I'm laying track right now and want to be prepared for future block detection and signaling. So I want to place track feeders where they can easily be wired to a block detection system. I have some questions on where to detect. 1. My current thinking is to place three detection sections in each block. A 1" or so section of track at the begining and ending of the block, and the large section in between. like this: =|=!=======!=|= (where each block is double gapped at the ends, and each detection section is single gapped to facilitate direct home wiring.) My thinking is that if I have 3 detection sections per block, I can determine not only occupancy, but also direction of movement. Are 3 detections sections per block necessary, or even desirable, or can you just as easily detect direction of movement using a different method? If I use 3 detection sections, is a 1" section sufficient, or should it be larger, so it can be used as a braking district also? 2. What do I do when I come to a turnout? Do you detect occupancy within the plant? Or should I just be checking for switch position before running a train through a plant? An image of my layout is attached. Thanks, Greg
Maybe a little better schematic I'm getting views but no responses, so I made up a couple of schematics to maybe clarify what I'm asking. I have a solid understanding of where my blocks and interlockings belong, what I'm trying to figure out is how much detection I need for each block, and wether I need it for interlockings. A block schematic of my layout: Each block is in color and abuts an interlocking (in gray) or the end of the map. Signal locations are also shown. I've been wiring each block with 3 detection sections shown below: My thinking was that I could detect direction of travel with this method. Is this necessary to detect direction of travel? Is it undesirable to have multiple detection sections within the same block? If I detect multiple sections within the same block, should they be larger the 1" (what I'm preparing for now)? Is there a simpler method? Should I detect occupancy in my interlockings? If yes, how? Do I place a detection section on to the point end of a turnout, the legs, all three? or wire it up to detect entire interlocking as a whole? or maybe on the outer limts of the interlocking? I'd like to simultaneously run unopposed trains through an interlocking (example 1 east 1 west both straight track) The ultimate goal for all of this detection is to implement signaling and computer operations. Thanks for any input. Greg
What program will you be using to run your trains? I think some of them will know what direction the train is moving if you tell it what way it is going when you start off. Then as it moves around it will follow it. If you moving from one block to the next then you will know direction of travel. Rather then hook it up with more then one current detection, you could compliment it whith IR Sensors. This will allow you to know when a train gets to a particuler point or when the end of the train has gone past that point. Where you have your pink sections of track in the loop you could place the IRs there.
I think you're plans are a little out of the scope of most modellers, Greg. I'm having a time getting my head around what you are doing and I'm in the signalling game in 1:1! Let me think about it....
Greg, Great question. I will be watching with interest as I have been looking at almost the exact same scenario. I wasn't (neccessarily) using the end blocks for direction information, but rather to have the capability of letting the computer run a train (commuter trains to run a schedule, including station stops, while 1 or 2 operators run facing commuters and/or freight operations) and be able to stop it in a siding/station automatically. I'm also contemplating wether a turnout should be it's own section/block. Direction info is a good side benefit though. So in my scenario a passing siding has three detection sections. The following things occur to put the train into the siding: -The preset schedule for the train is built and a script created for the computer to execute. -The computer starts the train from staging and it begins the trek around the layout. The computer handles signalling as appropriate. -The computer routes the turnout for the first siding in the script to direct the train into the siding. -As the train enters the siding, the computer begins to slow more with each passing segment. So as the train hits the turnout the speed is reduced 30% (from 60smph to 42, for example). -As the train hits the first section in the siding block, another 30% reduction in speed. (42 to 29) -When the train hits the middle section, it slows considerably depending on the signal aspect, (50% reduction...29 to 14) and the computer flags the next section to respond to the signal aspect. -Assuming the signal aspect is red--which it would be in the case of a meet--or if it is a commuter station stop, then the final section does a FULL STOP for the train. -If the train is being detected in either turnout section, an alert is sent to the "dispatcher" and signals will continue to correctly protect the block. (and maybe cut power to the facing turnout?)
It really comes down to what type of detection and how are you going to implement the logic. I have written on signaling a layout a couple of times in the past here on Trainboard on the basic requirements for signaling. There are 5 things you need: 1. Detection - The most realistic is current detection such as that provided by the Digitrax BDL168 and BD4 boards. If this is used you don't need to break your blocks up into multiple little blocks. Also, you don't need to double gap just single gap. I would recommend you make your turnouts blocks. The block would run from the signal head through the turnout until you are at the signal for the opposite direction. 2. Turnout position reporting - You need to provide information on the position of your turnout in order to properly display signal indications. This can be done with stationary decoders such as the Digitrax DS54 and 64. 3. Logic - This is the hardest part. How are you going to do this. Hardware is the cheapest but requires relays or dedicated boards. The more complex the tracks the harder it gets. Software with a computer is the most expensive (unless you have the computer) but as the track work becomes more complex the better and more cost effective it becomes. 4. Direction determination - This really falls into the logic area but I break it out because most people forget about it. Unless you are only going to have single direction then you need to think about this. If you use a computer and software it is very easy to do and it doesn't require having multiple sub-blocks within a block. 5. Signals - The final thing you need are the signals and the device that will drive them. They can be driven by either a dedicated signal logic board or some signal driver device such as the Digitrax SE8c or CTI signalman boards, or Oaktree boards. If you want to go the computer method I can help you. If you want to go the hardware way, you will have to wait for someone else's input. I am using Digitrax and CTI hardware. I have a friend who is using NCE, CTI, and Oaktree. We both use RR & Co software but there is an active community of CTI hardware/software users. JMRI software is available for free but I have never used it. David
Thanks for the responses. :teeth: David, Thanks for the input 1. I do plan on using BDL168s and/or BD4s unless something better comes along in the year or so before I get to wiring block detection. So you think I should be fine with one detector per block? From what I understand about direct home wiring I single gap each detection section and double gap the ends of the detection zone, is that right? As for detecting within an interlocking would you break up your complex interlockings into sections like this where each color would be a different detection section? 2. I'll be using DS 64's, but I haven't been able to determine how you accomplish position turnout reporting, I know with the DS 54 you use a microswitch for position reporting, or 2 microswitches for exact position reporting. Are you familiar with how you accomplish this with a DS64? I wrote Digitrax and they just said it was "built in" to the decoder. 3. Logic - I'm still scoping out software, and want to get my layout up and running so I have something to test a demo on. I like the RR&Co. I have several computers, but will probably eventually buy a cheap laptop to run my layout. 4. Direction - I plan on doing a lot of idustry work and will be needing to change direction within my blocks frequently to run around trains, spot cars, etc. It's good to know that it is fairly easy to implement using software. 5. Signals, I'll probably use SE8c's, but am still on the lookout. From what I understand about the SE8c, without software your pretty much limited to bidirectional signaling or the tree legs of a single turnout, so everything above has to be implemented before I get to signals. Doug, I'm also interested in how to set up braking and stopping scenarios like you described. You brought up a good point about protecting trains within a turnout, which would make it seem like a necessity to detect occupancy within an interlocking. Josef, Thanks , I hadn't really considered IR sensors, but may check them out. Thanks Greg
I do believe the CTI solution for signaling is cheaper then digitraxs. From what I have seen and played around with I like it a lot better to.
1. I do plan on using BDL168s and/or BD4s unless something better comes along in the year or so before I get to wiring block detection. So you think I should be fine with one detector per block? From what I understand about direct home wiring I single gap each detection section and double gap the ends of the detection zone, is that right? Response: Yes, one detector per block is all that is required. In some special situation like crossing signals you will need additional sub-blocks. Single gaps are all that are required for blocks. The only time a double gap is required is if you break your layout up into power districts or normal gapping for potential short circuits at cross overs and reverse loops. This is not part of the detection though. The BDL168 has 4 "zones" on the board. This allows you to have blocks in different power districts on the same board but all 4 of the zone can be in the same district. As for detecting within an interlocking would you break up your complex interlockings into sections like this where each color would be a different detection section? Response: Yes, you blocking looks good although you have a couple of singles in one location that are not required. 2. I'll be using DS 64's, but I haven't been able to determine how you accomplish position turnout reporting, I know with the DS 54 you use a microswitch for position reporting, or 2 microswitches for exact position reporting. Are you familiar with how you accomplish this with a DS64? I wrote Digitrax and they just said it was "built in" to the decoder. Response: I have not looked at the DS64 but it probably operates the same as the outputs on the SE8c for turnout control. A message goes out on the LocoNet that says a turnout is being thrown to a specific position. The assumption is that the turnout has moved to the commanded position. The DS54 didn't assume this, it used a physical feedback mechanism. I have used both devices and both work but you have to cycle the turnouts control by the SE8c to make sure they are in synch with the software. 3. Logic - I'm still scoping out software, and want to get my layout up and running so I have something to test a demo on. I like the RR&Co. I have several computers, but will probably eventually buy a cheap laptop to run my layout. 4. Direction - I plan on doing a lot of idustry work and will be needing to change direction within my blocks frequently to run around trains, spot cars, etc. It's good to know that it is fairly easy to implement using software. Response: I have implemented an auto detection of direction logic with RR & Co with a manual override. This allows a train to enter a signaled block from one direction and then exit the other direction with proper signal indications. The only problem with this is if you want proper signal indications you may have to have immediate access to the computer. 5. Signals, I'll probably use SE8c's, but am still on the lookout. From what I understand about the SE8c, without software your pretty much limited to bidirectional signaling or the tree legs of a single turnout, so everything above has to be implemented before I get to signals. Response: This isn't correct at all. The SE8c is a dumb signal driver board. It controls up to 32 signal heads that are not tried to any specific track configuration. You can make each individual output do whatever you want with computer or switch commands. I have a couple of output generating the flashing lights of a crossing gate on our club layout. On my home layout I have one output triggering a crossing bell. Also, don't forget that the SE8c board has 8 turnout outputs (Tortoise Only) and input for 8 detection blocks with the BD4. This is the most cost effective board I have found. It is more complicated to wire than CTI boards so there are cost/ease of use trade-offs. I hadn't really considered IR sensors, but may check them out. Response: If you are using a computer software, don't forget you need an interface to get IR sensor information to the computer. CTI boards offer this if you want to go with IR. David
Thanks David, I've considered using the SE8C to drive my tortoise motors, but decided against it because I wanted the route control of a DS64 and more importantly, I thought the SE8C didnt' report switch postion to Loconet. But after reading your reply, I'm reconsidering this. I already have one DS64 which will support 8 routes (which is more than enough for this layout), so for the price of two more DS64's I can instead buy an SE8C and get the added benefit of a signal driver. Great Tip! I downloaded the TrainController manual last night for another read through and realized this is where I got the idea for multiple detection sections per block. The manual discusses using multiple sensors to set up braking districts. I was thinking it was needed for direction of travel. I do have to add that this looks like one slick program. Thank you again for taking the time to reply. Greg
LOL, thats a little bit different. On my freinds layout we are using hte traincontroller and I remeber setting up the breaking as you described. Not to sure how it works, since I don't use traincotroller.
