|
Up
| |
| Kato SD90
with Soundtraxx DSD-AT100LC Decoder |
| Situation |
| One of my train buddies had bought four
identical engines (different road numbers), Kato’s Union Pacific, SD90’s.
He wanted to run them on our layout at the show and possibly at
home. He also wanted to have
sound in at least one. The
cost of several would just be too expensive.
This engine then, was to be the one engine that would be run most
of the time. I had done
several others for this train buddy and had put things in them like ditch
lights and red running lights for some NS SD70’s.
He asked to put ditch lights in this, the lead engine, but no cab
lights or other running lights.
And interestingly, we have avoided cab lights because they are more
of a novelty and a use for one of the six functions that Digitrax has in
their new decoders. They are
important to me to show the capabilities of DCC so I put them in mine.
Running lights, especially truck lights, are also something that
has not caught on; however again, I do it to mine to show off. |
| Decoder Selection |
| Since we are limited by space, I wanted to
put in just one decoder and picked the Soundtraxx DSD-KT100LC.
A quick call to Soundtrax found that this is no longer in
production. The substitute
was the DSD-AT100LC. The
particular model (820041) was a 2nd Generation EMD sound with
selectable horns. It had amps
matched to the engine and was a drop-in replacement for the factory
installed light board. On the
other units I will use a Digitrax DH123-PS (short plug) and leave the
lights alone. |
| Space |
|
Critical in this unit to some degree,
there was a lot of room, for whatever reason, under the Q-fans at the rear
of the unit. It looks like
they originally designed the undercarriage and weights to fit an SD70 and
didn’t “up-size” it to match the larger SD90 cowl.
In any event, I was looking for some air holes for the speaker(s).
The fans are always a good choice as they point up and under normal
circumstances usually have open grills.
Not so lucky in this case.
Even though the fan detail is marvelous with 10 blades showing
through the 52-inch grillwork and open, at least at the top, these were
closed units on the bottom.
Options included drilling holes or otherwise opening the bottom of the
fans without disrupting the top grills and molding or, as I chose, cutting
off the fans and replacing them with a detail set from a detail
manufacturer. This was
actually simpler than I had expected. The fan set is four pieces with a separate three-hole piece
just right for mounting details and hiding the ugly saw marks from the now
detached fans.
The drop in was longer than the
original board. I was going
to use LED’s, as I usually do, and Kato had already provided them.
I just had to unsolder them and then solder them onto the board
WITH a proper sized resistor since the Soundtraxx board is built to
accommodate electrically 12-16volt incandescents.
They do not recommend LED’s because there is still a 1.5v AC ripple
even when the light is supposed to be off.
Some LED’s will have a faint glow with this residual voltage.
There are several solutions; the best is to not do anything about
it. Most prototypes will use
“Rule 17 dimming” anyway.
The next space consideration was the
size of the speaker(s). I was
going to use 2 one-half inch speakers but read the “NOTE” on the
instruction sheet for the DSD-AT100LC which stated that the decoder was
designed for a single speaker.
So, I elected to use an elongated speaker which was easier to
mount, more power output, more surface area than the two that I originally
planned on as well as easier to enclose.
The speaker would not fit in the space under the fans without a
little shaving. I didn’t want
to shave the speaker so I shaved the enclosure.
This is a separate removable cowl from the main, outer hull.
This made it easier to file and shape, kind of… |
 |
 |
I spent an awful lot of time shaving and even thought
that I should have just cut off the area and filled it with the proper
width of styrene. The issue
would have been the retaining pins for the inner fan cowl to the outer
hull. These were in a spot
that I would have to cut off.
I thought it best to continue with the filing.
I also felt that it would be a tighter fit for the speaker
enclosure…maybe not. |
 |
 |
| Now that the choices had been made it was
time to make these choices work.
After carving out the spot for the speaker, sever the rear light
pipe about ¼ inch from the edge of the third fan hole.
This will allow the addition of an LED, shrink-tubed (a new verb!)
onto the end of the light pipe.
I polished the end of the light pipe before attaching the LED.
Cut the leads back to a length that provides comfortable solder
length without interfering with the speaker enclosure. |
 |
 |
| Now, directing our attention to the
speaker. This one has four
mounting holes. While good
for mounting, here we must fill them to remove unwanted air leakage.
I used plastic 2-56 screws.
You could use any left over sprue plastic and just glue them in. I couldn’t find any plastic that size (.080). |
 |
|
Next create a box.
Here we just need three pieces of plastic, all 0.630 inches wide.
Cut them to length to form the bottom and two sidepieces.
We will have to seal as we make; so use silicon adhesive that will
stay soft. Recommended
“gasket” cement (RTV for room temperature vulcanizing, Permatex or generic
brand) from a local auto parts store was used.
The trick here is mounting the capacitor.
I usually mount it very close to the speaker and in the enclosure. On this enclosure the plastic touches the speaker magnet.
Not much room here. I
had to mount it externally.
Just keep it close. Also,
remember to drill the holes for the wires and pull them through BEFORE you
glue on the final cover. You
may want to mark them for plus and minus.
A lot of posts on this subject suggest that you should try it both
ways and then mark the better sounding polarity.
At this point you should be ready for
the electrical. |
| Electrical |
 |
This decoder is considered a “drop
in”. While it fits in the
space of the other board, it is far from the intended NMRA “plug in”.
First, lets consider the motor connections.
To take out the light board you must squeeze the connections to the
motor to release them from the holes.
You would then drop in the new board.
I left the mounting tape on the board for insulation and and extra,
but unneeded, adhesion. Next,
cut back the left and right, front and back, wires from the track pick ups
off of the trucks. Solder
them into the holes provided.
I pushed them up from the bottom and soldered on top.
Carefully straighten out and flatten
the connections to the motor.
Then bend them over with needle-nose pliers just above the “tabs” towards
the board. When you do this
they should both just fit over the board and in between two through holes
(called through-vias in the printed circuit board business) and solder to
one or both. You should
pre-solder or tin the motor connections with a dab of solder to make it so
that you are not putting a lot of heat on the board.
Note that these vias are connected to the motor “tabs” on the side
that would ordinarily be used for Atlas engines.
A stroke of genius, I might add, by Soundtraxx engineering to allow
this board to be used for both manufacturers’ engines.
|
|
Add a 510-ohm, ¼ watt resistor, in
series with both LED’s. I
usually connect the resistors to the board if there is enough room.
The rear LED will have to be connected by wires to the board
because the speaker and enclosure is now in the way.
You might consider putting a connector between the board and the
LED. I did not as these
usually fail over time with corrosion.
Separate one-wire sockets are on the
board to connect the speaker.
The jury is out over whether this is a good idea considering my experience
with long-term corrosion. You
really need to do a lot of repair work to get an appreciation for what I’m
talking about here. Brass,
copper, even tinned copper oxidizes constantly.
They are more susceptible when there is a current flowing.
If you want to avoid the problem you could solder the connections
but I would wait until I tested both polarities before soldering. I soldered. The
connections had popped out every time I opened her up and I was afraid
that over time this might happen and cause a short.
Connections done, lights in, I test
along the way. I had already
tested the motor and pick-ups before adding the lights and speaker.
I had already programmed in the address and ran it around the loop
a couple of times also. Once
with the 03 address and again with its new 8284 address.
It ran rather rough considering the quality of Kato engines. |
 |
Button it up
|
|
Considering that my normal path is to
almost overhaul an engine, even if brand new, I continued with my now
famous Train Buddy™ regimen of engine service.
This consists of basic lubrication with three kinds of lubricant at
appropriate places and gauging and repairing any deficiencies in gauge
with the wheels. It may also
require truck interference testing, especially on large, 6 wheel truck
diesels or articulated steam.
Removing any interference (even Kato is not infallible), I am ready to
test again with the shell off before I commit.
This is a valuable test, since; many times a unit will fail when
the cowl is pressed into place because of wire or board pinch and
compression. Sometimes wires
even get pressed into the flywheels, drive shafts or the motor itself.
While we all try to be very careful here, it happens all too often.
Test again with the cowl or shell in
place. Test the lights (you
should already know from previous testing whether they will stay lit as
suggested in the Soundtraxx documentation or go out).
If they don’t go out and you want them to, then a simple diode,
capacitor arrangement will insure that you get full DC with no AC
component. This takes up more room in the shell and must be carefully
insulated. A short or
incorrect resistor value will destroy the light output functions on the
decoder.
Ready to go.
You might want to add other light functions but this decoder
doesn’t have any. Its major
feature is sound, so enjoy.
If you want ditch lights, beacons and the sort then you could use a TF2 or
TF4 from Digitrax that is very space conscious to add light functionality.
Just make sure to program them with the same address.
In order to do this you must separate the decoders electrically
from each other while programming.
The programming track for Digitrax is not designed to handle
programming more than one decoder at a time.
You could also lash this up with a front engine that has all those
features and get a great combo.
My train buddy John will be running four of these beauties
together. Can’t wait for the
next show to wow the youngsters and oldsters with this sound combo. |
| Programming |
|
One last note.
The standard program for this would need to be modified for the
prototype. The three horn
sounds are somewhat industry standards.
Look at the horn that came with the unit.
A single bell (horn bell here) means probably a Wabco.
Three bells of varying sizes, sometime in two different directions,
is a K3 (Leslie usually) and Five bells, ether in one or both directions
is a K5 or five-chime (horn-speak for five separate musical notes or
tones) horn. While there are
some exceptions to this, usually related to era or road, the manufacturer
of the engine has done their homework and has picked the right horn for
that engine. Again, many have
gone through a progression of horns from Wabco, single note in the early
50’s to late 60’s to K3 and/or K5’s up until recent times.
Only good prototype history digging, along with a picture of the
unit that you are modeling, will give you the correct prototypical answer
(if you even care). In this
case it was a K3 so Soundtraxx CV115 was set to 1 which is a “medium”
horn.
These are the other CV changes that I made from "factory default".
-
CV2 (start) = 14
this makes the engine just start to move when the speed is "01".
-
CV3 (acceleration) = 10
this makes the engine hesitate a little when starting out as
though it was dragging a long line of cars.
Allows the sound throttle to match the locos speed better.
-
CV4 (brake) = 10
again used to match sound and speed.
This means that even when you turn the throttle completely down
it will drift for three or more feet before it comes to a complete stop.
The faster it is going the longer this distance is.
(real world you know.)
-
CV115 (Horn) = 1 This is
the setting for medium horn.
According to Soundtraxx this is a 3 chime (bell or note) horn.
According to the model, that is what is physically modeled on
this engine. My research
verifies that this is correct.
Don't change this setting (unless you just want to hear the other
options.)
-
CV116 (notch) = 11
Assuming that the top speed of the engine is when the throttle is
at 99 (I didn't verify this) and there are 8 notches and they are linear
then splitting the 99 % range of the throttle into 8 pieces gives us a
notch setting of 11. This
shouldn't change.
-
CV120, 121 and 122
(volumes) = default (for
the bell and exhaust it is set to 128 and the horn is 192.
These can be changed while
operating. Look in the book for "operations mode programming" |
| Test and Refine |
|
While ordinarily I would have caught this sooner, it was only after I ran
it one last time that I noticed it.
The rear light would go on faintly every time you blew the horn!
I called Soundtraxx again, this time to get a sanity check.
Was this normal, did I do something wrong or was the board
defective? As it turns out,
for some odd reason, on this board, and only under the circumstances
mentioned with an LED, this happens.
Well, I set out to “fix” it.
My solution was crude but worked.
I put a 240-ohm ¼ watt resistor across the leads of the rear LED.
This creates a resistive load for the driver transistor and also a
voltage divider circuit. Too
small of a value (220) and it will keep the light from coming on at all.
Too large of a value (310) will not have any affect.
You can do this test for yourself carefully put a 500 ohm
potentiometer across the LED with the lid off and vary it until you get
the best results. Measure the potentiometer value and replace it with the next
size commercial fixed resistor.
Now for the (ugh) 73 part detail set. |
|