I dont mean to be annoying or anything but im looking for some feedback from other modelers. I enjoy railroad modelling plenty but i also enjoy electronics and i’m actually studying for my degree in this field. I have started designing a few circuits for model trains that i thought i might be able to sell at train shows and such. I’m just looking for a little feedback about what other modellers look for when they buy this sort of stuff. [:)]
So anyways…
How many modellers out there actually desire a working signalling system on their layouts but either cant or dont have one for some reason? Especially one that could follow a particular prototype.
Should i make my system super easy to install and alter (no soldering iron required) or make it a little more complicated but cheaper?
Do people perfer a system that costs more but has more bells and whistles (so to speak) or one that is simpler but easier to use and cheaper?
Im asking these questions because i dont want to be selling 10 different circuits that basically do much the same thing. Any feedback is very much appreciated because i really have no idea what fellow modellers are looking for. As far as i can tell a lot of the stuff on the market seems very expensive [%-)]
But anyway…thank you to anyone who helps me out here and sorry to anyone else that believes this post is a waste of time and space like some people have been saying about other posts…
My requirements for a large club layout’s signaling system would be, primarily, make it DCC compatible if you must use track current sensors, but preferably use IR sensors. I’m not too sure what you could do with a signaling system that would add “more bells and whistles,” as you put it. A three-position target signal, i.e., green-yellow-red 3mm LEDs on a signal mast, should suffice. Because you’re only concerned with the electronics involved, don’t get confused if other people start clouding the issue by asking for B&O style signals, or blade semaphore signals, etc. It would be up to them to find the type of signals they want to use with your electronics, so you need to study the various manufacturer’s signal masts so you know how they’re wired. And remember, too, that a blade semaphore is going to require additional circuitry to drive a Tortoise switch machine for moving the blades.
I am interested as I have my oregon rail signals built and ready to install. They have leds as I tried one with mini bulbs from miniatronics and they melted the target. So I changed them all to leds. I right now am still running DC. I have 9 grade crossings that will need wired so a cheap but well working systems would be nice. DCC is nice but there are those that can’t afford it, or just to old to change so I’d say make it compatable with both DCC or DC
I’d like to see a setup that would allow for the signal to change from green to red as the train passes it, then if there are no trains within two “blocks” of the signal, either shut down completely or go back to green once the leading train has passed the signal. I suppose that transponding might be the easier way to go on this, although I have seen the IR detectors in action at one of the local HO clubs. I just didn’t think it was quite what I would have looked for in my own operating layout.
I don’t know how well the various software packages that are out there to dispatch and control switches, etc., actually do as far as control. Maybe you’re on to something they haven’t covered yet.
When i said about including “bells and whistles” i meant things like the ability to control both color light signals as well as 3 color searchlights, including the option for approach lighting, the option of running single direction only or bi-directional on a single track, or the ability for the dispatcher to override the automatic indications, etc. Its really easy to make the simple circuit that will only power a red-yellow-green color light but adding the extra things makes it a little more complex and more expensive.
Also why is it that IR sensors are preferrable to some people than current sensors?? I would think most be people prefer current because its basically the same system as used on the prototype…the only downside i can see is that it will only detect locomotives, resistor equipped wheelsets, or lighted cars.
As for the prototype aspect i was wondering about railroads that have systems that arent just red-yellow-green. Is it not true some prototypes include aspects such as double yellow or flashing yellow?
Thanks for the help once again because when i do get around to making and selling these i want them to be what modellers are looking for. I want it to be the system that best meets their needs and all that [:)]
I much prefer current sensors. That give you train detection for an entire block.
There is an excellent model railroad electronics chat group on Yahoo and one for the CMRI system. Check them out and you’ll get a good indication of some ideas people are working on.
I mentioned IR detection in my post because it would be too difficult to go back and re-wire a 20x40 foot HO scale club layout that is already completely wired for DCC operation, and the expense for all of the members to put resistor wheel sets on hundreds of pieces of rolling stock in order to use current detection.
I have stumbled upon a way of making it sort of cheap. However there is one sticking point which comes under the “easy to implement” section. To make my system easy to implement i proposed to used a collection of screw terminals where the stripped wire is inserted into a hole a screw tightened to hold it in place. Easy! Problem is that these screw terminals happen to be the most expensive part of the whole unit - almost the price of all the other parts put together!!! [:O]
So anyhow, for my prototype that i’m building, im using plain old PCB pins that are dirt cheap but also mean that during installing the installer would have to solder the connections for all wires coming to the board or leaving it.
At the moment my inital designs use a system where each signal has its own small circuit board mounted under the layout directly beneath the signal itself. Each board only a couple of inches in length and in its most basic form means that each unit is connected by only 3 wires total plus a fourth to the occupancy detector, in a daisy chain fashion - similar to stringing 3 bus wires and then mounting the signal controllers at each point. If a the user wishes to place another signal along the line between 2 others its a simple matter of cutting the wires and adding the new unit in the space. Likewise if a signal is removed then the loose wires are simply joined together like the old unit was never there. The user could easily just run these 3 wires along side his DCC bus wires (if thats what he is using). If more features are added, more wires may become necessary. By doing it this way it reduces the amount of external wiring required for signals that arent absudly close together. This may be desirable to club layouts who have signals placed a distance apart? It allows the installation of a fully operational ABS system with signals placed sparingly and as more time and funds become available, users can install more signals filling the gaps and so on.
Try http://www.jlcenterprises.net/ they have a rather nice system maybe you can improve on it. Personally I prefer the IR sensors as they detect anything not just a current draw. Also a system with a huge number of inputs is desirable. I have to put microswitches on all of my turnouts so that when I do get the system up and running it will know whether or not the turnouts are thrown. Also being programable with some form of Basic is handy as with a little tweaking the entire system can be adjusted for very efficient operation. Another thing I’d love to see is clear instructions in the literature before purchase so you know what your getting into without going out and buying the manual (Seperately). Also soldering is no big deal I’d rather do a bit of soldering than pay twice the amount for a board.
Well Mr. Tyson my responce isn’t exactly what you were looking for, but this is what I wish for.
A whole new architecture would have to be designed. It would be a plug and play system. Something that can take advantage of the Lenz Railcom technology ( http://www.lenz.com/products/index.htm ). It must have a software program that Railcom (or NMRA generic version) teaches it where the locomotive is and what it’s doing ( the user would mark reference points) and the program knows what appropriate aspect the signals should be and control turnouts as a interlocking system.
Take a look at the turntable indexing system New York Railway Supply is using to get some idea as to what I’m getting at ( http://www.nyrs.com/ny00001.htm ). "The turntable’s indexing positions are programmed into the PTC by the user, who “teaches” the PTC where the lead and service track positions are. The PTC does not use infrared diodes, tab stops, or any mechanical methods for indexing. Once programmed, the PTC remembers the user’s track locations-- even when the system power is turned off".
Also the software program would have a point and click graphics display of a CTC machine ( and /or choice of other prototype system) and another board (window) showing train locations. The whole thing should be able to switch between automatic and manual modes.
A system like that would probably cost more than a brass steam locomotive or a super delux DCC system, but be worth every penny.
I would recomend you “compartmentalize” the units. Some for detection. Some for logic control. Some for signal driving, possibly for different signals. Approach lighting an option, or possibly not. A standardized power supply. All inter-working. If it was compartmentalized, the same circuts could be used for everything except detection, or what ever. Don’t have a particular preference for solder or screw connection - just something solid and good.
MR mag. has published such circuts in the past. See “Teton Short Line” in MR. Also, some from 5+/- years ago. Target, 3light, sema-4, etc.
Personaly, I am working on a “train order” signal, that clears just as the train goes by. Probably use light detection. Would want manual control for the clear - either full green (no orders) or yellow (pick-up on fly) or no change - FULL STOP & SIGN FOR PICKUP!
And Manual clear for that stop.
Thank you to everyone who as replied to this post and i have thought long and hard about everyone’s suggestions and opinions and started refining my designs to incorporate as much as i can.
The system i’m designing is (like Randy recommended) already “compartmentalized” although i had never heard that word before! But anyway its designed with 2 basic units with an optional third. The first unit is the occupancy detector, whether it be current sensing, IR sensing, or something else. Whichever type is used the outputs are identical and can be customized. The second unit is the signal controller which is desgined to work with practically any occupancy detector because the inputs can also be customized. The signal controller itself in its most basic form will provide its output through a transistor configured with an open collector for a vacant block, and an earth connection for an occupied block. In more sophisticated model this could be customized as well. The transistor ouput should be able to drain anything up to 1 amp. I’m hoping to make it able to switch up to 60V as well. This allows the outputs to directly drive a signal using current sinking method, or drive further logic circuitry or power relays, motors, soleniods, etc. The third unit is optional and that is the signal driver unit which would be used for signals that are not basic red, yellow, green, or searchlights. Examples are semaphores, flashing aspects, grade crossing units, switch motors, and such.
My plans so far call for different tiers of complexity with cheap basic units to more expensive, more complex units, although in most cases the units will be interchangable regardless of complexity. For example a simple occupancy detector inputting to a complex interlocking driver circuit. I have to find the balance so that there arent too many tiers and people becoming confused about what they want. All the features of each system would be clearly explained to comparisons presented for example.
Don’t forget to consider the power required for PRR position light signals. They use three lights in either vertical, angled or horizontal position and need more power to operate and switch them.
As far as your question for the screw terminals, as far as I’m concerned I would go with solder connections since they are much more reliable than screw terminals.
ndbprr:
In my initial design each output line is fed into a transistor configured for an open collector output. This means that the ouput switches between an open circuit to the earth rail. If your signals are configured for the current sinking method, which is common anode (where LED’s connected in parallel share the same power source) then it will easily be able to handle PRR position lights because the sinking capability will be at least 500mA and possibly even a full 1amp. If this sounds confusing dont worry because i plan to publish plenty of data, wiring diagrams, suggestions and examples in the manuals and on the internet. Another thing about this method is that a resistor connected to the board by the user will directly control the amount of current supplied to the LED’s. In this way it does not matter what power supply is connected nor what the voltage drop is across the LED itself. By doing this the user can easily control the brightness of his LED’s and mix and match colours, models, and manufacturers without having to consider what power supply he is using. The only thing he will need to think about is how bright he wants them. The board itself will have its own power source to provide power to the signals.
Frankwin:
I have since located a cheaper supplier of screw terminals that could significantly lower the cost of including them. However whether or not people prefer soldered connections or screw connections the boards will be identical with the only difference being whether PCB pins are included, or screw terminals included. I may be able to offer both variants together for those who have a particular preference.
Thank you all for you input once again. Everyone’s opinions and constructive comments are a huge help in refining my designs and also the ways i will go about getting them to the public.
Looks like I will need to order it.