The rest of the story
I had been on the hunt for a CF7 with the angled roof to join my two round roof versions for about 13 months. I eventually stumbled across a new old stock version at a dealer. Lucky me, I thought.
My latest conversion of several locos from ESU Select Direct to Tsunami 2 provided available decoders. Conversion of the lights on the new unit to SMD LEDs would be straight forward enough followed by soldering wires and downloading the proper sound file from ESU. A little weathering to match the existing CF7s in the roster and let the good times roll.
But then I saw a photo of a CF7 with a roof top rotary beacon and 3 chime horn. This was a relatively minor difference from the Athearn version and certainly worth the time and minor effort to achieve a distinct look. I had the Details West RB 106 rotary beacon and moving the ground plain antenna stand from above the headlights to behind the AC was no biggie. I also had the horn.
Decoder Install
My repurposed decoder was a Select Direct board style whereas the first two CF7s received the regular, and much shorter, Select. The upside was the Direct includes LED resistance built in. So a few solder joints later and we are on the programming track with the LokProgrammer working its magic.
But nothing works after the program files finished installing. Zip, nada. Long story short, I left the room as it takes about 30 minutes for the sound files to transfer but the loading was interrupted. Don’t know why or care. The programmer screen showed the reassigned road number, etc., but many of the other CV values were missing.
Nothing to do but reload the file while remaining in the room to make sure it completed.
Test Run on the Mainline.
A brief session on the driver screen verified file loading complete. Now to the mainline for a spin and tweaking.
What the heck??? Who applied the handbrake? This loco will not go above about 40 smph. And why is there an intermittent hesitation. And it wobbles!
As it is new, I set the throttle at speed step 25 and let it run around the main to loosen it up. After 30 or so minutes there was no improvement. It was still stiff as cactus juice in January so I put it back on the programmer to verify the CV3, 4, 5, and 6 settings. Same as the other CF7s.Rats.
A side by side comparison on the main with the other CF7s confirms the new one definitely comes from the other end of the speed gene pool. Must be the motor as my standard procedure is to disassemble the drive line and trucks, clean and lubricate.
New Motor
Based on the comments from the Overlord of Re-motoring Chris Palomarez on the MRH podcast, I searched eBay for a Kato motor and quickly found a new one. It is the same model as used on many of their SD locos. It arrives in days so in the interim I disassemble and uninstall everything again.
The Kato motor, while having a different approach to motor connections, etc., fits in the Athearn frame with no modifications needed. Like a brother from another mother.
Removing the Kato light board is simple as is trimming their motor leads and adding 30 gauge grey and orange leads. I used the larger Kato mounting screws as the Athearn versions were too small to secure the motor tightly in the frame.
After reinstalling the Select Direct and re-soldering everything, time for another mainline test. This motor knows how to creep in speed step 1 with the original CV2 – 6 settings and no BEMF tweaking. It is still a little slower in the mid and upper range but that may improve with break in. Compared with the Athearn motored CF7s, it looks like a speed table may be needed to allow them to run smoothly together. Or, I suck it up and buy 2 more Kato motors.
The wobbles and the stutter.
There is still an issue with intermittent stalling and body wobble so another partial disassembly follows. I am wary of repeatedly soldering and unsoldering a board type decoder so I leave everything wired and carefully remove the trucks to inspect the bolster area.
The loco frame has some roughness where the truck bolster rests so I try to smooth it with my smallest file. The truck bolsters’ contact pieces are rough and need light filing, too. I presume the uneven surfaces are responsible for both wobble and intermittent hesitation. But while attempting to manipulate everything into a position to file, I fatally flex the decoder board. I swear I heard it moan.
Not a problem as I have another Select Direct. So, everything is removed, unsoldered, motor, trucks, etc., but at least I can do the filing to both frame and truck bolster. A trial fit shows the body is stable. Now to reinstall everything again and test. Again.
Ran really well for about 6 inches before the power starts stuttering. All other locos are working fine so it is neither the NCE nor dirty track. Now the headlights do not work but the rotary beacon does. Body off again to inspect; all solder joints are good. But this is an older Athearn frame where each truck has only track feed for one of the rails. The other rail’s power feed is transmitted from a wire soldered to a clip which is screwed into the frame.
While I removed the clips and solder all the connections on the trucks, I could not get a good solder connection to the frame clip but I believed I had secured the screw well enough for a reliable connection. Nope, just a hint of movement was enough to cause the intermittent power stutter.
I scraped all the frame paint away in that area and tinned the bottom of the Athearn brass clip. I soldered a steel washer to it and reapplied the screw. Tight as a drum and no movement.
But in all this install, disconnect, re-solder, trying to force the screw tighter, and general thrashing, I damaged the second decoder. I console myself that It was a good thing in disguise as the Direct board is really too long to fit in the CF7 body. Especially when a RailMasters DSM-8 is installed in the rear of the hood.
Ordered an ESU Select, received it and installed it. Got so carried away with joy I forgot the Select does not have LED resistance built in. So after programing and loading the sound file, I put it on the mainline and pushed F8. Great sound. Nice movement at speed step 1. Then I turned on the lights and they looked good for about 2 seconds. Poof.
In yet another disassembly while I wait on more LEDs to arrive. But by this point, it is not maddening but funny.
Weathering
I should have known better but I began the weathering while I waited in the Litchfield Station light order. The body being off and all would make it easier. A valid presumption if the loco were not cursed.
I touched up the cab paint where I had removed and reinstalled the number board appendage but the ATSF yellow dried with a darker tone compared with the factory yellow. I have used this product before and it has always been a dead-on match.
I repainted the cab top in the new paint and it looks good and any minor differences will be masked by the weathering. Sure. I removed the handrails and began with my standard wash of India ink and Isopropyl using a wide, very soft, brush. Good for killing the shine and adding depth around seam details.
Wow, who would have guessed the new paint reacted differently to the ink solution? Oh what the heck, the powdered pigment will blend it in. Nope. The new painted grabbed the pigment and it all became not a dust color but a deep forest mud brown. And it was splotchy. Rats.
A coat of Dullcote was next. I was hoping it would break the evil spell and the cab would look good. Nope. But before I give up and strip the body, Why not do a isopropyl coat over the entire body? What is there to lose? It looked fairly good and evened out the blotches. Joy abounds in the garage. I used a small brush, just damp with the alcohol to thin out and blend the last vestiges of any blotches on the cab roof and adjacent areas. The curse has been lifted.
The Culmination
The lights arrived and were installed but I changed which LEDs I used. I went with the 1.8mm versions instead of the SMD versions to simplify the wiring.
For whatever reason, the loco is now speed matched to the older CF7s after switching to the Select from the Select Direct. The three locomotives run as one up to 25 speed steps. Good enough for me on my small railroad.
Motor and Lighting Conversion process
Lens Prep
The light lenses are .0625” dia. clear styrene rod, cut to length, and polished both ends to permit maximum light transmission. While the applied shrink tube is butted to the cab walls and prevents almost all stray light from making the cab glow, you can also paint the sides of the rods silver followed by a final coat of black paint. If some paint finds its way on to either end, clean with isopropyl or other plastic friendly solvent.
The headlight lens in this installation will be .750” long.
The rotary beacon requires only minor modification as the light globe is already cored. Use a .0625 dia. drill bit turned by hand to enlarge slightly the core diameter to accept the clear styrene rod. Cut the rod 7/16” long.
Test fit a length of .0625 shrink tube over the styrene before gluing the styrene rod to the beacon lens with Faller Expert. Use a .09375 dia. drill to create the path through the cab roof.
Body Prep
The design of the headlight housing and its attachment to the cab does not lend itself to an easy conversion to LEDs as they come from the factory. For simplicity and ease, remove the cab headlight/number board part. It is lightly glued to the front cab wall and comes off easily using an Xacto blade applied along the cab seam. This is also a good time to drill the windshield wiper mounting holes in the front cab wall.
The light openings and the wire pathway into the cab require filing to accommodate the new LED assemblies. Use a small rat tail file to increase each headlight opening just to the point where the .0625 dia. clear rod slips through easily. I also use an emery board to lightly polish the rod where it will go through the opening and then cut them .625” long. As before, polish both ends and paint.
The factory wire hole into the front cab wall requires more work as it has to allow both light rods and their shrink tube to bass through. To be on the safe side enlarge the hole to .1875” and make the opening flush with the cab interior roof wall.
Lights
Multiple types of LEDs can work, 0402 and 0603 SMD LEDs, and 1.8mm LEDs. Larger LEDs can be too large for HO from a space and brightness context. SMD wiring varies from magnet wire, to Micro Litz, to other sub-30-gauge varieties and are typically about 6 inches long. The 1.8mm LEDs have rigid leads about 1.1875-inch long.
Magnet wire is very pliable and comes with the ends already tinned (the common is slightly longer for easy ID). The micro litz is very thin multi strand wire with a clear jacket which can be a challenge to strip if you choose to shorten the leads. Whenever possible I use the leads as is and secure the excess out of the way with silicone, adhesive transfer tape, or bundling the wires with dental floss.
Light Assembly
There are several processes for securing SMD LEDs to styrene rod. Faller Expert cement can be used to join the LED to the rod. When the adhesive cures, cover the lens and LED with the shrink tube and apply heat to shrink. As the LEDs are tiny, and their leads long, stabilizing the joint while the cement cures can be a challenge.
An option is to cut the shrink tube slightly longer than the clear rod (maybe an extra .09375”) to allow the LED to recess inside completely. Make sure the LED rests flat against the inside roof wall and then heat the shrink to tighten capturing the LED and styrene rod.
Depending on the room available in the cab, if needed, a 1K resistor can also be wired in series on the common wire and also be captured in the shrink tube. Just add more tube length.
For the 1.8mm LEDs, the rounded tip of the bulb will fit inside the 1/16 shrink tube which is the process I used on this particular CF7. Put 2 small drops of AC 180 degrees apart to hold the LED in place while applying and heating the shrink tube. In this application, I used ¼ inch dia. shrink to capture both headlight LEDs.
On the rotary beacon, I used the 0402 SMD for space considerations. Insert the wire of the rotary beacon through light the cab roof hole from the outside and use a toothpick to apply a tiny drop of CA to the base to hold in place
Secure the light assemblies to the inside of the shell with acrylic silicone caulk. A small dime size dab works well when cured. It is also easily removed if a LED needs replaced. Never happened to me. No, especially on this locomotive.
The Select decoder wiring soldering is straight forward. Note, not all of the leads from the decoder are needed. In this instance, the brown leads are for the speaker. The rear pink and light green are not used but I did not cut them off but rather captured them with Kapton tape. The traditional blue (lighting common) and white are for the headlights. The dark green in this instance is Aux 1 (rotary beacon).
There is only one common lead and I need three. I created a mini harness which will feed the three lights in the cab. Think through your installation as to how you will connect the wires. For instance, will the soldering tip reach the wires without damaging any part of the shell? In this application, the answer is no. The wires must be attached to the LEDs before cementing the light housing to the cab. This means I need additional length for connections.
Know the implications before you proceed to eliminate reworking. But first and foremost, do not buy a loco assembled on sacred burial ground.
