i've been wondering how to implement a block detector that detects an entire train without needing some/all cars to draw current to allow purely electrical block detection. I don't want to add resistors to every truck to allow purely electrical block detection.
An approach i've not read about is to combine electrical block detection with optical detectors that monitor each end of the block so that when the engine leaves the block and the electrical detector is no longer active, the optical detectors will remain active until the last car leaves the block. The wired-OR of all 3 detectors indicates the block is occupied. The only condition where this won't work is if cars were left in the middle of the block.
has anyone tried combining electrical and optical detection as i've described?
greg - Philadelphia & Reading / Reading
Greg,
Most of the commercial optical circuits have a 'memory' or delay(many times adjustable) so there is just an 'entrance' and 'exit' detector. Even if the train is 'between' detectors, the signals will stay red until the 'timeout' runs to completion. Logic Rail has both optical and IR detection systems. I would highly suggest going the IR route. Light sensitive optical detectors do not work in low light conditions, and can be very sensitive even with normal lighting. IR systems are 'rock solid' when it comes to detection.
https://www.logicrailtech.com/index.htm
Myself, I use South Bend Signal Company IR detectors, and they are very reliable. The detectors are $30 per block, and willl drive most HO signals available, including semaphores. Their logic will drive a full ABS system with approach lighting.
http://www.sbsignal.com/
IR systems are not perfect solutions as they can lose a car between blocks if the time delay runs out. But there is no metal wheelset modification needed, and and from my experiance, they are more reliable detecting a train. On the other hand, the current sensing systems are more prototypical. I have 8 blocks of ABS installed and the system has not failed me in over 5 years. You can 'build' your own detectors, but for $30/block I got a working solution with great technical support.
As far as using up too much 'juice' with resistors, they draw so little that i would not worry about it. The big issue is modifing the wheel sets or buying new ones.
Jim
Modeling BNSF and Milwaukee Road in SW Wisconsin
Unless your blocks are abnormally short or you regualrly tun huge trains, all you really need are the loco and the caboose to be detected with resistor wheelsets. It actually doesn't take all that long to make up a bunch of resistor wheelsets - I prefer 2 per car, all teh other club guys did 1 per car. There's a good link I have on my home computer showign how to make them. The only reason it takes as logn as it does it you have to stop and let the glue and the conductive paint dry before proceeding.
The load is absolutely negligible. 100 cars, with a pair of 10k resistor axles per car, at a DCC voltage of 15V, is .27 amps.
Each detection type has advantages and disadvantages - resistance detectionc an't really do specific point detection, but it is far superior for block occupancy. When you want to trigger an action when a train hits a specific point, IR/optical detection is far superior.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
I have been using current detection on several T-Trak modules with great satisfaction for a couple of years now. The NCE-BD20's will detect current from a decoder sitting idle in a block. I plan on having several short blocks on my home layout that will be shorter than some of the trains I run so having resistor wheels on just the caboose may not be desireable. I am concerned about the effect of the cumulative resistance of having so many resistor wheelsets in parallel. Using a pair of 10K resistors on one car reduces the total resistance to 5K ohms. So a 10 car train with one resistor wheelset per car would have a total resistance of only 1K ohms, not including the locos themselves. A 20 car train would lower that to 100 Ohms. If current draw with two resistors is .27 amps, would the draw be 1.35 amps with ten resistors in parallel? Three N gauge locos running MU'ed pulls .31 amps with my PSX breakers set for 1.27 amps. Are my calculations correct?
No, sorry but your calculations are not correct (and you'll be relieved that they aren't!).
For sake of argument and keeping the math simple, let's assume an equivalent DC voltage of 15 volts across the rails. Each 10K ohm wheelset will draw 0.0015 amps (15 / 10000). Since all of the loads (resistors) are in parallel you simply multiple the current going through each resistor by the number of resistors. So, 20 resistor-equipped axles will draw 0.03 amps (30 ma). 100 of them will only draw 150 ma. Certainly neglible drain on the booster compared to a locomotive!
Chuck
OK. Thanks. I see where I made my mistake.