package, so for the average do-it-yourself guy, myself included, building such an interface device is a lot easier.
You can find the new schematics and source code zip file in the MRBus section.

September 4, 2009: Info only
Well, it's September already, and early fall has come to our beautiful land, early grapes and apples are ripe, and temperatures have cooled down a bit.
We're preparing for the upcoming FREMO meeting (October 23-27) in Southern Germany, and it looks like this is going to be the first layout, where multiple CTC-capable modules are going to work together. Things like the new dispatcher panels have been postponed, but the essentials of the concept have all been implemented into the software. So, we're looking forward to this exciting meeting! Cheers

April 20, 2009: Special functions for FREMO modules
As mentioned elsewhere, I've developed this signaling application for modular arrangements, which are set up by the americaN group of the Friendscircle of European Model Railroaders FREMO. americaN features single track modules in N-scale, and is operation-oriented. The traffic control system preferred by most members is dark territory and TWC; CTC and signaling came along with my entry into this group. In order to create acceptance for CTC and the signaling modules, it had been crucial to design a system which does not require any changes/additions in existing as well as in future modules. Moreover, cars are not and will not be equipped with resistor-axles, making their detection with current sensors impossible. This last issue has been partially overcome by using infrared sensors at every block boundary, but other things impede correct functioning of the signal system.

For instance, if a spur track coming off the main track is not electrically isolated from the latter and not fed from the APB feeder, a locomotive on it will draw current through one of the adjacent current sensors, keeping the block occupied and, more problematic, the direction of travel set, even if the train is clear of the main track. One possible remedy would be to mandate isolation of those spurs, but this understandably is not acceptable to most of the group members that aren't interested in signaled operation in any way.

The remedy in development realizes two software functions:

active clearing of occupancy by ignoring current sensor detection
active occupancy by forcing occupancy even if no train is detected by current or infrared sensors

Stay tuned for updates.

March 21, 2009: MRBus-Computer Interface 2.0 (USB)
The USB interface for MRBus presented some months ago has finally reached a stable version. It uses an ATmega128 controller, and an FTDI serial converter cable. You can find the project in the MRBus section.

February 28, 2009: Going nowhere slow
As university and work (hey, it's cool to work with the real thing!) keep me quite busy, there hasn't been any progress on signaling projects. I even had to withdraw my participation with Hazeltine at a FREMO meeting, which I'm going to take to only the HBD module.

November 22, 2008: Functional Keyboard
Small bugfix to the Functional Keyboard source code, which becomes 1.0a. After an error, the selected CP is now retained instead of defaulting to CP 0.

November 14, 2008: Hot Box Detector update
The programs used to configure the Hot Box Detector have been updated! The previously posted versions were incompatible with firmware 0.9, a fact I hadn't realized. I'm ashamed.

November 13, 2008: CTC Functional Keyboard ready; visitors
The Functional Keyboard, which can be used as means to send commands to CTC field circuitry is ready, hence schematics and source code have been posted on the respective page. I like this thing very much, and can safe me a lot of buttons and switches :-)
The counter on this site tells me that the number of visitors has reached 900. Thank you all for your interest! Seeing the number constantly climb is a great motivation to continue!

October 8, 2008: Major update on the signal system
Finally, the promised update happened. The APB intermediate module has become version 1.2, in line with APB interface 1.2. The CTC control point module presents two different version, one is what I'd call a template, whereas the second is a version I have running. Take a look at the Signaling section!
Update: The Electric Lock switch has also reached postable version 0.9a. Please see the Electronics section for more information.
Update 2: A new module has been presented in the Signaling departement.

September 30, 2008: Signal system and new project
The time working on the signal system at the beginning of August had been well spent, and all of the modifications are actually up and running. At the moment I'm busy with work, and the winter term at university is about to start, so I am short on time for my hobbies. I hope you will understand, and promise to allot a day to updating the website.
There is also a new project in the works: an electric lock switch. You can find it in the Electronics section.

August 4th, 2008: Signal system and CTC panel application
After some problems yesterday, I'm making progress on the signal system. I'm in the middle of making a switch to a new version of the APB block interface, as well as the indication management in the CTC Panel application (necessitated by the requested modularity), and this reflects in the state of this website: some information is old, and some new. I hope to get that fixed in a couple of days.

August 3rd, 2008: HBD source code
I finally had gotten myself to work on the signaling system of Hazeltine, when some parts of the equipment failed, especially the monitoring devices (MRBus computer interface, power source, ...). Damn! Thus I decided to at least post the current version of the Hot Box Detector source code, without any modifications (unlike I had promised). You can find the zip file in the 'Technical' section of the HBD page here.

June 26th, 2008: Final exams
June is nearly at its end, and the last exams in this term are coming up. Monday will see three of them in a row, and the last will be taken on Friday, 4th of July. Keep your fingers crossed for me! I promise I'll update the site afterwards :-)
One other thing: An article about our FREMO group and remote dispatching can be found in the July/August issue of the NScale Railroading Magazine, page 27.

May 24th, 2008: Update of MRBus files
Following a new release of WinAVR and the avr-gcc compiler, the interrupt routines of MRBus for AVR have been adjusted, so version 1.1b is now available. Please take a look at the Download section in the Signaling area.

May 19th, 2008: About the source code of the HBD device
It's been over a month now since the last update of this site. This time has been well spent on a major rework of the signal system installation in Hazeltine CTC siding. The existing circuitry was a "workbench", and would have needed substantial improvements. I chose to start over again, using the same MRBus system and the same logic, of course.
Therefore, any work on the HBD has been dormant. Particularly, before I'll post the source code of the device, I will have to rework it to be of any use (at the moment it is short of being a huge mess). I apologize for the long wait.

Apr 7th, 2008: Report from the first field test
As promised, I've put together a short report about the �bach-Palenberg meeting. It revolves mostly around the Hot Box Detector, and therefore is located on the device's Installations page.

Apr 2nd, 2008: First field test of HBD - success
From March 27 to 31 I was in �bach-Palenberg (Germany) and attended a FREMO meeting with my modules Hazeltine and the prototype of the Hot Box Detector. I'm very happy that I can say it was a big success! Photos coming ...

Mar 1st, 2008: Want a Hot Box Detector?
An important decision has been made: if enough people show interest, we will produce a series of devices with PCB and front plate, assembled, programmed and tested. Please see the Detector's page for more information.

Feb 26th, 2008: Revision 0.9d for the HBD schematic
An error in the HBD's schematic has been found and eliminated: the relais activating the radio must be inserted between the 10k potentiometer and the Radio Microphone input.
Additionally, I'm getting ready to install the first prototype by end of the week. I plan to add train length calculation in scale feet. This feature, as well as speed and axle count, can be individually turned on and off with the HBD program tools. A future version will see use of a 3x4 matrix keyboard.
Update: There is also a small bugfix to the APB intermediate signal controller software available. You can find it here on the module's page.

Feb 4th, 2008: Complete: new variant of the CTC node software
The logic upgrade for the CTC node, which allows easier handling of turnouts and routes, is fully functional and tested. See Electronics -> Signaling -> Modules -> CTC node.

Jan 26th, 2008: CTC node
The source code of the CTC node has been uploaded, along with a PDF showing programmable inputs and packet summary. See Electronics -> Signaling -> Modules.

Jan 24th, 2008: A lot of updates
There is quite a bit of new stuff about both the signal system and the respective CTC panel application. See Electronics for the first, and Software for the second.

Jan 21st, 2008: New schematic for the HBD
See the HBD's Technical page for a new version of the schematic. This should be fairly close to production units.

Jan 14th, 2008: First update - Erstes Update
There is some new information available about the Hot Box Detector in the Electronics section.
Es gibt nun Informationen �ber den Hot Box Detector in der Sektion "Elektronik".

Thanks to us-modellbahn.net for providing the web space for this site.