Weekend Update 11/11/12

[Dave V]

The old, oversized signals...



This is how it looks now.  The signals are a bit dim; I'm using 1kohm resistors...  I can probably cut that at least in half, as I'm using 560 ohms on the common annode too.  If that sounds excessive, toasting my first one 5 years ago made me a little gun-shy!

[peteski]

This is how it looks now.  The signals are a bit dim; I'm using 1kohm resistors...  I can probably cut that at least in half, as I'm using 560 ohms on the common annode too.  If that sounds excessive, toasting my first one 5 years ago made me a little gun-shy!

What color are the LEDs? yellow or white?  Do you have any specs for them (like forward voltage or operating current)?  How much voltage is dropped at the output of the circuit driving the LEDs? What is the power supply voltage? Once you have that info, you can safely figure out the right value resistor.

For example, if these are white LEDs then they typically have forward voltage of around 3V.  Current of around 5-10 mA (0.005-0.01A) would make them glow quite bright.

If they are yellow LEDs, then they probably have forward voltage of around 2V and operating current of 10-20mA (0.01-0.02A).

I assume that your supply voltage is 12V and the voltage dropped on the output of the driver circuit is 0.2V

To pass 10mA through the White LED, the resistor value will be 12 - 0.2 - 3 / 0.01 = 880 ohms (use 820 or 910 ohm resistor).
To pass 15mA through the Yellow LED, the resistor value will be 12 - 0.2 - 2 / 0.015 = 653.3 ohms (use 620 or 680 ohm resistor).

This is all that is needed to calculate the values (not a rocket science - just grade school math).    Of course, those were only examples as I don't have the actual specs of your circuits.

[Dave V]

Peteski,

I have all those specs; I know I could get away with less than 200 Ohms; just nervous considering that they're out of production.  Will probably switch to 250 or so and see what that gives me.  I don't want them headlight-bright either.

[Philip H]

Dave,
Those do look so much better.  If Bernie ever decides to produce them again, your installations need to be his catalogue shot.

[Hiroe]

Having made the decision to switch scales about 4 weeks ago, I've gotten more done in that time than I had in the previous 5 years in N.

I recently rebuilt my curved-module-side assembly jig, and have started building additional curved modules.

Here's the laminate jig:


One of the first ones out of the jig:


After a coat of Spar Urethane, and starting to lay track:


The grey ones next to it are some of the ones I built for N scale, 8 years ago, with new track installed on them.

These curved modules are simple and easy enough that it takes me approximately two evenings worth of work to go from raw materials to a finished module frame. Laying track and installing wiring is a third evening worth. I have yet to build the jig for assembling the catenary wire for the curves; but I estimate that once the jig exists, I could probably crank out the wirework in a single evening.

The possibility of having a full circle of these running for the EPTC meet in May just got a lot more likely.

--Drew

[peteski]

Peteski,

I have all those specs; I know I could get away with less than 200 Ohms; just nervous considering that they're out of production.  Will probably switch to 250 or so and see what that gives me.  I don't want them headlight-bright either.

Dave,
no wonder that you keep blowing up your LEDs!   
You have the specs and the simple 3rd grade math to calculate the proper values and yet, you just want to try a 200 ohm resistor (which most likely is too low of a value if your power supply is 12V). 

Or is the circuit powered up using a 5V power supply (which would most likely make it ok to use 200 ohm resistor) .  If the simple calculation I showed earlier shows that a 200 ohm resistor will not allow more than 20mA of current through the LED then you're fine.  I'm just guessing here since I don't know the details of your circuit. Nothing to be afraid of. 

You can estimate brightness pretty well using the resistor calculation. Again, I would have to see the LED specs. White LEDs will give you plenty of light with low currents. Yellow LEDs are not quit as efficient and need more current.  Again, without knowing the specs, I can't make any suggestions.

[Dave V]

The photo probably doesn't show it well, but the signals are now much brighter. 330 Ohms on the common anode and 270 on the leads... Looks just right for my eye.  Not headlight-bright, but clearly visible to the naked eye even from a slight angle.



Nothing says Pennsy quite so much as an N5c cabin car with Trainphone antennas braced by a pair of position light signals...

[Dave V]

Dave,
no wonder that you keep blowing up your LEDs!   
You have the specs and the simple 3rd grade math to calculate the proper values and yet, you just want to try a 200 ohm resistor (which most likely is too low of a value if your power supply is 12V). 

Or is the circuit powered up using a 5V power supply (which would most likely make it ok to use 200 ohm resistor) .  If the simple calculation I showed earlier shows that a 200 ohm resistor will not allow more than 20mA of current through the LED then you're fine.  I'm just guessing here since I don't know the details of your circuit. Nothing to be afraid of. 

You can estimate brightness pretty well using the resistor calculation. Again, I would have to see the LED specs. White LEDs will give you plenty of light with low currents. Yellow LEDs are not quit as efficient and need more current.  Again, without knowing the specs, I can't make any suggestions.

I'm certainly not afraid of math...   

What I didn't want to get into was that I got what I misinterpreted as conflicting information from the manufacturer of the logic board, the manufacturer of the signal head, and my own math.  Perhaps I was too lazy to really recheck all of the math, but based on the info I had I went with the most conservative of the three and worked my way toward a safe brightness.

They are yellow LEDs.  2.5V max, but I have them attached, through a Digitrax DCC system, to Logic Rail Tech signal animators that should output 5V.  The LED specs are thus:  http://alkemscalemodels.net/ASMMain/SignalLights.html

So, using math I get 5 - 0.2 - 2.5 / 0.015 = 153 Ohms.  That's IF I believe I'm really getting only +5V out of the signal animator.  But again, I don't want headlight brightness.  Then add Bernie's diagram with 1 kOhm, but then he's saying +5-12V input, with me using +5, so on the low side of that.  Using a simpler formula I got 125 Ohms which is what the gentleman at Logic Rail Tech indicated.  I had misread an e-mail from him suggesting 400-500 Ohms; I missed that he was talking 12V DC ground versus the 5V output on the animator.

I appreciate your help.  I wasn't ready to respond with specs earlier because I was at work on my iPhone.  That said, I'm satisfied with the current brightness.  I don't think I need any further help at the moment.

[Sokramiketes]

Have you thought about moving the signals before or after the crossover instead of the middle?

[Dave V]

Have you thought about moving the signals before or after the crossover instead of the middle?

Considered it but used existing holes for these.  That's how I did the other interlocking.

[Cajonpassfan]

Actually, since these are trailing point crossovers, the signal locations are almost right; just need to make sure to stop short of the fouling points...
Looking good, and nice animation, regardless of the math
Regards, Otto K.

[nkalanaga]

330 plus 270 would give you 600 ohms, which should be plenty for most LEDs.  Plus, since the 330 is on the common anode, the current through that has to be divided among the LEDs on at any given time, so I suspect you have closer to 900 or 1K in practice. 

[Dave V]

330 plus 270 would give you 600 ohms, which should be plenty for most LEDs.  Plus, since the 330 is on the common anode, the current through that has to be divided among the LEDs on at any given time, so I suspect you have closer to 900 or 1K in practice.

Right, and I have another, more pressing problem with the signals right now, to wit:

Confession time...!  I had one of those bolt-upright-in-bed moments last night when I realized on all 3 signals I've installed so far, I wired the "restricting" aspect into the "approach" connector for the signal animators.  In other words, where most railroads have "yellow," the Pennsy had a diagonal row of lights.  In Pennsy-land, the diagonal for "approach" goes from bottom left to top right.  I wired mine all going top right to bottom left, or "restricting," a relatively rare indication that most Pennsy signals were not even equipped to display (but Bernies' can).

Sounds like an easy fix, right?  For one of them it is.  I can swap the one lead for the other and done.  The other two signals have no working lead #6 (I mistakenly was using lead #4 as approach).  For the first signal I built, I believe that was the lead I toasted testing the head with a DC source inadvertently switched to pulse power (D'OH!).  The other signal, the lead came unsoldered from the back of the signal head during assembly.  For that latter signal, it's since been completely assembled and painted with the lead I thought I didn't need pretty much separated from the head and the soldering point obscured by paint and other details.  Well, can't I just rotate the LED board on the signal head casting so stop is now clear?  Yeah, if I hadn't used enough medium-viscosity ACC glue to weld an anvil to the ceiling...   

Ah, well...  Now, if it bothers me long term, for those two signals I could have the signal driver skip "restricting" by splitting the clear or stop leads into the "approach" connector.  So then, the signal drops after the loco passes and stays dropped for 20 seconds after the last car and then jumps back to clear.  Or I can enjoy some really cool animated signals and not think too hard about it.

Either way, Bernie, if you're following along and you ever decide to re-release these, I'm in for a few more.  Hell, I'd stockpile a bunch for the future basement monster.

[unittrain]

Awesome signals! Since Traincat probably wont be producing his targets I will have to resort to my original plan I got small LEDs from Ngineering that I'll have to get applied to a board. Hopefully there will be someone who can do this.

[peteski]

Dave,
if you manage to separate the circuit board from the head casting I have the skill and equipment to repair/replace/resolder any of the LEDs and wires.  Since these signals are so rare, I can try to fix them for you.

Looking at the link you posted earlier, the head casting is metal and the circuit board is standard glass-epoxy material. The LEDs are also made from a similar. Since you will be disassembling the signal anyway, take it apart and then soak the head assembly in acetone overnight.  That should soften the ACC glue. Then clean the circuit board as much as you can and send it so me. I'll fix whatever needs to be fixed. The replacement LEDs are still available (but discontinued) at Digikey for about 10 cents each (or less in quantity).

I have several projects on the plate right now but I woudl be able to get to this sometime in the early 2013.