M&K Junction Railroad

M&K Junction Railroad
Another train of eastbound coal crosses the Cheat River

Wednesday, September 29, 2010

Weaver Pacific Decoder Install

The Weaver Die-Cast Pacific can be equipped with a decoder almost (I did say almost!) without cutting a wire. Most importantly, the tether between the loco and the tender does not have to be cut or rewired. Nor do you even have to go into the tender, as I've already traced all of the wires for you. I actually did this installation in 2009. No video of this installation as I did not have a video camera back then.

The Pacific's performance on straight DC was OK, but slow speed starting and smooth running at realistic speeds left a bit to be desired. This was the perfect candidate for a first-time steam DCC conversion. How I converted this Weaver die-cast steamer is applicable to die-cast locos made by K-Line, Lionel and MTH as many have identical or very similar mechanisms.

The schematics for both the loco and tender are below (click on the image for a larger view), I'll refer to the loco diagram later, the tender is included for your information:

Following the standard convention, the tender picks up current from the left-hand running rail and sends power forward to the locomotive through the tether. Since power from the right rail is not available in the tender, but power from both of the left and right rails is available in the locomotive, it became clear that it would be better to mount the decoder in the locomotive rather than have to bring the right rail power back from the loco to the tender and then send the DCC-controlled motor current back forward. In this way no rewiring of the tether would be necessary.

Locomotive disassembly came next. The loco shell comes off with two screws mounted under the cab, approximately under the backhead, and two screws mounted out-of-sight in the steam cylinders. Watch out, there are a lot more left-over switches from the 3-rail model under the cab of the locomotive, do not undo these screws; also avoid the screws attaching detail parts such as the injectors.

With the shell off, the locomotive looks like this (click on any photo for a full-size image):

The red and black wires emerging from the boiler shell are for the front markers; these are LEDs and you must observe polarity in hooking them up. The black wires on the right are for the front headlight; this is a regular bulb. The red and black wires on the left are for the cab light. Curiously, while the markers and headlight have connectors to ease disassembly, the cab light does not. So to separate the boiler from the frame I popped the cab light out of the rubber grommet that holds it in the cab and undid the wire nut that attaches the cab light to the boiler shell.

There are two circuit boards mounted in the rear of the frame, under the cab. I’ll refer to the ‘T’ shaped one into which the tether plugs as the ‘Tether Board’ and the other as the ‘Main Board’. See the photo below:

Begin by removing the main board. Unplug the connectors, W2-W5, on the right and bottom edge of the board (as oriented in the photo). Remove the single Phillips-head screw that holds the main board to the frame. Cut the two red wires as close to the main board as possible. These wires are carrying power from the rails to the main board and you can reuse them later. Put the main board aside, it is no longer needed.

Now you will have an empty space in which to put a decoder, but there are some posts in that space which, if they shorted out to the decoder, would destroy it (see the next photo).

To prevent a short, I insulated the posts with heat shrink tubing. The photo shows all three posts insulated. Although the insulated posts will hold the decoder off of the frame; just to be doubly sure I covered this entire area with electrical tape, except for the vicinity around the stud for the trailing truck.

Now for the decoder; I chose the Lenz Gold Maxi because one objective for this conversion was to obtain smoother engine performance and Lenz is known for its motor control capabilities (see next photo). This is a 3.5 amp decoder with 6 function outputs and quite capable of handling this locomotive and, coincidentally, it fits in the space that has just opened up.

Every DCC installation requires that the motor be completely isolated from track power. To do this we must remove the brown wire from the screw located between the second and third pair of drivers on the right hand side of the frame, this wire normally supplied right rail power to the motor.

Now you have a choice. You can reuse the two red wires that you cut away from the main board to supply track power to the decoder, this is the simplest connection. If you do reuse these wires, attach them to the decoder's track terminals; the decoder doesn't care which rail is which. However, these wires are small and the solder joints to the tether board are poor, this could lead to reliability problems – the solder joints might even give way during the rest of the installation.

I chose instead to put a new wire on the screw where the brown wire was just removed; the red wire that I used follows the DCC color code for the RHS running rail and provides a robust, low voltage-drop connection between the rail and decoder. I used a crimped-on terminal, but wrapping the stripped end of a wire around the screw would work (next photo).

For the same reliability reasons I soldered a black wire (DCC color code for the left hand side rail) to the point on the tether board with the red and gray wires (refer to the schematic for the loco above). This point has the connection to the left rail via the tether and the tender. By using a more robust wire (I used 20 ga.) and making a good solder connection I have eliminated a potential trouble spot.

I soldered gray and orange wires to the motor terminals (per the DCC color code) to provide a larger gauge of wire as well as to replace the typically poor oriental solder work. The motor in many of the "China" steam locomotives is in "backwards", that is the motor is towards the front of the boiler with the shaft pointing rearwards into a gearbox mounted, in this Pacific, to the center driver. This does not make a difference except that you have to guess which terminal on the motor gets the orange wire. Put it on the same terminal that had the brown wire that was attached to the frame of the locomotive. If you guess wrong it only means that the loco will go backwards when you command forward and vice versa. If this happens don't worry. Most recent decoders, like the Lenz, have a CV that allows you to select which direction is 'forward'.

The gray and orange wires will have to be snaked alongside the gearbox to reach the motor terminals. I found it easier to do this by removing the top of the gearbox with the motor attached to allow the larger gauge wires to lie alongside the gearbox. While I was at it, I ran blue and white wire (DCC color code for the headlight) along with the gray and orange. Replacing the top of he gearbox will retain the wires running alongside it.

Was all of this rewiring strictly necessary? No. But I haven't seen a common consumer product from the orient that had good soldering. Several of the wires broke loose during the installation. Once I reassemble this loco, I don't want to have to fiddle with it again. Therefore I rewired it.

If you choose not to rewire, run the two red wires from the tether board to the track inputs on the decoder, run the two black wires from the headlight connector to the headlight outputs of the decoder (blue and white) and reuse the brown and gray wires already attached to the motor to connect to the decoder's motor output.

The next picture shows the Lenz decoder placed in the space. The seven-terminal strip on the left are the 6 function outputs plus common. The nine-terminal strip on the right is the track input, motor outputs, headlight and backup light plus common; as well as a pair of terminals used with a reed switch for station stop effects.

It's not entirely clear in this picture but the decoder is sitting on the three posts that I had insulated previously. One stud is longer than the others and the decoder is high on the right end. The fact that it is not level has no practical effect.

The next photo shows how I secured the decoder in place. I insulated the left hand side of the decoder board (near the seven-terminal strip) top and bottom with electrical tape. Then I loosened the tether board and captured the insulated end of the decoder board under it. Finally, I re-tightened the screws on the tether board trapping the decoder in place; a neat and simple way of holding the decoder in place. Click on the photo for a larger view and examine the picture in this area and you'll get the idea.

This last picture shows the decoder installation almost finished. Notice that some wires are attached to the nine terminal strip. Starting at the top you can see the read and black track inputs. A few terminals down you can see the blue and white wires going to the headlight. Finally the last two terminals are the orange and gray wires going to the motor.

Not shown in this picture are the wires going to the backup light. Refer to the schematic for the loco above. The backup light is connected to the wires going to socket W2 on the main board and the corresponding solder pads on the tether board are illustrated on my diagram. Connect these wires to the terminals on the nine-terminal strip for the backup light (normally blue and yellow).

Hookup of the headlight and backup light are simple as these are incandescent bulbs. Either wire can be connected to the blue and either the yellow (backup light) or the white (headlight).

Connecting the markers is not so simple. It was not clear to me, and nowhere in the Weaver documentation is it mentioned, whether the markers are bulbs or LEDs. On closer inspection I assumed that they were LEDs. The Lenz documentation does not indicate the polarity of the function outputs. I determined the polarity with a meter, traced the connections through the tender and connected the LEDs with the proper polarity. However, I assumed that there was a dropping resistor in line with the LEDs somewhere. Wrong. So when I activated the function I got a brief flash of red from the rear markers then snap, crackle, pop from the rear and then a whiff of burning. I had blown the rear markers and the function outputs of the decoder (or so I thought)

What I should have realized was that the current-limiting resistor was on the now-discarded main board. The only salvation was that the Lenz gold decoders are overload protected and the function outputs could be reset using a CV.

So as of this writing the rear markers have not been replaced, nor have I hooked up the front markers (they should be white instead of green anyway). I'll get around to both of these chores eventually.

Was the installation a success? Absolutely. The loco now has a speed of 2/3 smph at speed step 1; and I have not made any adjustments to the factory settings on the decoder. I could probably get it to go slower. I've previously posted video of this loco going through it's paces and you can find it here:

Lenz decoder Performance

Wednesday, September 8, 2010

Benchwork Experiment

This is my latest section of benchwork. I have been wanting to build some benchwork out of 3/4 inch plywood because I am tired of picking over the lumber pile, even the premium lumber pile, to find straight dimensional lumber. I've have been dragging my feet building new benchwork until I could try this experiment; now that I have, the experiment has had decidedly mixed results.

People have written about how they have had their lumber yard rip a sheet of 3/4 in. ply into 3 1/2 or 4 inch strips. Call around as I might, I could not get any yard to rip the sheet for me. So I picked up a sheet of ply at Home Depot (this could be my first mistake) and ripped it myself.

I do not have a table saw, so I set-up and ripped each strip using a circular saw and a saw guide. This resulted in some annoying dimensional differences between strips. Furthermore, each strip had two beveled, not square, edges where the hand-guided saw tilted beyond 90 degrees. Worse still, as each strip was cut free, it immediately curled!! WTF!!

This was not what I had expected. I stopped cutting after about half of the sheet was consumed. It produced a number of wide strips plus two 1 1/2 inch strips for leg braces. The legs of the benchwork would still be premium 2x4s.

I placed the strips on the basement floor for over a month in the hope of straightening them out - no luck, they were still curled, although not as much. The strips were every flexible, so I believed, rightly, that they could be built into a reasonably square structure; but their utility over dimensional lumber was fading rapidly.

The strips were very lightweight, which also did not bode well. When the basic frame structure was assembled it was light but not very rigid at all. When the legs were attached and the benchwork attached to the concrete basement wall at three points, the structure was easily distorted with little pressure. The fact that the rearmost longitudinal member is a two-piece affair to fit around the drain pipes that you see in the picture probably did not help. The benchwork was more than strong enough to hold up the railroad, but at this location that was not enough. This section of benchwork is to the left of the basement walkout door and has to be rigid enough to maintain the alignment for the lift-out section to span the doorway.

To salvage the section, I doubled up the rightmost two cross-members with 1x4 dimensional lumber to make it more rigid in the area where the lift-out section will attach. More leg cross-bracing was added to help stiffen this area, and the section will become more rigid when the benchwork sections to the left are added. These sections will be made out of conventional 1x4 lumber.

If this is not enough I may add some cross-bracing in the plane of the grid to add stiffenss by the door. If all else fails, I may be able to hang the subroadbed from heavy-duty metal brackets attached to the wall thereby making the alignment independent of the benchwork altogether.