Best Of New Brass EP-2 assembly & Mod clinic

[u18b]

Time to work on installing the socket in the second vestibule.

I decided to make a change this time.  Since the black wire is soldered to the shell..... why not solder the socket right to the shell-- no wire.

I marked and drilled the hole.  Made it a little closer to the wall this time and a bit smaller.



Hard to see in this photo, but I scratched the metal to get a fresh spot---- right side of the hole.



Fluxed and added solder to the spot.



I added the red wire to a socket with shrinkwrap as before.  The pin for the black wire is tinned and then bent over to the solder spot.



I then touch the soldering iron to the pin and the solder spot.  Space is tight and it is tricky.  A little excess solder got on the shrink wrap but it did not burn through and there is no short.  Socket is much more secure now.



I soldered a pin to the red wire and ran it to the motor as before.



Works.  I now have a jumper between both Cabs.  Of course the key is to plug the wire in correctly AND to make sure the same wire is on top.  In this case, the red wire.



This now dramatically increases electrical pick up for the locomotive.

Before, each motor in a cab picked up juice from 4 wheels on one rail and 3 wheels on the other rail.  Now each motor picks up power from 8 wheels and 6 wheels respectively.  The original 1983 version EP-2 appeared to be similar but the jumper was only for the front truck (the red wire in my scheme).  The chassis drawbar itself was the other conduit for juice.  And that worked as long as you had bare metal touching bare metal.  A painted loco killed the whole scheme.  This is better.



So to assemble the whole loco, plug the jumper into one Cab.  Then slide the wire through the boiler room.  Hook the boiler room to the Cab.



Plug the jumper into the second Cab and hook up the boiler room.



Next step in the modular wiring will be the operating headlights.

[spookshow]

My apologies for my denousity, but did you wire these up for two decoders or just one (or either/or)?

Thanks,
-Mark

[u18b]

My apologies for my denousity, but did you wire these up for two decoders or just one (or either/or)?

Thanks,
-Mark

Is that a typo?  Hmmm.  I never heard of that word.

Anyway,  I'm going to do this in stages.

1st step is analog with lighting.

2nd step will be 2 decoders.

and then I might try to add the extra wiring harness for 1 decoder.

It is supposed to SNOW in Florida today and tomorrow.  Yikes! 

LOL.

Hope to get a lot of modeling done in the next days.

More coming!

[Kisatchie]

It is supposed to SNOW in Florida today and tomorrow.  Yikes! 

I'm in east north east Louisiana, and we have 4 inches of snow on the ground. I just want to stay inside, but my dog loves the snow!


Hmm... my tree house could
use some more insulation...

[spookshow]

Whoops, that was supposed to be "densousity" (as in "being dense"). Spelling it wrong only proves that I have it   

Cheers,
-Mark

[u18b]

Time to move toward decoders and lights.

First step is to use a motor tool and carefully remove any inner projections from detail parts soldered to shell.  We need all the room we can get.



I took some thin styrene and made a rough template for what would fit in the nose.




Same thin PC board we used earlier.  I've scored a line about 11 mm from the edge.  That will be the height.



The width will be about 14mm.



Now round the corners for the roof.









Test fitting.  Needs a little more rounding on the corners.



Board fits.  Now removed the metal foil away from the edge for insulation purposes.  As before.

[u18b]

There were two tiny decoders that were possibilities for this project. 
I kind of lean toward TCS decoders and like the Z-2.

But in this situation, the Digitrax DZ125 is the winner because it is tiny and has solder pads that are easily reachable (replacing wires on the Z-2 is harder).
Here I've taken the shrink wrap off. (I suppose I've just voided the warranty  )



There are some VERY tiny tipped soldering irons out there.  I just take a file and sharpen a tip.



I've tried orienting the decoder different ways, but I think the wiring harness will work best going up instead of sideways.



I have VERY carefully unsoldered the stock Digitrax wires.  But WAIT!  Danger.  Digitrax uses a good deal of solder (making for a good joint) but I have created some solder bridges where solder touches more than one pad.



Adding just a tiny amount of flux and touching it with a clean solder tip usually worked to get the solder where it needed to go.



However, one solder bridge was persistent!  I could not get the solder to separate.
I eventually got success by pressing a #11 Xacto knife blade in the middle and THEN touching the soldering iron.



Success!  Board is ready for new wiring harness.
I have photos and another board for reference as to how the wires go back.
But for now, note the #1 on the board which is a reference.

[u18b]

Just double checking.
Decoder fits nicely in the nose.  Easier to tell without the wires.



Here I'm just laying everything out.
Decoder is on the circuit board.
Blue wire goes to the slightly lower pad on the left.
The white wire will go to one of the last pads on the right.
I plan to wrap these LED wires around and solder in the back.



I made marks for the slots for the wires.
Also, I marked the positive side.
The way most decoders work, the blue wire is the common positive pole.
The function wires go to ground when activated.
So the blue wire is +.



After the notches are cut, I flip the board over and plan.  I need three sections.  One for the blue wire.
But I need two sections for the function wire because a resistor is needed for the LED.
I mark the surface with a Sharpie.



A motor tool with a very fine bit, working VERY carefully can cut the metal layer but not through the fiberglass.
I've made my three sections.



Now the coating is removed by gently rubbing on superfine sand paper.


Board is ready.

[u18b]

Now, this loco looks sort of like a push-me-pull-you.
Which side is front..... since the white function wire goes to the front and the yellow goes to the rear.

Well, on the old NJ EP-2, both Cabs were identical.  So just pick an end for front.

But not on this new model.
Looking at the top of the hoods, you can see that one end is relatively clean, and one is busy.
The clean end is front (or the A end), and the busy end is rear (or the B end).


Furthermore, the B end has some extra extended sheet metal (in red) which the A end does not.



Ok.  Here I have superglued the decoder to the fiberglass side of the board.
The blue wire is installed and run to the back.
Since this will be the front decoder, I'm using a white wire on the right.

[u18b]

Here is the back side with the blue and white wires installed.
I will probably go back later and file those solder joints a little flatter.  Space is at a premium in this project.


Time to add the wiring harness plug to the decoder.

I'm allowing a plug length of about 37 mm.  That might be too long, but we'll see.
Each tip is fluxed and tinned.  The exposed wire is cut very short for soldering to the decoder.



I soldered the black, red, and gray wires.
Yikes!  Another solder bridge blob.



I used the Xacto knife method described above to separate the pads.
For good measure, I soldered the orange wire a bit to the outside.



Shazaam!  We have a modular decoder.



Here is my chassis wired for analog for initial testing.
Those loops jumper (join) the red and orange.... and the black and gray wires.
I now pull those jumpers.



The plug is lined up color-coded to its socket.



The plug is so small, I usually use a fine screwdriver to push it in.



Yay.  We have DCC.
(but does it work... or did I screw it up with all my hotwiring?
To the layout.



It works!  Whew! 

[u18b]

For those of you possibly asking..... do you really need that circuit board.

Well, yes, primarily as a support for the PC board resistor.

But as I have told you before, we are dealing with millimeters here.  Space is VERY tight.

In this project, small is good.

So I'm going back to the drawing board, and I'm doing away with the circuit board.

I've made good progress. but it will take a while to write and post.

Now, I could have waited until the loco is about finished before posting anything.
But I decided at the beginning that I would post stuff exactly as it happens, so that folks could see:
experimentation
errors
discovery
etc.

And besides, after a few hours, the posts are locked in and I can't delete or modify them any more.   

[u18b]

Well, in order to complete a thought.....

Here is the PC board.  Sections are divided and a flat board resisitor has been soldered n place.



So I then went and removed that board to go in a different direction.
but it was not a waste of time since that board would be what I would use if I were going with Analog.

Time to test out some LEDs.
I bought these from TCS.  They come in an anti static box.
ngineering.com sells smaller LEDs, but these should be just right.



The instructions are helpful.. The squarish side is the positive (or blue wire) and the rounded one is the function wire.



For wiring of the LED, I used magnet wire.  I bought this double spool from Ngineering.com.



I stretched out some lengths.  Flux the very tip and touch into a solder bob for only a fraction of a second.
The wire is coated, but the coating melts with heat.
You can see in this ultra-close-up that the paint covering is peeling from the green wire a little too far back.



I then solder to the LED while it is still on its sprue.
Red wire goes to positive, green to the negative.



How did I know which value resistor to use?

I guessed (experimented).

TCS gives a chart of resistors for the LEDs.  But they are at 3.2 volts while most function wire are at 5 volts.

I didn't feel like doing all the math to go from 3.2 volts to 5 volts.  So I just bought varying sizes of resistors and tried to pick one which left the LED adequately bright but not excessively so.



So I tried 4.7 k ohms.  Light seemed good.



One crucial test was to hold the LED with my fingers.  If it gets hot, then I know to use resistors
 with higher values.

[peteski]

How did I know which value resistor to use?

I guessed (experimented).

TCS gives a chart of resistors for the LEDs.  But they are at 3.2 volts while most function wire are at 5 volts.

I didn't feel like doing all the math to go from 3.2 volts to 5 volts.  So I just bought varying sizes of resistors and tried to pick one which left the LED adequately bright but not excessively so.


So I tried 4.7 k ohms.  Light seemed good.

One crucial test was to hold the LED with my fingers.  If it gets hot, then I know to use resistors
 with higher values.

Um, if the tiny LED was to feel hot in your fingers, the damage to it might already be done. 
I'm also not sure where you got the "5 volt function" thing.  Most small DCC decoders will have the function voltage equal to the track voltage minus the 1.4 volts across the rectifier diodes,  and another fraction dropped across the function transistor.  Let's say, track voltage minus 2 volts.  That will be closer to 10 than 5 volts (assuming the track voltage at 12V).

The math involved in the calculations is grade-school level (basic 4 math operations).  It really isn't that hard. 
However from experience, I can say that a 4.7k ohm resistor will be quite safe to use with that LED with the voltage common from the DCC decoder.

Here is a quick calculation for you.  Voltage from the decoder's function output (10V)  minus Vf of the LED (3.2V) is 6.8V. That is the voltage which will be across the resistor.  Co calculate the current flowing through the resistor you divide the voltage across it (6.8V) by the resistance (4700). That is 0.0014A (or 1.4mA). Since the resistor and LED are connected in series, the same current goes through all the series-connected components.  Thus, the current through the LED will also be 1.4mA.  That is very safe (LED can usually pass 10-20mA without any problems).

[u18b]

Here is the now-wired LED on its little circuit board.  These with the board, I find, are a whole lot easier to handle than plain LEDs.



On the back side, you can easily (with an optivisor, LOL!) see the square pad and the round pad. 
Square pad is the positive so red wire goes there.



The problem is these LEDs on their boards are convenient, but they are a tad too big for the headlight housing.
So I take the LED and file the corners down to make its footprint smaller.



You can see that it is now significantly smaller, and there is still plenty of solder pad left.



Front side.


Test fitting inside the headlight case.  It now fits.



Hard to tell in this photo, but I "painted" superglue all over it for a layer of insulation.
I painted the LED and a bit of the wires.



Now a little micro- heat shrink tubing.



I'm ready to go back to the redesign of the decoder.

I unsoldered all my wires except white.



Here is the new trick.  The reason for using the old circuit board was having a place for the resistor.
So I tried supergluing the resistor to the side of the decoder.
The funny thing is... it is the same thickness as the board.  So it will be unobtrusive.



I took off the white wire and soldered it downward (the opposite direction).  I then ran it down to the bottom of the LED.



Here is another view.



Setting it in front of the motor looks good.



I then trimmed my plug wires with no excess.



Side view.


And now bent down into position.


By pulling it off of the floor, I leave room for the base of the shell, and the retaining screw.


I added Kapton tape to insulate the decoder.
I did leave one pad (blue) and the top of the resistor (white) open for their respective wires.


After painting just a little bit of superglue inside the light housing and letting it dry for some extra insulation over the brass, I installed the wires and LED.



Starting to look good.
It just fits.



I lay the shell next to the body and measured the wire length over to the decoder and trimmed.



Here, the magnet wires for the LED have been installed.
Ready to place the shell on and test it.



Houston, we have half an EP-2!



Now I have to repeat the whole process for the B end.

Wow, aside from the more accurate details on these new EP-2s, these headlights are one of the things that make this model stand out from the original NJ EP-2.

[Chris333]

Dang electronics are just as involved as building the locomotive.