New brass Milwaukee Bipolar EP-2 clinic- part 2

[victor miranda]

I have not purchased them.
they are new to me since I last purchased leds.
I would assume they are bluish because of how hard it is to get warm white.
... all the tiny LED I have met are blueish.

have you any tamiya clear orange paint?
this stuff (http://www.amazon.com/Tamiya-Models-Acrylic-Paint-Orange/dp/B00KMWUWFY)

that has been my fix for these lamps.

well, I mix some yellow and orange to get a thicker layer of paint.

[u18b]

I swear.
You guys are going to be the death of me.




The next time I buy life insurance, I will have to watch out for the question that says: 

In the last 5 years, have you engaged in any of the following: bungie jumping, cliff diving, asshatery, deep sea diving, etc...

I have no idea if this thing works.
But just to show you the scale, here it is with a MT coupler spring from a 1015.



This is at the very limits of my skills and sanity.

I will still have to find a good high resistor so as not to damage this-- if it even works.

[victor miranda]

!!!
wow

how it looks when lit is the magic test.

I have to share that the more I solder to these tiny leds the better I get.
(and I am expecting to lose and toast a lot as punishment for that bragging.)

I mention it because you seem to be doing some amazing work there.

my eyes hurt from thinking about it.

I am trying to see how your leds are powered.

when I was thinking about how to set them into place of a PCB

I thought up a target. all the LED pointed in and the ring on the outside
for the tails (cathodes and anodes)

yours I can tell.

and I deny all accusations of attempting to do you in.

I would have just done the one LED and not done the 3 hole mask.
... we can't all play first fiddle...

victor

[peteski]

Ron, you have graduated into the ranks of the microscopic LED wranglers!    Awesome!

But I still cringe at hooking them all in parallel. Not the best practice. At least they are the same exact part (hopefully from the same production batch).   As you mentioned, brightness is the least of your worries as you can infinitely control that with the current limiting resistor.  Bit of rsan at a very low current you might see the effect of your parallel hookup (the brightness might be uneven between the LEDs) and the light emitted will be on the bluish side.

As far as the Chinese LEDs on eBay go, the white color might not be to your likng (not warm enough).  The N-Gineering LEDs can be had much less expensive. Just contact Tim (the owner) and ask for a quantity discount.  Something like at least 50 or a 100 of each type.  The price is much more reasonable, and those are some of the best 0603 and 0402 size warm white LEDs available anywhere.

[u18b]

Here is how they are wired.

I made two cuts on the tiny circuit board (black dashed lines.)

Polarity is indicated.

Brown wire is the positive (goes to blue on decoder).  Soldered in two spots as indicated.

Green wire goes to negative and white or yellow.  Soldered only at the end (top).




At this moment, I have no idea if I have solder bridges and thus a short. 

So I don't want to test this with my decoder and burn it out..

Later tonight, I will probably hook this up to a 10 k variable resistor (pot) and hook it up to my analog track.

Make sure I am at 10k.
Turn  on the track power, and gradually lower the pot until I see life.

First goal is to see if all three even come on.

If they do, then I can work with my decoder.

[u18b]

Pete.

How else would you hook this up?

If it was in series, then each one would be dependent on the other (like those nasty old christmas lights).

Also, wouldn't the 3rd in the series have a voltage drop of the other two and thus not be as bright?

I know you know much more about this.
I must be missing something.

[victor miranda]

Hi U18b,

parallel LED do not have current stability.
if one of the three decides to use more current
say due to heat the other two will not follow
so one light will get brighter.

the usual fixes are a resistor for each led or series.

for the most part I have not worried on this issue
keep your current as low as possible to avoid a thermal runaway.

victor

[peteski]

I guess you never looked at the best practices for powering LEDs.    LEDs are not like incandescent bulbs (they have a very non-linear voltage vs. current curve).

Every LED die is unique and has its own current/voltage curve.  If the LEDs are from the same manufacturing batch they are all matched fairly closely, but even the same LED from a different batches can have electrical properties different enough that if hooked up in parallel, they will not all light up evenly (as some LEDs will have lower forward voltage which will make them conduct more current than their parallel-connected peers).Each LED should ideally have its own current limiting resistor.    If you have multiple LEDs to be illuminated together the connecting them in series using only one current limiting resistor will assure that the same current passes through all of them (which means same brightness).  LEDs are reliable enough that the chance of one going out killing the entire circuit is very remote.

What you are doing is not going to blow up anything in most cases, but it is not really recommended.  I guess by having electronic background I'm more anal about this than most other hobbyists.  But it doesn't hurt to do it right either.

Without me going into lengthy details, here is a good tutorial I found by Googling: http://www.quickar.com/ledbasics.htm

[u18b]

OK.

Well, in this particular application, the goal is to keep the voltage pretty low-- so that individual lights actually show up.

So, like y'all said, it should not really be a problem.

[u18b]

Well, your praise was a little premature.

There was a moment when it worked.
Not sure what happened.

So tonight, I made another one.
Took me a little over an hour and wasted 2 LEDs.

Put a big old 10k resistor on it.

And it worked.   Light was a little dim, but it worked.
Then I moved the wires trying to position it for a photo- and gone.

Short?  Don't know.  But that was two in a row and a lot of time.

Soooooo........



After using between 6 to 8 hours working on the headlights, I'm ready to move on.
Not all was wasted.

I'm going with two LEDs behind a blackened disc.
That was acceptable.

Thanks for everyone's input.
It was worth trying.

[peteski]

Oh well, it was a fun experiment either way.

I have been thinking about the dual LED. Is it really needed?  If you put a light diffuser (disk of .005" white styrene) behind the 3-hole disk and then just place a single 0603 LED in there, that would give you enough light.  You might need to run the LED brighter than if it was without the diffuser, but with white LEDs being so efficient I think it would be plenty bright without even getting to its maximum operating current (which is probably 20mA).  I think that 8 or 9 mA would make it bright enough, even with the diffuser.

Making the top disk (with 3 holes) out of 0.010" black styrene would save you from painting it black - it would not pass any light through the plastic.

[u18b]

Oh well, it was a fun experiment either way.

I have been thinking about the dual LED. Is it really needed?  If you put a light diffuser (disk of .005" white styrene) behind the 3-hole disk and then just place a single 0603 LED in there, that would give you enough light.  You might need to run the LED brighter than if it was without the diffuser, but with white LEDs being so efficient I think it would be plenty bright without even getting to its maximum operating current (which is probably 20mA).  I think that 8 or 9 mA would make it bright enough, even with the diffuser.

Making the top disk (with 3 holes) out of 0.010" black styrene would save you from painting it black - it would not pass any light through the plastic.

Yes.  I don't regret trying.

As to doubling the LED.  Yes, I think it is better.

The problem is the holes are spread out across the headlight.  And the one LED shines broughter when a hole is directly over it.

Doubling the LED increases the surface area of the light.  So that more holes line up.

The net effect is a more even light source coming through the diffuser/disc.

[peteski]

Doubling the LED increases the surface area of the light.  So that more holes line up.

The net effect is a more even light source coming through the diffuser/disc.

I'm not sure you understand my idea. I don't think you are currently using any diffusers - just an opaque disk with 3 holes.

I think that a 0.005" white styrene disk acting as a diffuser (placed behind the disk with 3 holes) with a single white LED placed roughly in the center behind the white diffuser and running with around 10mA through it would give you the brightness and relatively even illumination of the 3 simulated headlights.

[u18b]

Oops.  You are correct.   I had a brain freeze. 

I may try that. 

I'm finished with the lights for now. 
They will actually be the last step. 

I need to experiment now since the 3 and 4 light versions are different from the older single lights.

I'm actually about finished constructing E-4.
The only thing left is to blacken the wheels.
Then air erase and paint.

I'm thinking about going with Tru-Color. 

Anybody use their Milwaukee colors?

While I'm waiting on the paint to come in,
I'm going to get back to work on E-5.  It is identical to E-4.

My woodworking friend is traveling.  When he gets back we will work on the wooden boxes.

[peteski]

When properly applied, you can't beat TruColor paints!  The finish will be smooth (no orange peel), and their semigloss finish is ready for decals without any additional clear coats.  But if applied too dry (too much air pressure and/or too far from the model), they (as with any paint) will develop an orange-peel coat.

When I airbrush them, I don't go over 20 psi and I spray a light coat first, then in the same session, a wet coat or two.  The wet coats are the key to smooth finish.  I usually spray about 6" away from the model or sometimes even less when getting into hart to reach areas (then I crank the pressure down).    That way the paint spray has no time to partially dry on the way to the model. The only times I go over 20 psi is when I need much wider coverage. Then I also increase the paint flow.  I use Badger model 200 airbrush.

I would use  a Tamiya light gray primer on brass. Or even their white primer if you want the colors to pop (assuming that it will be maroon and orange).

I don't know the accuracy of colors though. However with all the real-world variations, unless you are trying to match some specific loco in some specific time period, from some old photo, I wouldn't worry about the accuracy.  Plus will you try to match the 1:1 colors or use scale (lighter) version of the colors?