Hi, I have a reverse loop at each end of my layout. They are both in hidden sidings. The layout is basically a single-track dog bone. I have designed electronics to allow the continuous running of trains using Peco point motors and switchgear. It works perfectly when operated manually so I set about fitting it in. I had intended the system to be operated automatically by reed switches but to my disappointment when I fitted a reed switch into the system the locos failed to set it off! I was under the impression that the magnets around the motor would be enough to trigger the reed? This is an N scale layout. I have used reeds before with OO but with larger motors I guess there is a stronger magnet? I looked at using inferred detectors from www.micromark.com but they are triggered by any moving object, I only wanted the system to trigger as the loco went over as not to burn out the point motor. Does anybody have any experience with reeds in N or any better way to trigger a system with a train? Many thanks, James
Good question. Can't wait for the responses. My layout is similar except I have the ability to go around one of the reversing loops with a continuous oval. I too want to cover the reverse loop pictured below with another reverse loop that will serve as a coal branch. I have been thinking about reed switches and moreso about the photo-cell type triggering. Also with DCC, I am considering some of the sensor mechanisms available. Timely thread... I was thinking about starting the same.
Pleased to see I am not the only one trying to do this. I am not using DCC but the problem is the same, what triggers the point motors? I did experiment with micro switches years ago but that only led to derailments!
You might want to look at Heathcote Electronics there in the UK. They are the the producers of the IRDOT that Micro-Mark sells. They have a complete line of electronic controls. Do a Google search for Heathcote and you will find them. David
Going IR or whatever is gonna punch the cost up, so first off I'd try gluing a small magnet under the loco. In the fuel tank for diesel; on a pilot truck or in the ashpan for steam? As it's N space might not allow this, but these days there are some very powerful small magnets available. Or you could consider putting a magnet under the first car behind the loco. Practicality of that depends on how you run your trains of course, but it would avoid multiple switch operation if you have 2 or more locos on the point.
I have been giving this a lot of thought, I can’t justify the cost of inferred and I don’t want to have to stick magnets to everything. But I think I have solved it. So how about this for a solution. Instead of a reed switch or inferred sensor I wondered if a track circuit detector would work, some way of using the power in the track to trigger something with out disrupting the power to the train? So, I cut one of the rails leading to the point twice leaving a section of track about ½ and inch long. There must be no electrical connection left. I used a piece of link wire to jump the power across this piece of track so only the short piece is un-powered. Then I connected the short piece of rail to the input side of a relay and then return wire to the uncut rail. The theory behind this is when a loco, or any thing with metal wheels, runs onto this short piece of track it will complete the circuit and power the relay. The train did not have any loss of power or stall as the section was so short . The relay I used has a sensitive trigger so it fired with lightning speed, even if I had the controller on the lowest possible setting and bridged the gap with a screwdriver . The output side of the relay will be connected to the 16v AC circuit for the point motors. The risk here is with the loco and stock having lots of metal wheels, the relay will trigger as every wheel enters and leaves the short section, the point will keep trying to switch and burn out. So, if you use a Peco Accessory Switch or equivalent, which is switched by the point motor, to feed the point motor it will protect itself. So I set this up, a train enters the reverse loop and travels all the way round, just before the point it hits your short piece of rail and triggers the relay and point motor. Once changed the feed to the motor is cut and although the relay will still work the point motor is electrically cut of. The train rumbles past. The next train enters the loop in the opposite direction and will travel all the way around until it gets to the other short section of rail. It triggers the other relay, which changes the point and then cuts of the point motor and we start again. The appealing thing is the cost. The inferred system cost £14 ($25), a relay costs £1.69 ($2.90) and considering you need two for each point it can be expensive. Kitbash - I don’t now how well this would work with DCC! If you try this I would be interested to know? This worked perfectly; I can draw a wiring diagram for any body that is interested? Regards James
James: I think I got what you are saying but if you could add a wiring diagram it would be great. And I think it will work with DCC I will give it a try. Dan
Nice idea, James. The issue regarding the metal wheelsets is really mute as the only time it will be an issue is when the wheelsets are bridging the ends of the rail, which can be rectified with a sufficient break between the ends and an insulative piece glued in place (or insulated joiners used) to provide mechanical strength. The only time the relay will operate is when a locomotive(s) crosses over the section (circuit through the motor). To prevent the relay constantly picking and dropping, maybe a small capacitor across the relay coils to hold it up a few seconds, enough to clear the locomotive? I'd like to see a diagram as well, please.
James, your idea sounds pretty good. At first blush, I don't see why it wouldn't work w/ DCC. I'd have to ponder this. I have been thinking about this and will probably go ahead and use something like the Digitrax Detection circuit. At $30 bucks a pop/4 outputs, I would need two at $60, + some chumpchange for a couple of indication lights, etc. While a little on the pricey side, I consider it a deal. Heck, the last engine I purchased (two weeks ago), an Atlas RSD 4/5 cost me 84 bucks "on sale". The benefit of the detectors would be a huge gain on my layout. Plus, it would allow me to begin covering that loop w/ the new coal branch loop that I am aching to start building. Just gotta make sure the loop that will be hidden is in perfect shape before it "goes underground". Please do post a diagram. I like your idea as a possible alternate.
From James' description and my own thoughts, it sounds like it's wired thus: Which seems a fairly effective way of doing it. My only thought is if the Accessory Power was DC, capacitors fitted across the relay points would reduce arcing and so extend the contact life.
Gats – The system relies on metal wheels bridging the gap between the rails to complete the circuit. It is not completed through the motor. If you look at the great diagram done by Another ATSF Admirer, if the short track section and the blue ground rail are linked there is a circuit but it is unpowered. If the brake in the track is linked the circuit is completed through the controller and it works every time. However after fitting it this weekend I decided that having the relay always firing was annoying so instead of protecting the point motor with the solenoid switch I intend to protected the return from the relay. Another ATSF Admirer – Perfect, I have made a few improvements as I installed it, once I know they work I will re post. I have now opened a Rail Images account so I can post a wiring diagram of what I now have. I will post when finished.
Of course, it is possible to wire the circuit so the relays don't switch constantly. (James: not meaning to steal your thunder here, just bored and think this is an "interesting" problem ) Rough sketch of one way to do it, hacking up my previous drawing: Again the points were last set to straight, so the "straight" relay will be ignored until the "curved" portal is engaged. Only problem I see is that the "short" from red portal to red track has to be long enough to guarantee the points solenoid fires. It might be worth including a capacitor discharge if you need just a car with metal wheels to trigger this reliably. Shouldn't be a problem with a Diesel / Electric model, as they have enough wheels to bridge the portal long enough to throw a tortoise An 0-6-0 might even do the trick. You know, this circuit seems simple enough and effective enough, I might use it myself... Imagine if you wire more than one solenoid to the relay... So as well as alternating clock-wise / counter-clock-wise on the reverse loop, but have trains go off down different branches every second orbit. Magic "hands free" trains.
I have now tried three different types of relay. The last one was a 12VDC 10A/125VAC Mains relay. I have tried a 2A 12Vdc Sub-miniature relay and one other. I am still learning with electronics but my theories are this. 1. The relay switches so fast that there is not enough time for a complete connection. (I find this strange, as I only need to tap a wire in the pins and the system fires 100%). 2. The relay is putting up a lot of resistance and taking all the power? 3. I need a bigger relay. I don’t quite understand what you mean by the capacitor discharge? Are you referring to the CDU or something else on the trackside to hold the relay closed long enough to complete the process? Any help at this point would be greatly appreciated as I am running out of hair to pull out. Regards James
Sorry, there was a bit missing from the last post, it should have read! Another ATSF Admirer – Fantastic, we singing from the same sheet. Your diagram is exactly what I have done. I have also, as you suggested, used another point solenoid from the same relays. This one reverses the polarity if the main feed to the layout this allows continuous running in a single-track dog bone layout! But I can’t brake out the campaign just yet. This is why I haven’t posted a wiring diagram yet! I am using a CDU to power the solenoids, if I manually short the relays by quickly touching a wire or even a screwdriver across the terminals the system works 100%. The points change with positive snap, each moving the full distance very quickly. But if I let the relay make the connection the solenoids struggle to move, the points only go half way. They sound like there is now power there. I have now tried three different types of relay. The last one was a 12VDC 10A/125VAC Mains relay. I have tried a 2A 12Vdc Sub-miniature relay and one other. I am still learning with electronics but my theories are this. 1. The relay switches so fast that there is not enough time for a complete connection. (I find this strange, as I only need to tap a wire in the pins and the system fires 100%). 2. The relay is putting up a lot of resistance and taking all the power? 3. I need a bigger relay. I don’t quite understand what you mean by the capacitor discharge? Are you referring to the CDU or something else on the trackside to hold the relay closed long enough to complete the process? Any help at this point would be greatly appreciated as I am running out of hair to pull out. Regards James
Capacitor Discharge Unit = CDU, so yeah, that's what I ment. Any system which uses the charge in a capacitor to fire the solenoids, instead of the trickle of the accessory output which can burn out the point solenoids. As to why it doesn't completely fire: consider the accessory switch closely. If it has a no-mans-land in the middle of the points movement where neither "straight" nor "curved" (blue) is connected, this could be why the relay isn't working... The accessory switch removes power from the relay (blue) before the solenoid has completed moving the points, thus stopping the (green/orange) power to the solenoid. I suspect accessory switches would have just such a dead area, to reduce the risk of having both "straight" and "curved" connected - which would of course short the frog in the classic use, and fry the power-supply / pop the circuit breaker. If by "Tapping a wire in the pins" you mean the relay outputs (NO, C), then this explains my thoughts nicely: the screwdriver/wire keeps the solenoid / CDU energized longer, whereas the relay stops after the accessory switch moves part-way and so removes the energy. In short, the relay needs to remain powered long enough to guarantee the point solenoid has travelled the full distance, and the accessory switch may be hampering this by going "open" too soon. In diagrams, to exagurate, a bad switch which "goes open" at about 10% of the point throw: (the black ball is attached to the throw bar, and connects the orange contacts as it moves) As for solutions... Re-jig the accessory switch to switch "later"? Could be difficult, involving anything from finding a different switch make to trying to use two switches in parallel - one for straight, one for curved. With this diagram, both "straight" and "curved" are connected for a long time, meaning the relay stays energised for the whole motion of the points. It could be very hard to find a switch that works like this, however.
I think I've got it. Although this will take some explaining. By using double-pole relays for each portal, it is possible to "latch" the relay until the (even exagurated) accessory switch makes contact with the other side. So keep the "straight" relay (RS) energised until the "curved" accessory contact conducts. By the time that happens, the points should be sufficiently well thrown to no longer be a problem. </font> RS : Relay Straight</font> RC : Relay Curved</font> RLS : Relay Straight, Latching</font> RLC : Relay Curved, Latching</font> RbS : Relay Break Straight Latch</font> RbC : Relay Break Curved Latch</font> As drawn, the train last hit portal "Straight", and energised that side. RbS is actually conducting now (that NC connector is open). So long as RbS is energised, RLS is ineffective, and so RS is ignored just like for the last diagram. When a train hits portal "Curved", RC energises, connecting RLC which will continue to provide power for RC, irregardless of the accessory switch opening the "Straight" contacts early. After some time, the solenoid moves to "Curved", making contact for RbC, which now "breaks" the "latch" of RLC, and so RC de-energises. The connection marked Accessory Power can be either a standard accessory power pack, or a CDU as you chose. The downsides are: </font> this is much more complex - took me nearly an hour to draw it right on paper - so more can go wrong and it will be harder to wire up right. </font> One or other of RbS and RbC are energised at all times, so they have to be rated correctly for their current to avoid burning them out</font> If relays in .uk cost anything like relays in .nz, this will cost almost as much as the equivilent digital circuit using optical sensors and some simple IC flip-flops</font> I think if I ever made a circuit like this, I'd go for the digital circuit, just so I can wire it to a computer. Heck, just make an occupancy circuit and let the computer throw the points James: Let me know if this makes sense or works in hardware
Correct me if I'm wrong here, but all this is going to do is derail every train as it comes out of the turning loop. Consider: Loco crosses detector at end of the main line. Detector throws switch to main route. Loco goes around reverse loop, crosses second detector on approach to switch. Detector throws switch to diverging route. Loco crosses detector at end of main line on the way out. Detector throws switch to main route, thinking this is another train entering the loop. Switch throws under train, derailing it. A few solutions: the detector at the throat has have to know which way the train is passing in order to ignore the loco as it exits the return loop, or the switch can be locked with a timer to prevent movement for 30-60secs after it has been thrown to the diverging route. Another option is to go the mechanical route: use a spring switch. I don't know if these exist in N, but they're often the prototypical solution to this problem.
Longwalker: the trick is to have both portals beyond the points (frog end), not one on the approach track (points end): -S---/--- -C--/ portal --- points --- approach portal --- / This way the train never hits the "wrong" one while crossing the points. otherwise it would derail! Now the only way to derail is if the train is long enough to be touching both portals at once.
