Best Of Pennsy Position-Light Signal Head - Shapeways Project?

[DKS]

I just re-read DKS' lead reduction proposal and if I understand it correctly he proposes to have single negative lead for all the center LEDs.  While it should be ok to hook up 4 identical LEDs (from the same production lot) to be connected in parallel and share a singe current limiting resistor, I personally don't like doing that.

I would install a separate resistor for each center LED. Those could be hidden either on the target or on the bridge. That way single negative lead can still feed all 4 center LEDs with their own current limiting resistors.  I know that I'm being overly conservative in my design, but that's just how I am.

Peteski, in your diagram, you have the position LEDs connected in series, and the center one not, so you'd need to use a different resistor for the center LED to keep the brightness the same as the rest. Now, with four signals in one group, you can connect a pair of the center LEDs in series, such that you'll have two pair. Then, if you don't mind one additional wire, you can use the same resistor value for everything. Or, you can connect all four of the center LEDs in series, and use a different resistor for them (assuming the supply delivers enough voltage to handle four LEDs in series, which should be about six volts).

[peteski]

Peteski, in your diagram, you have the position LEDs connected in series, and the center one not, so you'd need to use a different resistor for the center LED to keep the brightness the same as the rest. Now, with four signals in one group, you can connect a pair of the center LEDs in series, such that you'll have two pair. Then, if you don't mind one additional wire, you can use the same resistor value for everything. Or, you can connect all four of the center LEDs in series, and use a different resistor for them (assuming the supply delivers enough voltage to handle four LEDs in series, which should be about six volts).

Ah, now I see how you envision this.  That makes sense.  This would mean that there would be 6 (not 5) wires coming from the signal heads which have the chained center LED.  That is not a problem since the extra wire only interconnect between the signal heads (not down to the layout).

BTW, most yellow LEDs have Vf of around 2V, so if 4 were connected in series, that would be closer to 8V plus a couple of volts for the current limiting resistor to do its job.  That is not a problem since many singnal circuits operate from 12V supply.

[Mark5]

I'm gonna have fun with whatever Position Light Targets are available by the time I get around to it - I need Green, Yellow, and Red. 

[nscaleSPF2]

This thread is starting to take on a life of its own.  I like it when that happens.

Before we get too deep into simplifying the wiring, as Peteski pointed out,

I think the best way to wire the LEDs would be in some sort of a jig. Then take the LED/wire assembly and glue the LEDs into the target.

Preliminary testing shows that a light spray of 3M 77 adhesive onto the signal head is enough to hold the LEDs in place during soldering.  Looks like we now need a Shapeways fixture:  one without the mounting bracket and the shades.  Tape the fixture onto the bench, spray it with 77, put the LEDs in place, then solder the wires.  I like it.

Jim Hale

[towl1996]

Yes, the evolution of this project has been remarkable and I congratulate you on your efforts.  Are these going to be available for purchase? If so, I would like to reward you financially for your craftsmanship. 

[eric220]

Eric's in trouble though.

Yes.  Without pre-wired light boards like the ones Bob was promising, I'm in a lot of trouble.  As for hiding the cabling, I've seen plenty of photos of prototype bridges that had multiple conduits running down them to carry the cables.  I'm planning on using brass tubing to mimic the conduits and contain the wires.

[peteski]

To be honest, the best way to do this would be to have printed circuit boards made. As thin as possible, and double-sided. That would dramatically cut down on the labor needed to assemble these.  It would also be nice to have the targets and hoods made as a lost-wax brass castings. But this would really take this venture to a much higher level. Hmm.... didn't someone produce this type of singnal in the past? 

[John]

Hmm.... didn't someone produce this type of singnal in the past? 

alkem ...   and not sure what the status of traincat signals are now ..  this is a good alternative .. wiring led's isnt that hard .. just tedious

[nscaleSPF2]

Yes, the evolution of this project has been remarkable and I congratulate you on your efforts.  Are these going to be available for purchase? If so, I would like to reward you financially for your craftsmanship. 

towl1996, it is not my intention to profit from this design.  Once we get an acceptable print from Shapeways, I will make the model available to anyone who wants to order parts from them.  They cost about $5.00.  I expect that the jig will cost somewhat less.  This is only fair, as the design truly belongs to the forum members, many of whom contributed to it.

To be honest, the best way to do this would be to have printed circuit boards made. As thin as possible, and double-sided. That would dramatically cut down on the labor needed to assemble these.  It would also be nice to have the targets and hoods made as a lost-wax brass castings. But this would really take this venture to a much higher level. Hmm.... didn't someone produce this type of singnal in the past? 

Peteski, if anyone wants to take this project to a higher level, then they are more than welcome.  Somehow, though, I don't think that this is going to happen.  If the costs to produce printed circuit boards, castings, and assemblies were justified, then Alkem would still be producing PRR signals.  Never mind the ridiculously expensive signals from New Jersey Intl.

Jim Hale

[DKS]

Somehow, though, I don't think that this is going to happen.  If the costs to produce printed circuit boards, castings, and assemblies were justified, then Alkem would still be producing PRR signals.

TrainCat has already invested heavily in this project, including a very thin PC board with LEDs attached, and an injection-molded target. The PC boards have already been made; the tooling for the target, to my knowledge, has not been done yet. I had produced an accurate cast resin target as a proof-of-concept in the event TrainCat wanted to save tooling costs, but it was vetoed. Unfortunately that means I cannot sell my castings, as we agreed to avoid direct competition. Worse, owing to multiple dire personal issues, TrainCat is not currently moving forward with the project, but still plans to, leaving it all in limbo. I even offered to buy the entire project from him outright, but he refused.

[Dave V]

I'm so sorry Bob has had to suffer so much over the last year or two and I wish him all the best both in life and in business.  I hope he keeps cranking out the PL masts and signal bridges.  However, respectfully I don't believe we should feel obliged to remain in limbo until whenever Bob may or may not be able to move forward with the project.

I think it's a valid argument to wonder whether we'd see Alkem do more of theirs if they were profitable.  It's worth noting that Alkem sold out of their prewired PL heads.  Bernie can shed more light on what that means.

For my tiny slice of Pennsy, I'm willing to give this a go myself.  Microsoldering LEDs is just another skill I've yet to learn (better to do so now while I have steady hands!).  I'm assuming you solder before you install lest you melt the casting, LOL.

Time to fire up the 150W Weller gun!   

[eric220]

TrainCat is not currently moving forward with the project, but still plans to, leaving it all in limbo. I even offered to buy the entire project from him outright, but he refused.

That is one of the most encouraging things I've read in awhile in regards to TrainCat. I really like the solution that Bob came up with, and I'm eager to see his signals produced. Fortunately for me, I've got plenty to do before I'm ready to install signals, so I can wait. Good luck for the speediest recovery possible for Bob and his family.

[nscaleSPF2]

Hah!  I'm back.  Went thru another couple of design iterations with Shapeways.  Inverted the digital model, thinned the target to the minimum allowable (0.3 mm), and changed the shape of the hoods slightly, to make the lights a little more visible when they are not viewed directly from the front.  The last change will allow an engineer to actually use the signal to control the speed of his train, which is one of the reasons I started this project in the first place.

OK, for better or worse, here are what the latest signal heads look like after cleaning with Bestine, and sprayed with Krylon chrome, followed with Krylon flat black, per Bob Gilmore's suggestion.





Now, to install the LEDs, I made a jig by cutting off the hoods and the mounting bracket from one of the earlier samples, taped the face down, and sprayed the jig with 3M 77 contact adhesive.



The LEDs are wired per Peteski's recommendation in an earlier post.



Soldering these tiny buggers is not the easiest thing to do, but I do need to improve my technique.  The assembly is very carefully pried off the jig with a number 11 blade.  There is probably a better adhesive to use than the 77 for this application.



I then glued the LED assembly into the back of the signal head.



The mast and ladder came from Ben Renard at BeNscale.com.







Will post more photos after final paint and installation on the layout.  One of the better features of Ben's design is a large base into which the assembly is placed.  The base is large enough for a phone cord plug to pass thru, for the electrical connections.

Don't you love it when a plan comes together?

Jim

[DKS]

Don't you love it when a plan comes together?

Indeed! Well done!

[eric220]