Engine Motion Detection and Triggering Circuit

Nice job and especially the assembly craftsmanship

If anyone is planning to duplicate this for Rule 17, an asymmetrical time-delay could be added for, say, 50 cents in parts so brief stops don't trip the dimmer.  I noticed some engines with built-in Rule 17 wait a bit after coming to a stop before the headlight dims. But when starting, the headlight immediately turns bright.

and the other set of contacts was used to break the cab light connection so the lights were off when the train is in motion.

I LIKE this idea as not only turning off the cab interior light but also any under cab running lights one may have with those newly developed light chips!

Is it safe to assume it will work in conventional mode as well?

I figured it was close enough to have it go off when you stop, but the delay makes sense.  I did notice one minor issue, but it's not worth me going back, since I don't have a ready solution.

The voltage across the motor to actually trigger this is about 1.6 volts, after the drop across the Schottky bridge and the Emitter in the opto-isolator.  With the Williams GG-1 and no cars, it'll actually start moving before that, and you do have a period of uncertainty where the lights may flicker a couple of times.  This only happens when using the Cruise Commander 100 step mode and starting out in step 1.  Since the opto and bridge need to be totally isolated and self-powered, I didn't have a ready answer for that. Perhaps your delay circuit might be the ticket, what did you have in mind?  I was thinking of some hysteresis on the relay dropout would be the way to address it.

Since with even a couple of cars behind it, this isn't a factor, or if you start it out at anything but the first step of the 100 step mode of the Cruise Commander, I decided for this generation one circuit that I'd leave it as it was.  However, for the next time...

It is slick to be able to add these features easily to a standard TMCC locomotive, now if I could only duplicate the Legacy sound, I'd have it all!

I LIKE this idea as not only turning off the cab interior light but also any under cab running lights one may have with those newly developed light chips!


Is it safe to assume it will work in conventional mode as well?

No reason I can imagine that it won't work in conventional mode, it's strictly triggered from the motor voltage.  The only issue I see is the power supply for the relay, that requires 12 volts.  I think I'd probably use a lower voltage relay, and also add a super-cap to allow the circuit to work as you slow down and stop.  Of course, you need some voltage to light the lights!

It was designed with command mode in mind, that's pretty much all I run.

I think John designed this for TMCC operation.  For example, it appears he used a 12V relay.  However, lower voltage relays (e.g., 3V or 5V) are readily available.

John, I was thinking of the fast-charge, slow-discharge asymmetrical time-constant from a diode-resistor-cap.  To keep the cap small, the R-C would then drive a 5 cent transistor to provide current gain (and slow cap discharge) to drive the relay coil.  There's a chance this would solve your first-step problem.  That is, I believe the motor drive is pulse-modulating the motor at the first step so you might be getting some slim pulses at the opto output.  A fast-charge, slow-discharge circuit would effectively spread out these pulses.  Without the spreading, the average value is apparently marginally tripping the relay.

Post your mods, and I'll give them a go in the next circuit.

It was indeed designed with command in mind.  Conventional operation throws variable voltage (and no voltage at times) into the mix.

The above 25 cent circuit ought to do the trick. You might be able to experiment on your existing circuit with minimal re-work.

The opto output quickly charges the 1uF cap when motor is driven.  The N-FET functions as a switch to drive the relay coil with essentially zero gate current (unlike a conventional NPN bipolar).  The 2N7000 or BS170 are the 2 cheapest switching N-FETs I found at Mouser (about 15 cents).  When the motor turns off, the 1uF slowly discharges through the 1 M resistor with a 1 sec time-constant (1 uF x 1 M).  The zener limits the gate voltage to under 20V which is a typical limit for these low current FETs.

Since the FET turns on/off at a few volts, and you're running at 18V or so, it will take several time-constants (several seconds) to turn off.  You can adjust R or C to adjust time constant as desired.

The opto output transistor acts as the diode function of the fast charge, slow discharge circuit since it does not drain the cap voltage when off.  I looked up your relay and it appears to require a coil current of about 10 mA.  So let's assume your opto output provides 10mA when the motor is on. Since the charge rate of a cap is I/C, a 10mA current charging a 1uF cap ramps up at 10V per millisec (i.e., very quickly) so the relay will trip as soon as the motor starts. 

Just another thought, dont know how practical it is, is to install a diode package between the motor and reverse board,described here for conventional operation use to slow a motor

http://www.jcstudiosinc.com/Bl...=488&categoryId=

the string at each end,clamps about + or - 3 volts,depending on train direction. This can be used for lighting an LED or bulb or activating a common 3VDC relay coil.  Whenever there is voltage to the motor(s) the relay or LED will be on.

Dale H

gunrunnerjohn,

Regarding your circuit, is pin 6 of the Opto left unconnected, dc+, or grounded? 

thanks,

The pins that show no connection really have no connection.

John,  Nice job.  For clarification this is a DC motor?  How does rule 17 work in reverse for real engines, is it the same principle as forward.  G

That's a DC motor, the Schottky bridge rectifier obviously creates a single polarity to trigger the opto-coupler and thus the relay.

As far as Rule-17 for real engines in reverse, I'm not entirely sure.  I would guess that for diesels you'd still bring the rear light up to full intensity when moving, but dimming the steamer front light might not be affected in reverse.  However, it is in my models.

I'm going to assemble one with Stan's suggested mods, they look like they'll be a nice improvement.  I'm also looking at getting a PCB made once I hit on a final configuration.

If I remember correctly Lionel flips the polarity of both leads and you don't have a PV sitting at one of the motor leads. How would you modify this for a MTH PS-2 motor?

For MTH the relay switches the High PV (about 20VDC) to one of the motor leads depending on direction selected.  The other lead then is flipped to the FET to modulate the return to PCB Ground.  That is why the PV for heater and headlight still exist even in neutral.  One of the motor leads and one of the diodes is always powered by PV.

It would be neat to have automated cab light off and on with motion vice operator action, plus you could use a voltage drop to have your rule 17 also.  Would be a nice backfit to MTH engines.  G

No mod necessary for MTH, it works just fine.

Remember, the motor leads are totally isolated, they just feed the opto-coupler through that bridge.  The bridge insures I always get the right polarity, and the motor voltage fires the LED in the opto.

It works just fine in an MTH locomotive.  The only change that I can see ever being possible would be to have a filter on the output of the Schottky bridge to further filter the opto-isolator voltage, but it works fine without it on MTH and Lionel.

Good point Jon, I confess to using parts that I already had on hand.   I think I'll look at that option now.

Since I'm going to whack another one of these together, I'm going to use the AC opto-coupler (thanks Jon), and the FET delay circuit (thanks Stan), and I'll post the results I get with all the mods when I get it done.  My Digikey order is on it's way for the opto-couplers.

John;

No shame there!

I really like your original incarnation, and I think others here have provided some useful refinements.

Design by committee; ya gotta love it.

I for one will be very interested in the final outcome.

Rod

Well, the new prototype is done.  I'm pleased to say that the improvements both made a difference.  The AC opto-isolator lowered the motor detection voltage about 1/2 a volt, which was expected as I eliminated two Schottky diodes in the circuit, that being the diodes in the bridge.  The FET delay circuit works well also, it delays the turn-off about one second and eliminates any chatter, another useful mod.

Here's the prototype and the circuit as it stands now.  I've generated artwork based on the current design, but obviously I don't have PC boards yet.