Anyone have Ideas for an optical sync on the tsunami decoders? As I see it, the problem lies in getting track power to the cam (tan) wire.
Using the internet, and my almost nonexistent knowledge of EE, I came up with the following idea:
Driving the emitter side of an IR reflective sensor from FX5 (or FX6), and using the detector side to switch a photo-relay that would send track power to the sync connection.
I realize I'd need some resistors, and I have no idea where to get the power for the collector on the detector. But I also know we have some real electronics whizzes on here. So, any Ideas?
Problem is, it needs to see the track power to work. So what you would need is a method of controlling a switch that passes track power to the decoder for a chuff.
While optical is an idea, it probably would not be practical. Same as using a magnetic sensor, which would work until you ran over an uncoupling magnet...
A chuff cam works by passing the track power via the cam mounted on the wheelset and a wiper.
Here's a more developed idea, maybe someone more knowledgeable than I can tell me if it works.
Circuit Idea #1 Running a photo-sensor off of Fx5
(!Disclaimer - I have not tried this yet, and I have no Idea if it even could work. try at your own risk!)
Assumptions: Function common is 12V+ (per vague wording in documentation, and diagrams in same) Fx5 output = Ground. and Fx5 is always on (when you want sync). These may be entirely incorrect, but I'm hoping to hear back from soundtraxx soon.
Operation: With Fx5 on, the IR LED in the sensor is always on. When a reflective band passes the sensor, IR is detected and the output goes High, switching on the track voltage side of the Photo-relay (IC 1) and generating a "chuff."
In this case, the black and white pattern shown in the documentation from soundtraxx should be inverted.
Parts: (links open in new window) AQY272 $12.37 (this could possibly be replaced by a cheaper model, but I went with 2A capacity on the load side)QRB1133/4 $2.80 (more expensive than an LED and a phototransistor, but already packaged and aligned)R1 - I know that the sountraxx documentation says 680 ohm, 1/4 watt. But my little resistor calculator gave me 220 ohm 1 Watt for 50mA forward current at 12V $0.50
Parts: (links open in new window)
AQY272 $12.37 (this could possibly be replaced by a cheaper model, but I went with 2A capacity on the load side)
QRB1133/4 $2.80 (more expensive than an LED and a phototransistor, but already packaged and aligned)
R1 - I know that the sountraxx documentation says 680 ohm, 1/4 watt. But my little resistor calculator gave me 220 ohm 1 Watt for 50mA forward current at 12V $0.50
Is it practical?
I suppose it depends on which scale you're in. The largest part is the sensor, and it can be trimmed down (slightly) to fit. So it might not work for N, but possibly HO, and anything larger should have no problem. Seems easier to me than trying to solder a disk on the back of a driver.
In any case it's only 3 parts, 2 of which are quite small. The unmodified sensor is 17mm wide at it's widest, but the thickness is only 5mm, plenty small to fit between the drivers in HO.
I hope I can get some useful feedback on this from someone who knows more about electronics.
Your concept is fine. The circuit is in the ballpark but you won't need any 2A rated parts. Your current draw is going to be mainly that of R1 (the other stuff will draw some but not much) and at 12 volts, that's gonna be about 0.05A (50mA). Odds are really good that you can increase the value of R1 to reduce the draw even further. Even a small 2Nxxxx part will handle that-- especially if you get it in a metal can which will offer more heat dissipation, but I doubt even that would be an issue in this case.
The other thing you might consider is that you could also probably use a 'slotted-disk' type arrangement and there may be smaller parts available for that-- (hint: think computer mice-- chock full of lots of interesting parts-- though you'll wanna get the 'ball' type and not the 'opto' type).
Another point to make is that you probably don't need a perfect reflection-- I can tell you from experience in other arenas that you might just be able to flood the disk with light and put the photo-diode (receiver) very close to the disk surface and get enough of a blip from that to work. You might need to adjust the "amplification" though.
Another possible idea might be to use small magnets and a hall-effect sensor. Same general concept and they make hall effect sensors very, very tiny. The only issue is the gauss rating of the magnets & sensor, and whether or not there could be any influence / interference from the motor or external items, like uncouplers. Probably not, but its at least a consideration for the first one (prototype). Look at Digikey or Jameco for appropriate hall-effect sensors.
John
Thank you John, for the additional ideas. I also thought of an additional "refinement" of my original circuit. While poking around on digi-key I found some bridge rectifiers that are the same size as the photo-relay (10mm square ~3/8" for us metric-challenged folks) If that was wired to the track pickups, that would free up the function output for other uses. They can also be found with very small output currents, so direct driving of the sensor/switch might be possible.
Again, smaller scales and smaller locomotives might not have enough room, but my current project is an S scale 0-4-0T, so I don't have much room to spare.