New brass Milwaukee Bipolar EP-2 clinic- part 2

[basementcalling]

You, sir, are simply a craftsman, and an artist of the first degree.

[u18b]

Gotcha Mark.   Until I see another one, I'm thinking it is a one-of-a-kind.

Peter,
Thank you for your kind words.

The work continues on this project.   I have one headed toward painting- though I'm going to have to make my own decals.

And I've started on the next one.   I'm getting better (as you would hope) with each one.  This past week, I did a few things different on the trucks that worked amazingly well.
And I'll be posting on that.

Aside from the few that I have left to build, I also have a stock NJ Custom Brass EP-2 that I will be rebuilding.   
This current thread is a bit unique- this is a super rare version.

However, lots of people have NJ models.  I plan to cover how to rebuild one of those using the techniques I've developed in this thread and the previous thread.   That should be a practical thread that can help a lot of people.

[Caddy58]

Hello Ron,

Admiring what you are doing with the Kumatas.
I look forward to see how you can improve the NJs

I have been tinkering with mine for a while and found that the motors tend to work against each other and that power pickup is mediocre.
So current game-plan is to use two decoders and speed-match the two halves like a two-unit diesel plus adding big keep-alives.
My alternative approach is to junk the motors and replace them with one big can motor.

But so far I have not done a lot of investigations on mechanical improvements, so your insights will be invaluable!

Cheers
Dirk

[u18b]

Hello Ron,

Admiring what you are doing with the Kumatas.
I look forward to see how you can improve the NJs

I have been tinkering with mine for a while and found that the motors tend to work against each other and that power pickup is mediocre.
So current game-plan is to use two decoders and speed-match the two halves like a two-unit diesel plus adding big keep-alives.
My alternative approach is to junk the motors and replace them with one big can motor.

But so far I have not done a lot of investigations on mechanical improvements, so your insights will be invaluable!

Cheers
Dirk

Dirk, if yours can barely pull itself... much less a train.... and it's very noisy and growly.... then that is average for these NJ locos.  Some are truly pitiful.

Fixing the traction issue and general function goes a long long way.   But then, I've found that 2 decoders makes all the difference in the world.   They are programed to the same address, and so run from one address.... but you program them separately.  in that regard, you treat them as speed matching two entirely different locos (which just happen to have the same address).

I use a digital speedometer and get them as close to 100 mph as possible.  But the nature of these locos and their gearing is that they tend to run better in one direction versus the other.    That's where Trim comes in.   If Cab A runs 100 forward but 90 in reverse, then adding revers Trim brings it closer to 100 both directions.  And when each Cab is running the same speed in the same direction, then the noise level goes way down and it becomes much more smooth.

The new and the old EP-2s have a couple of design changed (like the motors in mine is smaller, while the NJ motor are bigger) which make rebuilding them a bit different.   So I will eventually cover all that.

You mention one motor.    One strong motor is always better than two unless the two are perfectly matched- no matter what loco it is.

However, I can't even imagine how I would power an EP-2 with only one motor.

[u18b]

Turning my attention back to EP-2s.

I’ve still not moved forward yet to decals for the experimental scheme.  Hoping to get help in the next week or so.


So while E-1 is pause, I’ve started E-3.
And I encountered something new (sigh).

I had a truck that kept cogging.  I had already worked on it some and gave up a long time ago.   I decided to try again.

And I made a discovery.  One of the transition gears was defective.   You can clearly see that the hole is off center.   Thankfully I had a few spare parts from Kumata.



I’m really glad to have this solved… but it also means I’ll have to be sure and check all the other gears I have left.

E-3 now has the basic assembly with circuit boards and wipers, frame beams added, and tuned up as well as can be expected under analog (which is not great).

Now on to DCC…. But something new.

[u18b]

Adding sound to an EP-2 Bipolar

After a recent challenging but positive experience putting ESU sound into a brass Little Joe, I’ve decided to put sound into an EP-2.  Not sure how well it will go, and not sure I’ll add it to another one…. but let’s try and see what happens.

Here is our target locomotive.  Numbered E-3 and painted in the late cigar scheme.



Of course, before I do anything, I spent hours thinking and planning.   I measure.  I write down ideas and questions.   I think of alternatives.

I’ve done all that and here is the plan I will pursue.

1.  The ESU Nano will be located in the boiler room along with the speaker.  The Nano assembly will plug into the A unit.

2.  The Nano will require six wires into the A Cab (via a plug and socket).

Four of those wires will go to the wiring harness on top of the motor.  These will include the red/black and orange/gray wires.  The other two wires (blue and white) will be to the headlight (with a plug and socket for convenience.


3.   The Nano will ONLY control the A unit.   I have discovered that separate decoders is one of the keys to getting an EP-2 to run well.

4.  The wiring harness inside the A unit will need to be modified.  Instead of a four pin micro plug and socket, I’ll be mounting a six pin micro plug and socket.  This is because I have the four wires coming from the decoder and one red wire coming from the front truck of the A Cab (providing right rail power).

5.  An ESU LokPilot straight DCC decoder will go in the B unit in the same proceedure as I have been doing all along in this thread.

I don’t have any experience doing this (yet), but the LokPilot has a setting whereby it will imitate the v5 sound decoders.

When the time comes, I have no doubt some people on this board will help me get the two decoders to run well together.

6.  I still want a jumper wire to electrically tie the two Cabs together.  The socket in the B Cab will be just like I always have done.

However, the jumper wires of the red and black track wires will go from the B Cab to the track terminal pads on the ESU E24 board.

Also, because the boiler room will now be crowded, I’m going to use a lighter gauge wire- in fact VERY light.   The purpose of the jumper is simply to provide back-up power in the event that power blinks in on of the Cabs.   Thus each decoder of this EP-2 will draw juice from 14 wheels instead of 7.  Most of the time this jumper will not be really doing nothing since adequate power is provided by my wiper system under each Cab.  Power will only truly flow through these wires when one Cab blinks.   My point is- I’m expecting light under-rated wires will be sufficient for this purpose.  They are only a back up plan.

[u18b]

My first decision was the plug I was going to use.
TCS makes a 6-pin micro plug (left) which has all the pins in one row.   I like it, and will use it inside, but the plug to the decoder needs to be a bit stronger in my mind.

Thus I chose the 6 pin mini by TCS (right).  It is a double stacked 3x2 plug/socket set.

EDIT IN HINDSIGHT:  The straight-6 is better because space is so tight.  The 1 or 2 mm savings in height matter.




I then pressed the plug pins into some styrene to make indentations.  Then I could drill out the indentations for the pins making a drilling jig.





I used the jig to drill the holes in a thin double sided copper clad fiberglass board.   NOTE: next time I’ll make this much wider.

Notice that the bottom right pin is not isolated from the outside- that’s because this pin goes to the loco frame and shell and thus the left (black wire) rail.





For cutting very fine traces, I purchased a bit of ebay that has a very tiny burred wheel in it.  This work obviously took a pretty steady hand (and I was still a bit messy)

By the way, since this was a double clad board, I had to remove all the copper away from the pins on the back side.





This is my initial plan.  The socket and board will mount in the A Cab vestibule.   The decoder and Soberton speaker will be all one unit that plugs into the socket and resides in the boiler room.





The reason I made the board so narrow is so that it would fit in the vestibule (even though mounted on the inside).  The weight (as you will see) also has a part to play.

But as indicated above, wider is better.





I then cut a hole in the vestibule wall as shown for the socket.

EDIT:  The hole needs to be a little higher.  Top edge of the hole should be even with the bottom edge of the side cab window.  (otherwise not enough room for the decoder once snapped on).






Here is a test fit.   The board goes on the inside with the solder pads available.

[u18b]

I then roughed up and cleaned the brass in prep for solder.   Before I got going, I strengthened solder joints in stress areas.   In a  shot following, you’ll see that I added extra solder to the screw supports and to the drawbar mount.






Before I mount the board, there is one more prep stage.  The main Cab wall has a sharp edge (see left).  I use a motor tool and round that sharp edge because wires will go across this spot and we don’t want the insulation pierced.







I then solder the board to the inner wall of the Cab.  My initial solder joint was with the plug installed so as to get the board properly positioned.

Notice all the pads are isolated except the top left.  Also, solder spills over the edges to the wall to hold fast.   It is also held by some copper cladding on the back side edges.






I’m ready to mount the socket.  But how will I get to the back pins?   I trimmed the front row first, and then soldered the front row.




After soldering all the pins, I use a motor tool to cut the excess down.   I still have to clean all the metal particles out of the shell and board.






Here is the outside view.  The board is soldered to the wall and thus the socket is held firmly.   I press on the socket to make sure everything holds.  If I have to start over, butted to find out now.

Everything is strong.




I use special tools to clean the traces.  Years ago, I acquired some retired dentist tools.  They come in handy when you need them.




Now the check:
Do I have any solder bridges or shorts? 
I use my multimeter to check and make sure there is no conductivity between any of the pins.

Here, I’m checking to see if there is continuity between this pin and the shell (there isn’t).   The only pin that I WANT to have continuity with the shell is the top right from this view.




No two pins should have continuity.   I test all combinations.





Well, that took an entire evening.
The next modeling session is to create the decoder board/plug.

[u18b]

One more quick mod.

The Cab weights are a vital component of this locomotive.

But the weight will obstruct the circuitry we just installed.

So I need to modify the weight.

The weight fills the cab and rests on the gearbox.  But Kumata designed it for NJ custom Brass years ago with a notch for the jumper wire that ran through the boiler room.

I just need to widen that notch.  The stock weight is on the left.   I made the projections at the bottom thinner with a motor tool be removing material from the inside edges.




Here is the underside.  I have to remove material so the flywheel will not scrape the weight.






Here is the weight installed.   Notice that the sensitive pads in the middle do not touch.   I don’t care if it touches on the outer rim of the board since most of the whole mechanism is charged to the black wire.

[u18b]

Now I’m ready to start working on the decoder.
I’m installing an ESU Nano with the newer E24 board (not the awful NEXT18 board).

I pop the decoder off the board and set it aside.  No need to risk shorting it out with static during construction.

Here is the stock E24 board.





Now, I need to remove all of the wires except for the brown speaker wires.  Spreading out the wires here.

I took a photo so I could also REMEMBER which wires go to which pads.

In brief, notive the pairs are:
track  (red, black)
motor (orange, gray)
headlight  (blue, white)

A shortcut:  the B&W colors go together (white black gray)

A VERY fine soldering tip is required.






Here, the wires have been removed.





Here is the stock TCS mini plug inserted.  Notice black goes up.  That was the way I soldered the board inside)




The idea is to create one compact unit that plugs in.
So the mini plug will lay right on top of the board.   
I have the action end with pads I need on the right to give me room to work.  Speaker wires on the left.
The edge of the board is equal to the edge of the plug housing.





I worked inside out to keep the hot iron away from completed work.  So blue and white go first.





The red and orange.  This pair did not go as well.  The orange wire got too hot and the 90 degree bend melted the sheathing away a bit.

It’s only a little bit, but I didn’t want it to happen again. So I made an adjustment to the lats pair.







I pushed everything to the right so the the black and red wires did not com in at a 90 degree angle but were more straight when soldering.

Much better.





Now I pull everything back to the left where it needs to go.




Here is a quick little test.   The board should fit inside the boiler room.





Before I’m finished with this stage, I need two more wires.  I need jumper wires for the track (red/black) to run to the B Cab so that all track power is shared by everything equally.

I explained earlier that these are very thin wires because these only really kick in if the power blinks in or or the other.   If Cab A blinks out because of a dirty spot on the track, then the B Cab provides electricity to the A Cab keep going.   Think of it as a partial (micro) keep alive.

These thin wires came off some pre-wired micro LEDs.

[u18b]

The plug will lay right on top of some open pads on the board- and there was a tiny gap on the insulation on the plug.   In order to totally remove the possibility of arcing, I insulated the pads.




Time to lock all of this down since space will be tight.   I mixed some epoxy and sealed the wires to the board.   I want this to be one unit.   When I pull on this to removed the assembly from the socket, I don’t want something tuning loose.

I made a couple of applications.




I then turned back to the A Cab.    I had saved the wires I removed from the board earlier.  I now soldered the needed wires to my board behind the socket.  3 wires to the motor (red, grey, orange) and 2 wires to the LED (blue and white.

This was actually rather difficult.

When finished, I did another check looking for shorts.  There were none.




My epoxy is dry and time to do a little test.   





I attached the Nano… and it would not easily fit.
The caps on the Nano hit the drawbar support in the boiler room.  Rats.  Am I totally hosed and will have to start over???

In desperation, I bent the socket upward and soldered a brace under it to hold it up.





This worked- though space is tight.

I turned my attention back to the inside.   The wires must follow the inner roofline toward the front.  So I bent them down.




And then installed the modified weight.






Now it’s time to make a plug for the red, orange and grey wires to go into the socket over the motor.




Plugged in.   If you’re wondering where the black wire went, we don’t need it.    The frame and shell are charged to the black wire.





And here is a test fitting.   Tedious work, but slowly making progress.

[u18b]

Now time to add the speaker.  Space is tight.

I had about 6 mm to work with (bottom of the speaker to top of sound chamber).

I like to build sound chambers out of acetate because it is thin and tougher than thin styrene.








I turned the speaker box upside down and soldered the brown speaker wires.





I then flipped the speaker over to its final position.



I added epoxy to lock the speaker box down.  But I was sure to add the epoxy in the middle because I did not want to block the air holes in the corners.





Ready for the decoder.  Because of proximity to metal, I had to protect the decoder.   Those capacitors will almost sit on the weight in the boiler room floor and one close to the draw bar support.

TCS sells its clear ultra-thin shrink wrapping separately.  In this case, shrink wrap for the TCS M1 was a good size.   It is VERY sensitive to heat- fire will melt it in a heartbeat.  A hot hair dryer is better.

Of course, when finished, I had to cut the shrink wrap away from the E24 socket on the Nano




A quick test shows that we have sound!

[u18b]

Yikes!  I had an almost disastrous problem.

The decoder is awfully close to the draw bar post.

At first, I could not get the boiler room on!!!  The drawbar support on the boiler room would not fit in that gap.

Will I have to re-do the socket and circuit board?





I needed another millimeter.  And the only way to get it was to bend the socket upward some more.

I studied the socket and saw that there was a small gap between the brace I had soldered and the socket.

So I used my resistance soldering tweezers and raised the support bar and got that socket as high as I dared without risking breaking the board loose.

The boiler room will now go on (just barely).






At any rate, this has convinced me in hindsight that the straight in-line 6 pin micro plus and socket by TCS is the way to go next time (plug on the left). 

A millimeter or two matter!   That’s how tight space is.

I’ll use the straight-6 next time.

[u18b]

With the A Cab and Nano now working, time to turn my attention to the B Cab.

I installed the TCS 2-pin mini socket into the vestibule since I only need the black and red wires.

However, I made a couple of changes from my typical install.

Because space is so tight, I was worried the plug and socket would obstruct the Nano.   So installed the plug high in the vestibule.






The second change was the soldering of the circuit board to the inside wall of the Cab.

Based on my previous experience with small traces coming loose near the edge of the board when heating and soldering, I made this board much wider.

Soldering was much easier and better.






This meant that I had to trim the back edges of the weight a bit.

Since the weights were now very unique to each Cab, I marked each clearly.  That way, when everything is taken apart for painting, there will be no confusion which weight belongs where.

[u18b]

The next change was the decoder.   In the past, I used Digitrax DZ126 decoders.  They are good quality decoders, but I originally chose them for this application because they also have the soldering pads all lined up on the ends.

As seen here.




Keeping everything modular, I then installed a LED resistor on the side with superglue…






And then a socket for the LED plug and wires.





But now, I’m going with an ESU LokPilot for compatibility with the Nano.   It also has the needed solder pads all lined up on one end.




I removed the shrinkwrap and all the unneeded wires.  Keeping the orange grey red black (and yes, I wish the ESU people would get a color wheel and figure out what “red” means-- since their “red” is really orange, and light orange is what they call orange.)




The ESU decoder is smaller than the Digitrax DZ126,   I was going to mount the resistor and the LED socket on the side as I did before, but another option presented itself.

ESU places the yellow and the blue pads exactly opposite each other.   So it was tempting to simply solder the socket right there.  This would obviously only work for the B Cab.   Doing this for an A unit would be trickier because we would need white and blue.     But this seemed perfect for the B Cab.

Space is tight, but I was able to do it.






I then shrinkwrapped the decoder with TCS M-1 shrinkwrap.   I added a little epoxy to seal anything still a little exposed.





I then added a TCS 4-pin micro plug.    I always add superglue to the wires at the plug so noting pulls away.  If not, you might pull one of the pins totally free when unplugging.  Superglue forces everything to act as one unit




The LokPilot is installed.
Adding the LED socket raises the height of the assembly, but since the decoder is shorter, then I think I’ll break even and this will work (I’ll know better when I install the LED headlight).

Another issue is the resistor.  Instead of mounting it on the decoder, I’ll have to remember to install it in the line of the LED or I’ll blow the LED.