New brass Milwaukee Bipolar EP-2 clinic- part 2

[peteski]

Ron, I'm following your thread in interest. There are some excellent ideas utilized.

As for decoders with all the solder pads lined up at the edge of the decoder, any brand of decoder which has the 6-pin (NEM651) interface (either pins directly in the decoder or connected by wire) will have those 6 solder pads at one edge of the decoder bard.  As you discovered, Digitrax is not the only one.

[u18b]

Thanks Pete.

The reason I mentioned this is there are small decoders that are not thus designed.   At the time I began this project, the TCS Z2 was very small- but the wiring VERY hard to change since a couple of the wires did not solder at end pads.

The ESU LokPilot is really great for this project.

[peteski]

Thanks Pete.

The reason I mentioned this is there are small decoders that are not thus designed.   At the time I began this project, the TCS Z2 was very small- but the wiring VERY hard to change since a couple of the wires did not solder at end pads.

The ESU LokPilot is really great for this project.
p

That's true, there are few small motor-control-only decoders which do not have the 6-pin NEM651 solder pad layout

As far as decoders go, ESU decoders in general are great decoders, much more flexible and advanced than the domestic decoders are, plus the price is comparable with the domestic decoders.  The big advantages are that their firmware is upgradable, and they function outputs can be fully re-mappable. They also have more function outputs than comparable size domestic decoders.

[u18b]

I installed a plug for B Cab.   The plug is soldered to the Nano E24 board at the red and black track pads.




I then marked the plug and the socket with a silver sharpie so that correct polarity would be maintained.




Silver side to silver side. 
Raising the height of the socket on the B Cab made for more room.

[u18b]

OK.  With all decoders installed and each cab tested to make sure it runs, now it’s time to program speed matching.

Each Cab has it’s own decoder, and so I have to treat each like a separate locomotive.

But the DCC address of both will be 3 (which is DCC default- but also happens to be the road number of this one which is E-3).

This means I can only have one Cab on the track at one time.   They have to be programmed completely separate from each other.

If you are joining us late, I mount the motors with opposite polarity in the two Cabs.    Each Cab has the same identical wiring harness and configuration except….
There is a casting spue mark in the plastic housing of the motor.

Cab A motor is mounted with the sprue mark up
Cab B motor is mounted with the sprue mark down (thus opposite polarity).

That way, since the Cabs face back to back, when I input Forward on the throttle… the B Cab runs backward.   I felt this was MUCH more clear and easy than messing with CV29 and programming one to run backwards as forward- which can be wiped out with the first decoder reset.

The first step is to set the LokPilot to the same settings as the Nano (not sure why they do not come similarly set).  For example, stock acceleration is longer in the Nano than the LokPilot- so I made them the same.

Time to adjust Vmax and Vmid and Vstart.  A 3 point curve is all I need right now.

I use an Accutrack II Speedometer.
I warm up the Cabs a couple of minutes before measuring.

And then start trying to get the Cabs running as close as possible to 100 scale mph.   That’s my reference for this loco- even though it is a bit faster than the prototype.





Ok- this is close enough since there is a natural variability of +/- 2 mph.



Now, because the 4-axle power truck greartower is offset and not centered, an EP-2 Cab tends to run faster in one direction than the other.

That's where the TRIM variables come in.  ESU default is 128 (mid-range).

If Cab A runs at 100 mph in forward, but only 92 in reverse, then I increase the Reverse Trim variable and keep doing so until is runs 100 in both directions.

Remove A from the track and do the same with the B Cab.

Now, here is where ESU helps me more than Digitrax.   I am not a programming savant like some of you.   This comes hard for me and I often have to read directions.

So it was only recently that I discovered that in order to use Trim variables with Digitrax, you are required to input a 28 step speed curve.   That’s a pain.

ESU can do it with three basic speed steps.  Much easier.

[u18b]

When all is matched well, I then take both Cabs for a test drive WITHOUT the boiler room.





When I run it this way, I watch the slack in the jumper wire.    If it is very tight, then the lead unit is running too fast.

Here is a test video.   You can see that there is some variability when starting and stopping, but at running speed, the Cabs are pretty well matched.

Later, I can fine tune some more.
By the way, that white and blue wire wrapped around the headlight is the wires for the LED for headlight.   They are there just to get them out of the way and prevent a short.

[u18b]

And now the moment of truth.  A full initial sound test (well actually, I already made a few adjustments to volume levels).

On the EP-2 configuration, the Nano sound decoder is mounted to the A Cab.   I have to install the boiler room on Cab A first.




Then Cab B.   You can see the plug wire length is just enough to get my fingers in there to insert the plug into the socket.





The ESU file is the GG-1.

The clicking noise at start-up is the pantograph rising to the cat wire and then the generators and compressors starting.

The adjustments I made were:
increase horn.
increase bell
decrease whine
decrease annoying sound of shifting speed

Well, this has been a challenging install.   It took me a week of evenings.   But I think the extra effort has been worth it.

[jwaldo]

Ron, you are the inspirational madman who keeps me from going sane. And you've raised (or maybe lowered?) the bar yet again.

[peteski]

Ron,  while not regularly done, you can program 2 decoders (especially from the same manufacturer and same series) simultaneously on the programming track (and of course if they are at the same address, same can be done on main line).  That would make their programming less time consuming. The only possible issue would be speed matching (since there will be some mechanical differences).

 Or if you want to program each one separately (for things like speed matching), just program different address in one of them, then place them in advanced consist to run them together.  When in that case, you should be able to program the individually on the main line by using each loco's actual address (not the consist address).

[u18b]

Great observation.  It would definitely make programming a 28 step speed curve into two Digitrax DX126 decoders much easier.

But....
I'm thinking the ESU LokProgrammer will only recognize one decoder at a time.

I can't imagine that it would program both a sound Nano and a non-sound LokPilot at the same time.... would it?

[peteski]

Great observation.  It would definitely make programming a 28 step speed curve into two Digitrax DX126 decoders much easier.

But....
I'm thinking the ESU LokProgrammer will only recognize one decoder at a time.

I can't imagine that it would program both a sound Nano and a non-sound LokPilot at the same time.... would it?

LokProgrmmer does not use the standard DCC programming method, so no, I do not think that would work.  But LokProgrammer makes complex programing easy and fast (much faster than standard DCC programming.  As you probably noticed, it can program hundreds of CVs in a LokSound decoder (Decoder Data) in just few seconds. But you can duplicate the projects so you can use the CVs from the LokPilot in the LokSound decoder (or vice-versa).

[u18b]

Time to deal with that blue and white wire coming out of the front door of the nose.  Of course, as LED wires, they are a shorting hazard.   

My experience with failed decoders has been that red/black shorts are not that big a deal (the auto-shutdown of the DCC system takes care of those), but a short in the LED wires will fry a decoder.  I have killed a couple of decoders when the shell pinched and pierced decoder wires.

Figuring out what to do with this application required some thought because space is so tight.

In the standard DCC configuration for an EP-2, I mount the LED resistor right to the decoder.  I then mount a TCS micro 2-pin socket to the decoder.   I then wire up an LED with thin magnet wire with a plug so that the shell may be removed with no soldering.  Consequently, there are simply 2 thin magnet wires in the nose.

But adding sound changed everything.
Normally I have ONE red (right rail) jumper wire coming through the vestibule and running along the inside roof, but now with sound I have 5 wires.  So the inside of the roof is getting crowded. 

My concern is room for a possible socket and LED.   Do I run the thick blue and white wire all the way forward?

After some measuring, I decided to terminate the wires on the side of the inner shell.   This is in the area of the gearbox which is much thinner than the shell walls so there is a gap between the shell and the gearbox.

So I made a small 2 sided circuit board. 
I soldered the board to the wall to hold it in place.

I then trimmed and routed the blue and white wire to the side board with a 2.4 k resistor.




Just to be safe, I cleaned the gearbox side with an alcohol prep pad to remove grease and oil and then placed Kapton tape on it to provide insulation and  protection from shorts.





I then trimmed the weight just a little to make sure all fit properly.

And this is where I’ll stop for now.  I may cover the whole board with Kapton tape temporarily.

I’m not installing the LED yet because I have tons of work to do on this loco to add detailing- some of which involves heat and soldering (e.g., I have to replace the bell on the roof).  I don’t want more wires to contend with right now.

Also, the magnet wires are simply covered with lacquer.   Any solvent/de-greaser I may use later may damage the coating.

So adding the LED will be about the last thing I do on this loco.  And with this configuration, I won’t need a socket and plug.    The LED can be hard wired to the solder pad since there is nothing to unplug.  The LED and power source are a part of the shell assembly.

I plan to run the magnet wire back down to the weight, turn under the weight and go forward up to the headlight.   Epoxy will hold the magnet wire in place and avoid obstruction with the flywheel and motor.

[u18b]

This now concludes the sound installation on this EP-2.

[peteski]

Great job as usual Ron!
Good to see you fully embracing using PC boards in the installs.  While they take some extra time to make, they result in clean and need wiring.

[u18b]

Continuing from my last post..... because I had a Homer Simpson DOH! moment.
(

/>
OK..... adding a Nano sound decoder to the back of the A Cab required a change to the LED hook up in the A Cab.

Here is what I did before:  I added a small circuit board for the LED resistor and wire leads.



So yesterday, I was working on another EP-2....
(that's the spot I used before).....


And then, when I was handling the B Cab...... DOH!

Look at all that room in that crevice!    The LED board will fit easily right in there.



So now that I've installed my first sound decoder in an EP-2...

NOTE TO SELF:   Next time, Install Nano on the outside of the B Cab!!!