Guys. Seriously, Guys. Fleischmann has impressed the hell out of me again

[nscaler711]

Oh I gotcha... Ok that is pretty sweet. I thought you all were goo goo for what was just simply directional lighting.... I was like ok these guys have officially gone nuts. 
But I'll definitely agree that's pretty sweet....
MAGIC I TELL YOU!

[Lemosteam]

I just knew there had to be a mechanical element to the thing!  Cool, thanks!

[mmagliaro]

Mechanical, simple, and elegant.  No electronic sensors.  Genius, I tell you!
So it looks like there are pickups for the electricity on all the wheels, and then that bronze
axle/finger thing is additional just to provide the directional switching.  I suppose the only
thing that would happen if that finger thing got dirty is that the direction wouldn't switch,
but as long as the other pickups are functioning, you would still have lights.  They would just be pointed
the wrong way.

The thing is, that finger contact doesn't even have to be all that perfect.  It just has to touch for an instant
to trigger a switch in direction.  It "latches" on in that direction via circuitry on the circuit board.

This is really smart mechanical design. 

[peteski]

The only thing missing as to when compared to a DCC function-only decoder is that the lights cannot be turned off (unless you can precisely position the car in a neutral setting of the finger/wiper).  But other than that it is ingenious: works on DC or DCC and it is totally independent of the rest of the train or the locomotive.

Mike, are all 4 wheels picking up power, or only one wheel per axle?  Most Fleischmann wheelsets I have encountered have only 1 wheel per axle insulated.

[mmagliaro]

 I would think it impossible to turn the lights off no matter how you position the car.
That circuit is probably going to turn on and latch one way or the other and the only way to
switch it is to change direction.
It is kind of like a SPDT switch with no "center off" position.

Which begs another question.  Maybe I missed this, but does the car remember its state?
If it is lighted in one direction, and you turn off the power, does it reliably come back on with the lights
in that same direction when you power it on again?

[peteski]

Max, Mike's video clearly showed the lights in off state (Mark even described that there is a small dead spot when the car changes direction).  It doesn't look like it latches the last direction (other than mechanically).

[jagged ben]

Ok ... but we still have no explanation of how it actually works electrically.  And I for one would like something of an explanation.

 

Someone said no electronic sensors but there is still an electronic component here that is more complicated than a mechanical switching of polarity.  If I understand Mike's diagram and video correctly, only one of the little contacts touches an axle for either given direction.   And yes, if you get it just in between where neither contact is touching, the lights turn off.

So... Each contact completes a circuit that tells the lights to do the opposite thing.    How does it complete a circuit that can tell anything to do something?   I think it must be that that each axle has one insulated wheel like Peteski said, and the insulated wheels must be on opposite sides.   Track power is connected through the uninsulated wheel, the sensor contact, and back through to the opposite rail via the circuit board and whatever electronic component(s) control the LEDs.

So far so good, but how does the rest of the circuit work?  (Questions: if you deliberately complete both circuits, do both sets of lights turn on?  Same answer with DCC and DC, or different?  I suspect different, but this could be a big clue. )    There are three LED circuits, and Mike thinks he sees the resistors for those.  At first I thought the rest could just be diodes but I'm having a hard time with that idea unless there's a full-bridge rectifier on the bottom of that circuit board.   (The red and white would only require two diodes but the destination sign would require 4 all by itself, at first glance.)

I'm thinking it could be more clever.   I see a set of two components next to what I think is the the LED for the destination sign.  What if those are two transistors?   I haven't tried to draw out the circuit, but what if each sensor circuit triggers a transistor that closes the circuits for the right LEDs? 

[peteski]

As I mentioned, I would love to reverse-engineer this myself but I'm not willing to dish a hundred bucks just to do that.  Since this circuit works in DC and DCC, there might be a  bridge rectifier in there to provide power which does not change polarity with the track voltage.

EDIT
I looked again at the circuit board in the video.  One the front edge I see 4 LEDs (2 white and 2 red) for the lower headlights/EOT markers.  Then on the opposite side I see 2 yellow LEDs (for the destination board) and a single white LED for the upper headlight. Then in the center I see 3 resistors. That's it. But we don't know if there are any components on the bottom.


I am curious as to how it behaves in DC. I'm beginning to suspect that this circuit might be simpler than we think.  It might be all mechanical switching.

[mmagliaro]

Max, Mike's video clearly showed the lights in off state (Mark even described that there is a small dead spot when the car changes direction).  It doesn't look like it latches the last direction (other than mechanically).

Ah yes... so it does.   There just has to be a little gate or flip-flop circuit in that board that can latch the
state.  And yes, as you reminded me, it does indeed have an "off" state.   That little circuit alone would be a really handy
thing for somebody like Ngineering to add to their line of handy line of tiny module boards.

[peteski]

That little circuit alone would be a really handy
thing for somebody like Ngineering to add to their line of handy line of tiny module boards.

Unfortunately I think that the real "brains" if this circuit are the special contacts in the truck.  That might be more difficult to produce and adapt to various models.

[delamaize]

So to clear some things up, the power to light the lights is from both trucks. they are a silver colored wiper, that sits on the inside off all 8 wheels. The Bronze connections are ONLY for switching the direction of the lights. I feel like the track current travels thought those, and they are what is responsible for switching the direction, like a single pole, double throw switch....

[peteski]

So, it looks like the power is picked up from all 8 wheels.  But in the video each truck's underside looks different. So while power is provided by 8 wheels, the direction changing wipers are only on the truck directly under the the cab?

Can you see under the LED circuit board?  Are there any additional components soldered on that side?

Is there a way you can test this car on DC (manually pushing it would be fine).

That sure is an interesting circuit.

[jagged ben]

I looked again at the circuit board in the video.  One the front edge I see 4 LEDs (2 white and 2 red) for the lower headlights/EOT markers.  Then on the opposite side I see 2 yellow LEDs (for the destination board) and a single white LED for the upper headlight. Then in the center I see 3 resistors. That's it. But we don't know if there are any components on the bottom.

...
I am curious as to how it behaves in DC. I'm beginning to suspect that this circuit might be simpler than we think.  It might be all mechanical switching.

In order for this to be just mechanical, by which I mean no diodes besides the LEDs themselves, the destination sign must have two separate LEDs, one for each direction    But that's exactly what you're pointing out seems to be there, so you may well just be right.    I sketched a diagram for myself assuming one LED for the destination sign, but then I needed four diodes to prevent shorts.  (Actually one could forgo two of them if one was certain that the brass contacts would never touch at the same time, but that'd be pretty negligent engineering.)   But if there are two LEDs for the destination sign, each wired in parallel with the other LEDs for a given direction, then you can have the circuits for each direction completely isolated from each other and it would be exceedingly simple, aside from having a lot more LEDs than your typical American lightboard.   It'd be just like any old DC directional lighting except that the switching contacts on the axles account for DCC as well.

If this is true, and if one somehow forced both contacts to close at the same time and applied power, you'd get both sets of lights on in DCC, but in DC it would act like a conventional lightboard.   

I do wonder if there might be a couple capacitors on the bottom of the board to reduce flicker.  Otherwise I'd  worry more about the longevity of the contacts.   It doesn't hurt to have the 'return' pickup on all four axles, but your still relying on one switch contact to the axle, per direction, to complete the circuit.

[jagged ben]

So to clear some things up, the power to light the lights is from both trucks. they are a silver colored wiper, that sits on the inside off all 8 wheels. The Bronze connections are ONLY for switching the direction of the lights. I feel like the track current travels thought those, and they are what is responsible for switching the direction, like a single pole, double throw switch....

As I mentioned in the previous post, if the wipers on the wheels connect to the frame, you simply have full pickup for the 'return' of the circuit.  But the bronze contacts could still carry the actual current powering the LEDs, rather than throwing an electronic switch.  It certainly doesn't hurt to have more pickup on one side of a circuit than the other. 

[Erik aka Ngineer]

Thanks Mike! I run some Fleischmann stuff but I never expected the solution to be this clever yet simple.