Hey All,
I'm looking for some comments on track detection ciruits. On my last layout I designed and built custom circuits based on a modified Twinn T circuit. This time around I am less ambitious (and had to dispose of all my test gear and most electronic parts), and will probably go with manufactured units.
Do any of you have experience with currently available commercial detection circuits? Good or bad, I would like to know.
I am building in HO scale, and will be using both older Athearn/Atlas diesels as well as more recently released (i.e. lower current draw) locos. I am using DC, so while DCC-only circuits interest me, I won't be using any.
Also, any comments on "resistor equipped" wheelsets for rolling stock are welcomed.
Modeling the CN in Louisiana before it was there
Modeling an HO gauge freelance version of the Union Pacific Oregon Short Line and the Utah Railway around 1957 in a world where Pirates from the Great Salt Lake founded Ogden, UT.
- Photo album of layout construction -
DC or DCC? It makes a difference. For DCC I am using current coil type detectors. No voltage drop. They won't work on DC because they usually only detect current flow in one direction.
Resistor wheels - as long as you have metal wheels, they are not difficult to make. Thick CA, some SMD resistor, and solver conductive paint. You'll get sticker shock when you see how much the silver paint costs, but that tiny amount you get for $40-$50, I did about 100 wheelsets and still have half of it left. Wheelsets with slipper plastic axles, like the P2K ones most of mine are, need to have the plastic scuffed up and the blackinging ont he back of the wheel scraped off for good contact, the conductive paint doesn;t want to stick to the slipper plastic. On others like Intermountain, with metal axles, it's just a matter of gluing the resistor on leaving up between the axle and wheel back on the insualted wheel. That would use so little of the conductive paint you could probably do 1000 with one bottle of paint. Just don't let the CA get on the connection pads at the ends of the resistor, it it will be forever insulated.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
rrinker DC or DCC? It makes a difference. For DCC I am using current coil type detectors. No voltage drop. They won't work on DC because they usually only detect current flow in one direction.
rrinker Resistor wheels - Thick CA, some SMD resistor, and solver conductive paint.
For my track detection I am using NCE BD-20 current detectors to energize a relay for my relay based signal system. The NCE units work with DC and DCC but they need a 12 volt DC power source. The BD-20 switches the DC-. The BD-20s run about $11-$12 and the mini relays run $1.50-$1.75. I am spacing my signals based on my longest train so I may not need resistor wheelsets, although I have the resistors and paint. One resistor wheelset per car should be enough for me.
Theoretically, if you placed a rectifier on the secondary of the coil, it would be able to work - you'd need to adjust teh comparator circuit to work with the reduced voltage, and/or use more turns of the feeder wire to compensate for the fact that with DC you aren't always at full voltage. You'd also still need some sort of bias current to be able to detect a stopped train sitting in a siding. A better option for DC is Dr. Chubb's Optimized Detector, it's made for DC. DCC does make this easier though.
MikeyChris
I don't know what detection methods you have looked at, but there is a thread on inexpensive Infrared Detection circuits that might offer you an alternate detection method:
http://cs.trains.com/mrr/f/744/t/219552.aspx
(To make the link work, highlight it, then right click on it, then select 'Open Link' or words to that effect from the pop up menu.)
These will work with either DC or DCC and they might save you some money by not having to buy resistor wheel sets.
Dave
I'm just a dude with a bad back having a lot of fun with model trains, and finally building a layout!
hon30critter ... you an alternate detection method: http://cs.trains.com/mrr/f/744/t/219552.aspx
... you an alternate detection method:
Thanx Dave. I am interested in learning all I can about all types of track deection circuits, so this IR circuit was certainly interesting to me. I have usually preferred sensing track current because it seemed to me to be more "inclusive". By that I mean any resistive path accross the rails (regardless of position) would be detected, whereas photo detectors by their nature cover only a limited amount of track. However, my current layout is very simple (all one level, no hidden trackage, only 5' x 11') and most blocks will only be as long as a train anyway, so that point is moot. So, I may indeed try some form of photo detector, and if I do, IR will most likely be the type I use. One question though: The train room has a large flourescent fixturew (8' long two lamps) over the work bench, incandescent track lighting over the layout, and natural sunlight coming in through two medium-sized windows. Will all these lights of different color temoerature affect the response of the detector (i.e. cause false detection)? I noticed there were two circuits in that link you posted - one for incandescent lighting, the other for flourescent. So I gather the trick is to play with the biasing of the Darlington transistor and/or add an LED as a light source (in close proximity to the sensor).
Thanx again.
retsignalmtr For my track detection I am using NCE BD-20 current detectors to energize a relay for my relay based signal system. The NCE units work with DC and DCC but they need a 12 volt DC power source. The BD-20 switches the DC-. The BD-20s run about $11-$12 and the mini relays run $1.50-$1.75. I am spacing my signals based on my longest train so I may not need resistor wheelsets, although I have the resistors and paint. One resistor wheelset per car should be enough for me.
I'd be very suprised if a BOD20 can work with DC. It is a current coil type detector. The rating per the instructions is that with 5 turns through the coil, it can detect as little as 1.5ma. Sounds pretty good. But if you have a singel car with 4.7K os resistence through the wheels, that's 7 volts needed. Unless you have some pretty poor locos, 7 volts should have them moving at a pretty good clip. Totally stationary, 0 volts - no chance. If you look to older (but pretty simple and cheap to build) options like the Twin-T, they passed a bias voltage through the rails to detect standing trains without them moving. With modern low current motor is might still run the train. The Chubb OD shows 12V through a 2.2K resistor to provide a bias current when there is no power applied to teh detection section.
IR detectors, either refelctive or block the beam type, are perhaps mor euseful. There are no track conenctiosn at all. With a decent design, they can tell when a block becomes occupied and when it is clear, not just that a train is over the sensor. Depending on block and train lengths, a single sensor in the middle may be sufficient to operate signals, at least for reasonable looking signal animation.
The BD-20s are not direction sensitive so you have to look at the signals to see what direction is indicated. That $12 is just for the detection blocks. Each automatic signal has three relays, 1 for track detection + the BD-20, 1 home relay and 1 for distant controls. Additional relays are used for traffic direction, route selection and switch repeaters. Home signals have five relays. Signals are three color. So the cost adds up, but it is less than a commercially made system, is easy to trouble shoot and it works the way I want it to. I have a couple of BD-20s set up on a museums DC layout to control crossing flashers and they work great.
BTW, the BD-20s will detect the current drawn by the decoder in an idle loco on DCC so stopping a train will not shut off the current detector.
Yes, in DCC there is always power flowing in the rails. Or with older carrier control systems - always power, so a stopped train can be easily detected. With DC, stopped train = no current flowing through the rails so a current detector cannot work. OP said he is using DC. If detecting moving trains is good enough, it will work, but if you stop a train, detection will drop out and the signals will show the block clear when in fact it is not. It takes some extra doing to make block detection work in DC for disconnected blocks.
Some people have had problems with overhead lighting interfering with the very simple detectors that we have been referring to. It is crucial that the heat shrink tubing on the IR receiver be installed properly, and the there be no lights directly overtop of the receiver. The tubing has to shield the receiver so that no direct light from overhead lighting can get to the IR receiver. I haven't tested them with LED strip lighting to see if they will upset the detectors. Personally I'm not worried about that because I only intend to use the simple detectors in hidden staging.
I don't have a layout yet so I have only been able to bench test them under combined incandescent and mini florescent lighting. They worked fine.
My suggestion would be to build a couple and test them under your particular circumstances. If your overhead lighting messes them up then you have two choices - spend the money on the more complex IR detectors or use a different system. Rob Paisley offers a huge variety of DIY model railroad circuits. Here is a link to his IR detectors:
http://home.cogeco.ca/~rpaisley4/CircuitIndex.html#Light
There are also commercial detectors available but of course, the cost jumps way up. However, they do include systems that eliminate the sensor from flashing on and off as each car passes over it. That makes them more suitable for controlling other devices. The simple detectors are really only suited to illuminating LEDs because the signal does flash on and off between cars. If you want to use the simple detectors to operate crossing signals for example, you have to add in a timer to stop the on/off constant flashes. I think Rob Paisley's circuits include that feature, or at least some of them do.
Here is one example of a commercial detector:
http://www.azatrax.com/ir-model-train-detector.html
rrinker Yes, in DCC there is always power ... With DC, stopped train = no current flowing through the rails...
Yes, in DCC there is always power ...
With DC, stopped train = no current flowing through the rails...
When I used my old modified Twin T detectors, I just added a bias resistor (to each block) and power supply for same. It worked pretty well (on DC). The only problems I ever had were
1) noise causing detectors to drop out (cured with bypass caps) and 2) turning power to a medium to fast moving loco off very quickly caused the detector to change direction of detection (I assume due to back EMF from the motor rapidly changing from being a motor to being a generator). I was using direction sensitive detection to automatically throw turnouts and control track polarity of reverse loops. If I ran at anything close to realistic speeds all was good.
Once again, thanx for the comments.
Hi Dave,
hon30critter My suggestion would be to build a couple and test them under your particular circumstances.
My suggestion would be to build a couple and test them under your particular circumstances.
Absolutely. I always mock everything up first because (as we all know) Murphy is waiting to screw up any project!
Thanx for the links, they are helpful.