Replacing the Drive Gear

The little train has worked flawlessly until a friend visited. The train ran around the track once and then stopped. The motor was running but the train wasn’t moving. I figured something slipped and it would be easy to fix. No dice.

I took the engine apart to get to the drive gearbox and found that the main drive shaft gear had cracked in half. So I went to the Bachmann site for parts to order the replacement only to find that they are out of stock. So I emailed them asking when they would be in stock. They didn’t seem to have much interest.

I then searched the internet and found that the gear commonly cracks and a third party called “NorthWest Short Line” at makes a replacement gear for $11.95. I ordered two since Shipping was $5.95 regardless of the quantity. I also figured that the gear was likely to break again.

QtyItem #DescriptionPrice EachTotal Price
22223-6#1 Bachmann 4-4-0 & 2-6-0 axle gear 24T$11.95$23.90

The gears arrived today and I took the gear box apart, took the drive shaft to the garage with the gear and used a rubber hammer and a small socket from a socket wrench set to tap the gear down onto the drive shaft. After getting it perfectly centered on the little grooves that gripped the gear to the shaft I put the drive gears back together. I then put a bunch of nylon gear grease on them and reassembled the train.

In the process of putting the train back together I found that the mechanism that simulates driving the valves that open and close for the steam pistons was mangled. I tried to straighten it back up but couldn’t do it without taking the entire thing apart more thoroughly. No-one will ever notice it as I never noticed it until it was apart. So I left it as was.

When I put the train back together it worked perfectly. I ran it around the track a couple of times and then cleaned up. My hands were covered in the electrically conductive grease that coats all of the parts that pick up electricity from the track. I have more of the grease and will dab it on another time.

New engine Arrives

Well, the long wait for the engine was somewhat disappointing.  I learned that you have to really embrace this as a hobby when you are buying these things.

The package arrived this morning but I had to work so I couldn’t play with it.  Mike seemed as excited as I was that it had finally arrived.  I worked till about 1pm then headed to the doctor.  But, I unpacked the engine before leaving for work.  I got home at about 4pm and set it on the track.  It fit nicely and so I turned on the power with both engines on the track thinking that they would go the same speed.  Well, they didn’t.  The old train started running, the new train started running, and then the new train stopped running.   The old train almost ran into the end of the new one before I could get back to the power supply to turn it off.

The new train engine was stuck on a curve and it wasn’t immediately obvious as to what was wrong.  I thought that maybe a wheel had come off the track.  So, I pushed it through the curve and started again.  It did great down the straight away, but then stopped on the next curve.  I took the train off the track and inspected the wheels.  There was a odd little attachment on the engine between the middle wheels.  The middle wheels on this one, just like the old engine, lack the raised edge that keeps the wheel on the track.  This is how the little engine can make it around a tight radius curve.  Only two wheels on each side have flanges and so only two wheels keep it on the track.  The front two wheels just pivot and the middle wheel floats.  So, these little pins force the middle of the train to stay on the track.  But, they also prevent the train from being able to go around a curve.

So, my first reaction was to take the plastic mount off that had the pins mounted on it and try to run the train.  But, when I put the train on the track, it ran till it got to a curve, and then the back wheels jumped off the track.  I ran the train slowly and watched very carefully to find that the middle wheels were hanging off the edge of the track during a curve, and when they went back onto the curve, the train would lift up and hop the back wheels off the track.  On the old engine, the middle wheels are exactly the same and are set just a little higher than the front and back wheels so they never catch on the edge of the rail.

Figuring that there had to be something wrong with the train, I looked at it carefully and noticed that the track for the front pistons was slightly bent.  That got me to thinking that maybe the mount for the front wheels was bent and it was time to consider sending this back.  I flipped the train over on its original packing and started inspecting it.  I applied a little pressure to the piston track and it straightened out.  I looked at the front wheels and found that both were on some kind of spring system but one did not bounce back so easily.  So, I took the little screws out and pulled the front wheel mount off.  Two small springs were in the wheel mount but one was not seated in its little hole.  I put the spring where it should be and put it all back together.  With the spring in place, I noticed that the front wheels now sit a little higher (when upside down) than the middle wheels and so this should account for keeping the middle wheels off the track.

While I had it apart, I looked at the drive mechanism.  I had been wondering why this engine would be $500 when the other was $200.  Both have a canister motor.  Well, all you have to do is turn them over and you see the difference.  The old engine had a canister motor with a spiral screw that drove a gear which in turn drove the wheels.  That made it really easy to bog down the old engine because the wheels were spinning at a speed similar to the motor.  Further, all of the wheels of the old engine were linked together with gears making it have incredible grip on the track.  The new engine has a drive train with a series of gears off the canister motor that run down a channel to only the two center wheels.  The center wheels in turn are linked to the front and back wheels with the standard steam engine linkage.  The gearing has the motor running much faster than the wheels.  On the negative, the new engine doesn’t grip the track so easily but it still has no problem pulling the train.  On the positive side, it is almost impossible to stall out the motor because of the gearing.  I find it really interesting that the real train linkage between the wheels is what drives the front and back wheels.  Another improvement in this engine over the old one is that not only do the wheels for the engine pick up power, but the wheels for the tender pick it up also.  So, a 2 pin wire connects the tender to the engine allowing the engine to get power from a total of 14 wheels.  The front two wheels of the engine are just fake wheels that ride on the track but do not pick up power or do anything.  They do not pivot as much as the old engine (this one is shorter) and so it is not as obvious that they are fake.  An added benefit to this engine is that it is shorter so I didn’t have to make any modifications to the nose to make it go through the mounts that are close to the curves.

So, back to getting the train to run on the track.  I put everything back together and even put on the little piece of plastic with the two pins.  I then took it downstairs and put it on the track and gave it double the throttle as before.  It powered down the track and then half way through a curve before stalling out.  I pushed it through the curve and it flew down to the next curve.  It hopped the back wheels off the track.  I put it back on track and it took off for the next curve.  When it hit the next curve, it kept going throughout the curve but made a sound like a card in the spokes of a bicycle tire and then stopped just short of finishing the curve.  I put it back on the track and it took off.  It never stopped again and just hummed along.  I let it go around the track a couple of times to see if the original pattern would return but it didn’t.

So, I couldn’t just let things be.  I had to take the engine off the track and see why it was now running just fine.  Much to my surprise, the only thing different was that the metal pins were gone and had popped out of the plastic mount.  But, the plastic mount was still there and was somehow keeping everything from hopping off the track.  I put the tender on the engine and let it run.  Then a car, then another car, and finally the last car.  It ran perfectly.  I started writing this blog entry with the training running when I heard a click and looked over to see the engine and the tender racing along the track without the rest of the train.  The coupling had popped loose.

So, I hooked the cars back to the tender and let it go.  About 3 rounds around the track and the coupler on the tender popped loose again.  Hum…I inspected the tender coupler and found that it was a standard Bachmann coupler but the mechanism that holds the coupler closed was plastic instead of the metal that I had on the previous cars.  My theory was that the plastic would vibrate loose and let the coupler disengage.  The metal ones were heavier and so were not as prone to vibration.  So, I took a coupler from one of the sets of wheels that I had removed from the original train cars and put it on the tender.  The result was at least 20 perfect runs of the train.

So, now my new engine and the train are running smoothly around the track.  After watching it for a while I noticed that this train looks smaller than the old engine, not only in length, but in height.  Then I looked them up and this one is 1:20 scale and the other was 1:29 scale.  Oddly, the cars that I bought for the train are 1:20 scale and so everything now matches but having the animals in the train be so much higher than the rest of the cars and right after the engine, it looks like the engine is rather small. Time for some new cars!

Fixed the cheap engine

I was so frustrated the the engine was so erratic on the track that I just had to take the thing apart and find out what was wrong.  When running the engine I noticed that there was an arcing sound that  you could only hear if the engine was directly over your head.  I found another broken wire to the front wheels that pick up power.  So, I put a better wire in place and hooked it up.  What was happening is that the large wheels on the engine pick up power and so do the little wheels in the front.  The larger wheels move slowly on the track and had some buildup on the wheels.  So, if you watched the train very closely you could see that when the wheel was turning, as it would get to the buildup, the train would pause and an arc would jump from the track to the wheel.  This was making weld spots on the wheels.

So, I repaired the broken wire so that the front pickups (the primary ones) are again making contact, and I cleaned the big wheels.  They are pitted from the arcing so they may be permanently damaged.  With the little wheels picking up and the big wheels cleaned, the train runs smoothly again.  The problem will return because the wires will eventually break again.  The design of the carriage for the front pickups causes way to much movement in the wires and they eventually just break from being flexed so much.  This is a bad design.  I hope the new engine has a better mechanism for the pickup wheels.

If I were going to keep this engine for a long time, I would replace the wheels on the tender car with wheels from Aristo that have pickups.  I did this for the performer car and the car runs much better.  So, I could replace the wheels on the tender and wire the tender to the engine pickup circuit so that it had a better way to get power.  I may still do that just to experiment after the new engine arrives.

New Engine is on its way!

I got an email from FedEx that a 15 pound package is on its way.  That is the right weight for the engine.  I then checked the web site where I ordered the engine and it shows that it shipped.  The package is set to arrive on December 4 so I am getting excited that I will get a good engine just as the cheaper one has failed.

Problems with the cheap engine

I was running the train and noticed that it seemed to be erratic in its speed.  I thought that the track might be dirty so I got out some alcohol and cleaned the track.  It got slightly better, but kept acting like the train wasn’t making a good connection on the track.  I took the engine off the track and inspected the pickups and found that the front four wheels that pick up power did not seem to be working.  Further inspection revealed that one of the large wheels had what looked like weld marks on the wheel.  Then, I found the problem.  I noticed a small red wire sticking out of the bottom of the 4 wheel cart on the front of the train.  The wire was clearly broken.

So, I opened up the train and found that the drive wheels also pick up power and that the wheel with the weld marks was picking up power in place of the front wheels.  The same marks are on one of the wheels of the 4 wheel cart that was broken.  Anyhow, I got out a little piece of multistranded wire and soldered it to the train.  When I put it back on the track, the train was working as it did before.

I still wonder if the original engine will ever arrive?

Cannon Ridge Railroad: Heartland Locomotive Works Passenger car arrives

I was running the train and noticed that it seemed to be erratic in its speed.  I thought that the track might be dirty so I got out some alcohol and cleaned the track.  It got slightly better, but kept acting like the train wasn’t making a good connection on the track.  I took the engine off the track and inspected the pickups and found that the front four wheels that pick up power did not seem to be working.  Further inspection revealed that one of the large wheels had what looked like weld marks on the wheel.  Then, I found the problem.  I noticed a small red wire sticking out of the bottom of the 4 wheel cart on the front of the train.  The wire was clearly broken.

So, I opened up the train and found that the drive wheels also pick up power and that the wheel with the weld marks was picking up power in place of the front wheels.  The same marks are on one of the wheels of the 4 wheel cart that was broken.  Anyhow, I got out a little piece of multistranded wire and soldered it to the train.  When I put it back on the track, the train was working as it did before. 

I still wonder if the original engine will ever arrive?

Heartland Locomotive Works Paseenger car arrives

The FedEx guy showed up today and delivered a box.  I had gotten an email that it was coming, but I thought it was another set of wheels to replace the trucks on the other animal car.  Anyhow, the car was the HLW passenger car that I had ordered a couple of weeks ago.  For the price ($80), I figured that it would have interior lights and a bit more detail, but it has nothing.  Additionally, it only comes with hook and loop coupler and the trucks do not match the design of the Bachmann or the Aristo cars so a replacement of the truck isn’t possible.  After some finagling, I did find that a coupler from a Bachmann truck could be made to fit over the peg and form a reasonable connector for the car.  Why these cars cost so much is beyond me as they don’t appear to be of any better construction than the Bachmann cars.  The Aristo cars are by far the best constructed of anything I have bought.

New Aristo trucks

The FedEx guy came today and left a package for me.  It contained the Aristo trucks that I had ordered a week ago.  These trucks are far better built than the ones from Bachmann as they have true bearings to hold the wheels.  I can put a truck on the track and give it a hard shove and it will go completely around the entire room.  The hubs on the outside of the truck turn with the wheels giving them a good look of realism.  And, the wheels are mounted in authentic spring loaded frames so that they are true replicas of a real train truck.

The real reason that I bought the Aristo trucks was in hopes of replacing the trucks on the performer car.  The performer car has electric pickups that are simply metal brushes that drag on the outer diameter of the wheels.  This makes a whistling noise all the time and is quite annoying.  It also puts a lot of drag on the engine.  So, the new trucks have electric pickups like the tender car that are in the races for the bearings and so you get a good frictionless wheel that still picks up power.  I had no problem replacing the trucks as they fit the car mounts perfectly.  But, the Aristo knuckle does not fit smoothly with the Bachmann knuckle.

The second set of trucks that I bought have short arms on them for the knuckle and they sit about 1/8 inch lower than the Bachmann truck.  The result is that the arm on the truck does not clear a part of the Bachmann train car edge.  So, out came the dremel tool and away went some plastic.  After that little bit of work, I was able to put the truck on the yellow giraffe flatbed car.  I only put one truck on so that the car made the transition between Aristo and Bachmann knuckles.  This won’t do for long as the Bachmann trucks are silver colored while the Aristo ones are black.  The car looks off balance.  I’ll fix that later.

With this small change, there is considerably less noise on the track and the engine doesn’t bog down in the curves at low speeds.  So, overall, it was a good move….expensive at $25 for a pair of trucks, but a good move.

Dampening the sound upstairs

I was laying in the bed nursing my back when an idea hit me for reducing the noise of the track. When cleaning out the garage last weekend, I threw away an old inner tube for a bicycle.  I don’t know why I kept it around, but I thought that maybe that tube would provide enough rubber to reduce the transmission of sound through the joists in the ceiling. So, I got up and went to the garage to root through the trash for the tube.  It was gone.  Mike had had one of those moments when he isn’t lazy and he took the trash to his office to put in the dumpster.  He usually only does this with stinky trash, but I guess he was trying to clean up since his parents were visiting.

But, while rooting through the remaining stuff to be thrown away, I found an old doggy door in our pile to toss. This would be better than the inner tube in that it is 1/8 inch thick and made of a very soft rubber.  It had been discarded by Mike because it was ripped on one side.  We should have thrown it away, but he saved it for some reason.  Now I know why…it pays to be a pack-rat.

Anyhow, I took the flap and cleaned it up in the sink, and then cut it into 1/2 inch strips.  I unscrewed the track from the ceiling mounts and slipped the rubber under the track on half of the track.  I only did half the track and then went upstairs to hear how it sounds. I couldn’t tell the difference between when the train was on one side of the room from the other.  So, my choice was to either take out the ones I had put in or finish the job.  It occurred to me that maybe the sound would be transmitted by the ones without the rubber spacers even when the train is on the other side.  Kind of like putting your ear to a real train track to hear of the train is coming.  So, I went down and finished the job.

When all was said and done, I couldn’t tell a difference in the sound upstairs from before and after the rubber spacers were put in.  But, from downstairs the train is considerably quieter.  So, I left them in.

Now that I have tried everything I can think of to make it quieter, I am ready to put the mushroom caps in the holes to cover the screws.  The first batch of caps I had done were difficult and I had only done 12 of each size.  My first method was to hold onto the cap by the back side with a pair of scissors, and dip the head of the cap into the polyurethane and then set it upright on the work bench.  This took forever and I found that I couldn’t grip the caps very while after a short while.  So, this time I got a pie pan, poured some polyurethane in it to about 1/8 inch depth, and then dumped the caps into the polyurethane.  I was going to turn them over so that they were all head down, but I noticed that the caps were drawing the polyurethane through the pores of the wood and becoming wet on the side that wasn’t in the polyurethane.  I sloshed them around and then lifted an edge of the pan to get them completely submerged.  I then used the scissor tips to pull them out one-by-one and put them onto a piece of aluminum foil on my work bench.  I did the same with both the large 3/8 caps and the tiny 1/4 caps.  The caps formed puddles on the aluminum foil so I got worried that they would stick or worse…they would not fit into the holes.  So, I changed my method so that I rubbed the bottom of the cap on an old T-shirt before setting on the aluminum foil.  This worked well and since the caps had absorbed so much polyurethane, they remained shiny with just one dipping.  I will have to let them sit for a few days to make sure that they are completely dry before I try to put them into the holes.

Final Costs

In an earlier posting, I estimated the cost per linear foot of the track to be around $9. But, this was a rough estimate based on the cost of only the larger components of the track and on 100% utilization of components. In reality, there is a bit of waste on the metal track for the curves and the estimate did not factor in the costs of the track spikes or the screws needed to hold the track in place.

To get a realistic cost per linear foot of track, I added up the total cost of all consumable materials and divided by the number of feet. The table (below) shows all of the consumable costs needed to make my 81 feet of live track. This cost includes the costs needed for the prototypes and any waste material that was left over. The exception is that it does not include the excess metal track that was purchased due to poor estimation of the materials required for the project.

So, the final cost per linear foot is $15.87. This is still significantly cheaper than the $125 per linear foot for the Loco Boose track.

Consumable materials needed for track

Lowe’s SKU Description Count Unit $ Total $ Tax $ Gross $
1039 1X2X8 Red Oak Board 23 7.57 174.11 8.71 182.82
1175 1/4X2X4 Red Oak Scant 46 2.66 122.36 6.12 128.48
2234 4X3/4 PH FL Wood Screw 2 2.57 5.14 0.26 5.40
9362 1/4X2X2 Red Oak Scant 4 1.33 5.32 0.27 5.59
10777 1X2X6 Poplar Board 8 3.98 31.84 1.59 33.43
16422 3/4″x18MM Oak Cabinet Grade Plywood 1 44.88 44.88 2.24 47.12
19379 5/16″X48″ Poplar Round Dowel 14 0.68 9.52 0.48 10.00
19416 1/4″X36″ Red Oak Round Dowel 1 0.88 0.88 0.04 0.92
19437 3/8″ Oak Wood Button (Mushroom Top) 132 0.17 21.78 1.09 22.87
—— 3/8″ Oak Wood Button (Mushroom Top) – Mailorder 30 0.10 3.00 3.00
19444 5/16 Oak Round Dowel 64 0.88 56.32 2.82 59.14
27594 3/8″X36″ Poplar Square Dowel 105 0.88 92.40 4.62 97.02
35880 3.25 Oz PreBond Woodfill 1 2.98 2.98 0.15 3.13
45862 Minwax 1 Qt Satin Polyurethane 1 8.87 8.87 0.44 9.31
46575 Minwax 1 Qt Wood Conditioner 1 9.98 9.98 0.50 10.48
52068 2X3/8 PFH BRS WD Screw 2 0.98 1.96 0.10 2.06
86091 8 Oz TiteBond Orig Wood Glue 1 2.27 2.27 0.11 2.38
86092 16 Oz TiteBond Orig Wood Glue 1 3.98 3.98 0.20 4.18
86165 1 QT Dark Walnut Wood Finish 1 5.87 5.87 0.29 6.16
225794 2″ PGP Star Drive Screws 2 8.48 16.96 0.85 17.81
—— 0.05×3.5x.5 Stainless Steel custom plates w/4 holes 110 2.31 254.10 254.10
—— #4×1/4 Pan Head Steel Square Drive screws 400 0.03 13.92 13.92
—— Lagas Code 250 Nickel-Silver rail joiners 40 0.65 26.12 26.12
—— Lagas 1/2″ large spikes 1000 0.02 20.90 20.90
—— Lagas Code 250 Nickel-Silver 6′ rails 32 9.98 319.36 319.36
—— 1/4 Oak Wood Button (Mushroom Top) 100 0.08 8.00 8.00

 Total 1,285.69
Total Linear Feet of Track 81  15.87 1,285.69

Special Tools Needed for Project

Lowe’s SKU Description Count Unit $ Total $ Tax $ Gross $
Milwaulkee Table Saw 1
Dewalt Hand Drill 1
Saw Horse 2
Riobi 3/8 HP router 1
Carpetendar’s Square 1
127093 Skil Delux Router Table 1 69.00 69.00 3.45 72.45
171028 3/8″ Auger Bit 1 10.63 10.63 0.53 11.16
177059 1/16″ Titanium Drill Bit 1 2.96 2.96 0.15 3.11
195435 ALI 4×36 Sanding Belt, 120 Grit 2 5.86 11.72 0.59 12.31
195714 ALI 6″ DA Disc 220 Grit 1 11.29 11.29 0.56 11.85
204419 Skil CT 3/4X25/32 Straight Double Flute router bit 1 8.98 8.98 0.45 9.43
204422 Skil CT 3/8X25/32 Straight Double Flute router bit 1 7.98 7.98 0.40 8.38
204588 Skil 1/8″ Roundover 1/2″BB Double Flute router bit 1 12.98 12.98 0.65 13.63
206527 QT Turpentine 1 4.98 4.98 0.25 5.23
209208 Dewalt 1/4″ Brad 1 5.48 5.48 0.27 5.75
209210 Dewalt 5/16″ Brad 1 6.48 6.48 0.32 6.81
219285 1″ Metal Spring Clamp 16 1.98 31.68 1.58 33.26
227313 Bosch Carbide Tipped 5/16 X 1 Straight Double Flute router bit 1 13.97 13.97 0.70 14.67
241624 4″ Hand Clamp 6 6.96 41.76 2.09 43.85
243599 Firestorm 10″ 200T Saw Blade 1 12.97 12.97 0.65 13.62
251929 Task Force 4″X6″ Belt Disc Sander 1 99.00 99.00 4.95   103.95
251932 Task Force 10″ Bench Drill Press with Laser and Light 1 99.00 99.00 4.95   103.95
—— #4 Square driver 1 1.20 1.20 1.20
—— Lagas Code 250 Track Gauge 1 16.10 16.10 16.10
—— Lindsay Code 250 Rail Bender 1 80.70 80.70 80.70



At this point, the track is mostly finished.  All that is left for the track is the installation of the plugs for the holes that mount the track to the ceiling.  So, I started putting materials away and started cleaning up the pool table and the bar.  I want to be able to show people the various prototypes used in creating the track design, so I took the prototypes and mounted them on the wall near the TV.  I mounted the two ceiling mounts under the bar and put the final prototype on the mounts and put a car on the track.  I then took the leftover ties and glued them to a leftover piece of track wood with the intention that I will use it to display how the final track was assembled.

Steel Plates Installed

The steel screws arrived as I was leaving for work.  When I got home, I fed the dogs, then went to the basement and installed one of the plates to see how the screws would look.  They match in color perfectly and the square hole for the drill bit gives it a slightly odd look that works well with the track.  The color of the metal track, screws, and plates looks good together and the metal plates hold the track firmly.


I installed all of the plates on the inside of the track (facing the interior of the room) while listening to Transformers and then released the clamps that I had in place to keep the wood from moving.  As I released the clamps, the wood flexed slightly and set into place.  I ran the train for about an hour and noticed that the track slowly adjusted to the position of the wood.  In a couple of places, the track pulled slightly away at the joints leaving a 1/16 gap.  This happened just before each curve.

On Friday night, I watched the extra materials on the Transformers DVD and installed the plates on the outside of the track.  This was much more difficult because I could not use a normal driver to put the screws in and I couldn’t use the drill to tap the holes for the plates.  So, I took the little square driver bit that was for the #4 size screws, and a small wrench that could turn the driver bit.  With a lot of effort, I pushed with my index finger on my left hand while turning the bit with my right hand to start the screw into the wood.  After losing a bit of skin on the fingers, I was able to get all of the screws started into the track.  I then went back and used the wrench to finish driving the screws into the wood and locking the plates in place.

Temporary Engine Arrives

The engine that I ordered last week arrived today at work and I put it on the track as soon as I had fed the dogs when I got home.  The engine looked great.  I first powered up the track with a very low voltage until the engine started moving.  I let the engine run around the track backwards to watch how it behaved and ensure that there were no problems on the track.  All looked good.

But, I changed direction through the switch on the front of the engine and let it run forward at the same slow speed.  The engine did fine until it hit a curve and it stopped dead.  I checked what was going on and found that the front two wheels of the train not only pivot, but they also shift from right to left.  This allows the long engine to traverse curves without having to have any pivots in the drive train. The situation is difficult to see unless you look very closely at the way the front two wheels move with relation to the body of the train.


Unfortunately, this is not a situation that I had expected or planned for in the ceiling mounts.  All would have been fine except the front of the train had ears that supported a metal bar and the ears were catching very slightly but enough to stop the engine on the low speed. I have highlighted the ears in the following picture.


So, I took the nose of the train off the engine (it is leaning on my keyboard in the above picture) by removing two screws and then I got out the dremel tool, mounted a cutting blade, and cut 1/4 inch out of the plastic ears.  The two little metal feet that are hanging down below the circles are what I removed.  After mounting the cow catcher back onto the engine with 1/4 inch of the ears missing, the engine has no problem running around the track.  Notice that in the picture below that the bar that was on top of the cow catcher is gone.

So, with the cow catcher adjusted and the ears shorter, the train is able to move around the track with ease.  The next step was to start running the train a bit faster to see how it did at a decent speed.  Initially, the engine moved with variations in its speed, but those soon went away.  But, one thing I noticed was that as the engine went through the S-Curve, there would be a small clicking sound but the engine did just fine.  After watching it a couple of times, it was clear that the front two wheels that are floating with a very light spring were coming off the track and then popping back on.  I put the gage on the track and found that the track was not matching the gage any longer.  I tapped it back into position and the problem still occurred.  But, then I realized that what was happening was that the S-Curve was causing the wheels to shift across from one side of the engine to the other making a complete shift in a very short area of track.  There is a plastic seam in the middle of the slot that the wheels shift in and the seam was causing the wheels to catch.  After about 10 runs of the train around the track, I noticed that the wheels stopped clicking through the S-Curve.

So, now that I have the train moving smoothly on the track, I added the tender car to the engine.  The tender has two wires that attach to the engine.  One pair of wires has a label “light” on it, and the other pair says “sound”.  When the train goes forward, the light on the front of the train is on.  When the train goes backwards, the light on the tender comes on.  The tender has a slot where a 9-volt battery goes in and if you put a battery in, the tender gets a trigger from the engine every time the valve changes direction on the piston of the mechanism.

So, now the engine and tender have been tested and I can add cars.  I put the two animal cars on the train and the little caboose, but the caboose made a ton of noise with its metal wheel mounts.  So, I put the performer car on with the ringmaster logo to the back so that it lit from behind.  The engine was sluggish at first but after giving it a bit more power, the engine was able to pull the train without issue.  It slowed down in the curves, but otherwise did just fine.  I adjusted the pressure on the pickups on the wheels to free them a bit, and the train did much better.  Here is a picture of the entire train over the fireplace.


And below is a picture of the train pulling through the S-Curve.


Another view from under the 5th curve looking at the section that goes over the pool table.


And, finally a shot of the train pulling into curve number 5.


So, now my train is running and there are no problems.  I went upstairs to get something from my desk and left the train running.  When I came out of my office, I heard the sound of either a small jet or a large truck in the driveway but then realized that the sound was coming from the floor.  That is when I realized that this train makes a lot more noise upstairs than it does downstairs.  Downstairs you hear the metal wheels on the metal track, but upstairs, you hear a roar as the floor not only transmits the sound, but it seems to amplify it.  The track is mounted to the joists between the ceiling of the basement and the floor of the main level.  So, after discovering this effect, I realize that I won’t be running this train when people are sleeping.

Well, the big concern left is that the real engine, which is more than double the cost of this one, will have the same issue with clearance.  I don’t mind cutting on this cheaper engine, but I hope not to do the same to the expensive one!

Test connecting plate

I received the connecting plates today and immediately ran to the basement to test one out.  I am extremely pleased with both the color and proportion of the plate with the track.  Now, I have to decide how I am going to attach the plates.  I originally had planned to use brass plates so I already had a handful of brass screws ready.  So, I went ahead and put a plate up with brass screws.

I rolled the train car in place so that it would be easy to see the various colors of the track, wheels, trucks, and plates all together.  I don’t like the brass screws on the silver plate.  It makes it look to fancy.  So, I will search online and at the local hardware stores for screws that have silver heads.  The ideal screw would be one that has a square head like the top of a railroad tie.  But, I doubt I will find such a thing locally.

In any case, I am very pleased with the plates.  I really appreciate ID Plate’s ability to let me draw the plate in Visio and send them a PDF which they matched to create the product.  And, the price was not to bad.

Bought an alternate engine

Well, I just couldn’t stand it any more and so I called the place that I bought the caboose from Discount Trains Online and they had the Bachmann BM81094 in stock and could ship it out today.  Their price was the same as so I bought it.  I figure that this does two things.  First, it is like washing your car to get it to rain.  Now that I have paid for another engine, will probably find my engine and ship it out so that they both arrive at the same time.  Second, it gives me an alternative engine should anything go wrong with the one that I have ordered.  They have it in stock, but chances of it arriving this week are slim to none.  It will probably arrive next Tuesday.

When I got home, I also put some of the plugs in the ceiling mounts to see how they would look. They certainly finish off the look of the mounts.

Track completed

After feeding the dogs, I headed down to the basement and started working on the remaining 3 sections of track.  I started at 7pm, and finished up at 9:38pm.  I started with the straight section over the fireplace, then the 8th curve, and finally the little short pickup section.  I saved the short section to let me have a piece to work on that would be easy to manipulate and measure for getting the track correct without gap.  As it was, I ended up with a 1/16 gap on the last section and decided to leave it that way since it is between straight sections.

I rolled the performer car over the final curve and tested connectivity.

The other three cars were on the left side of the fireplace while the performer car was on the right long section.  After completing all of the sections, I rolled the cars together and had them meet on the 8th section.  So, as of 9:38pm on October 8, 2007, the physical track is complete.  I still have finishing touches to do, like putting in the plugs on the ceiling mounts and touching up the stain on the ends of some of the ties that got scratched during assembly.  I have 8 sections of nickel-silver track left over (about $88 worth), two 1x2x6 Oak Boards, and one track connector.

I also got an email from FedEx that a package had been entered to ship to me.  I got my hopes up thinking that maybe it was the engine, but the package was shipped from Texas.  Texas is where the steel connector plates were made and so I think it is the steel plates that are on their way.

Power continuity test

Today started the morning by putting the final coat of polyurethane on the remaining sections that I coated when I got home from DC last night.  I turned the pieces upside down in their final locations and then started tearing down my work area.  Now that I have completed all of the physical construction, I have no risk that I will need the router or any of my power tools.  So, I dismounted my router from the routing table, took all of the screws out of my workbench and broke the steadying board from the back of the workbench.  All of the tools went into one of the rooms in the basement and the board for the workbench was slid behind the dumbell rack in the server/weight room.  The board will be beneficial if we foster a dog that needs to be caged during the day because the previous dogs have torn up the insulation on the wall.  In the event that I decide to build another track in the middle area of the basement, I will have my jig for the curves ready.  And, should they ever lift the burn ban, this board will be very interesting to burn. After cleaning up the work area, I vacuumed and put the original furniture back in place.  It looks as if nothing ever happened there.

The next thing I did was to try to find some alligator clips to connect the power supply to the track so that I could test whether I had any issues with continuity.  I couldn’t find any clips, but I did find that if I simply leveraged the end of the wire between the track and a tie and could then test that I had power to every section.  Although I do not have an engine to test with, I do have the performer car which has lights so I can push the car around and make sure that the lights never go out.  This also lets me test to make sure that the joints between the sections are smooth and make little noise.  So, once I got everything connected, I was happy to find that I have good connectivity to every section that is currently up.

After testing for power, I hunkered down in the weight room and knelt on the rubber pads while I worked on the track that was on the concrete.  I started with what I consider to be the most difficult which is the curved section nearest the weight room door.  I got it done, and then rolled the performer car onto the section and validated that I still had continuity.  A picture is below showing the car with the lights with the camera in the corner under curve number 7.

After finishing curve number 6, I started working on the opposite side of the track near the fireplace.  The little 3 foot section after the S-Curve was the first section to be completed and then I built the 7th curve.  Below is a picture of the 7th curve with the animal cars rolled into place.

I felt like I was done for the day at about 6pm after the dogs reminded me that I needed to feed them.  After an hour of sitting around, I decided that I should work on another section.  So, I started on the long 8 foot section at about 7pm and by 8:30 I had it completed and in place.  Once again, I powered up the track and rolled the performer car onto the new section and then took a picture.

So, I am calling it quits for the day.  Sure, I could do the 5 foot section, but I don’t have an engine and I won’t have one for quite some time, so why rush.

Last of the ties are mounted

I am going into town for dinner tonight at about 5pm so I am pretty much done for the day.  I started at about 9am and worked pretty much constantly till about 3:30pm.  In this time I put the metal rails on the curve in front of the stairs and the straight section in front of the stairs.  I then checked the newly stained ties and a few were still tacky to the touch, but they were dry enough to slide on the metal of the router.  So, I got out my jig and router table, and routed the slots out of all of the new ties.  They were sticky enough that they picked up some sawdust, but this came off easily with the buffing pad for the between-coat roughing.  The nice thing is that the sticky part of the ties came off but the color stayed.  They had dried for more than 12 hours so I am pretty sure that the stain has penetrated as much as it is going to penetrate.

After routing out the slots, I took the two remaining sections (the one in front of the fireplace and the short section to the right of the fireplace) and glued the new ties to them.  As of 3:30pm, they were under the weights and pressure, and so they should be ready to be coated with varnish by the time I get home tonight.

When I got home at about midnight, I went downstairs to put on a coat of polyurethane on the last two sections of track that had the ties glued earlier.  I got the coat completed by 1am and headed to bed.  I will put another coat on in the morning and they will be ready by the time I get to them with the metal rail.

Ugh…no engine, but nicer people

Well, I got an email from saying that they cannot find the engine I wanted through any of their suppliers.  So, I have asked if they can get BAC81088 or BAC81097.

From: WST []
Sent: Saturday, October 06, 2007 3:45 PM
Subject: RE: Order 200715924 status?

The wire is on backorder and the BAC81493 I can not find at any of my
suppliers at this time

Sent: Saturday, October 06, 2007 4:13 PM
To: ‘WST’
Subject: RE: Order 200715924 status?

Thank you for the update. Go ahead and cancel the rest of the order and I will find another engine for the train.

Can you check to see if you can get BAC81088 (it is a backorder item also)? If not, how about BAC81097?

Thank you,

From: WST []
Sent: Saturday, October 06, 2007 4:25 PM
Subject: RE: Order 200715924 status?

I did cancel that and I can get BAC81088 on 10/31


Sent: Saturday, October 06, 2007 4:47 PM
To: WST []
Subject: RE: Order 200715924 status?

Thanks for your help! I called up your 888 number to find out what you had in stock and talked to someone that asked me to hold tight and that in about 5 days he would have more information on whether it is in the order coming at the end of the month.

The sense of urgency here is that I hand built an elevated track and am itching to put an engine on it. At the moment, I am just pushing the cars around by hand.

Can you “uncancel” the order for BAC81493 until he finds out more information?


So, I called them on the phone and happened to get the guy that orders the stuff.  He said that they can’t cancel an order for something like this without him being involved because these expensive engines don’t move very quickly.  He said that he had an order coming in at the end of the month that might have it in the box and to hold tight for about 5 days for him to find out the status.  He said that everything was delayed about 15 to 20 days due to the issues with lead paint from China and that it was probably sitting on a dock just waiting to be cleared.

So, I will call back in 5 days and see if they know anything new. not so good with customer support

OK, so I am a bit of a sensitive person when it comes to relations with other people.  Partly this is because I am from the south where we bend over backwards to try to help a customer, but mostly because I have run customer support centers for various companies for more than 20 years. Well, today I simply wanted to know where the engine is for my little train, and I ended up feeling like I will never order anything from again.  Here is the series of emails:

Sent: Friday, October 05, 2007 1:57 PM
Subject: Order 200715924 status?

Could you check on order 200715924 to see if you have an estimate for the engine? This is my first G train and so I have cars and track with no engine. 🙂



From: WST []
Sent: Friday, October 05, 2007 5:24 PM
Subject: RE: Order 200715924 status?

An estimate for what? I can tell you that, that item is on back order.

Sent: Friday, October 05, 2007 5:39 PM
To: WST []
Subject: RE: Order 200715924 status?

I did not intend to offend anyone with the request. Since I referenced the
order number “200715924” and referenced the engine, I thought I was asking
for an estimate related to the order “200715924” which would seem to only be
the estimate for when the order might be shipped. The original order was


From: WST []
Sent: Friday, October 05, 2007 5:54 PM
Subject: RE: Order 200715924 status?

Im sorry, I guess I was looking for a less vague request but I see what your
asking now. Your item CRE55453 is on back order which means WE are still
waiting to recieve it. You will recieve a confirmation when we recieve/ship
it. Thanks

Sent: Friday, October 05, 2007 5:59 PM
To: WST []
Subject: RE: Order 200715924 status?


I was meaning both BAC81493 and CRE55453. If CRE55453 is holding it up, I can go down to Lowes and get some wire. It is the engine that I am most interested in. I have been building the track and have the cars, but do not have an engine.

I first instinct is to see if you have another engine that is comparable that I can get sooner, but I think that would leave you with an engine from your supplier that you have no customer for if you have the engine on order from your supplier.


So, I admit that my original message could have been more clear by explicitly asking them to check for an estimated delivery time for item BAC81493 on order 200715924, but I still think that the original message conveyed that request and did not warrant the response. It was the initial response that really turned on the sensitivity for me. Did the support person bother to read the message? Did the subject line not convey enough information that I was looking for the status of an order?

But, I figured that maybe this support person had just had a bad day and that I should apologize and explain my message. But, I cannot figure out if the response is just clarifying why they responded as they did or if they are being smug.

Now, in my support centers, I look for these kind of short answers back to customers because the brevity in the response immediately conveys a lack of patience with the customer. But, in the end, my question was never answered as to if there was any sort of timeframe for me to get the engine. And, the result is that I really want to simply cancel the order and buy from some place else even if it is more expensive. One thing for sure is that I will not do business nor recommend to anyone. I realize that a wholesale place is basically able to undercut everyone by cutting everything to a minimum, but if you are going to respond to a customer support request, at least do it so that you do not offend your customer.

More ties have been stained

I got the miter saw out and cut the dowel rods for the ties, peeled off the stupid Lowes SKU stickers, put the gummed up sanding belt on the sander and took off the glue from the end pieces, and then stained the new ties.  While I had the sander out, I put the good belt back on and sanded down the last two sections that are to receive ties.  It was dark by the time I was half way through the ties but I finished them under the porch light. Hopefully it will not rain tonight and they can dry outside.  I will pull them in before I go to bed.

That’s it.  I have lost a lot of steam in the project because I don’t have an engine and so putting it all together is kinda useless other than for it to be a very elaborate display of a few train cars.  As you can tell, I am pretty frustrated with waiting 6 weeks for an engine and still having no idea as to when it will arrive. 

Wood Rail completed, short by 50 ties

In order to complete the measurements for the straight sections and the curves, I needed to hang the rest of the ceiling mounts.  I took this picture just after putting up all of the mounts and after starting the 7th curve.  You can see the 7th curve hanging on the mounts without ties.  You can notice from this picture that there are no mounts for the curve to the right of the fireplace as I had not started on that curve and could not put the hangers in place until I had cut the wood to length.


My method of putting up the mounts is to put the end mounts in place where the straight sections will end, put temporary straight sections in place, and then put the raw curve section in place over the ends of the straight section.  I then move the curve around so that I cut nothing off one end (the end that is straight from the jig) and catch the other end of the curve at the peak of the curve so that I get a perfect fit with a 90 degree curve.  I mark the wood with a pencil on the bottom, and then cut it with a saw erring on the side of making the curve to long.  I check fit afterwards and trim it down a very small amount until I get a perfect fit.

After putting up all of the ceiling mounts, I got a little goofey with the camera and decided to set it on the track and take a picture down the length of the track.  This shot is looking down the track toward the bathroom from just over the pool table.  After taking this shot, it dawned on me that it would be an interesting touch to put a camera on the engine and show a live feed of what the engine sees through the monitor of the computer in the basement.


Well, after taking a picture on the rails, I decided to take a picture from the unfinished section of track to the left of the fireplace before I put ties on it.


After getting the curves measured, I put the ties on the new pieces.  For the short straight piece at the end of the S-Curve, it turned out to be perfectly 3 feet in length and so I just had to cut a piece in half to get the stock for the section.  After cutting and making sure everything fit correctly, I put them on the workbench and glued ties to the pieces.


While the new sections were curing in the press, I got out the polyurethane and finished off the 5th and 6th curve as well as the long straight section.  Well, it dawned on me that the long straight section might be better served as two sections since the 8 foot section didn’t fit the entire wall.  But, after measuring it, it turned out that the length needed is slightly over 13 feet.  So, with the 8 foot section, I will need a 5 foot section.  If I used 6 foot sections, I would have a little short patch section needed just as I had to do for the bathroom.

After letting the sections set for a while so that the polyurethane was completely dry, I put the pieces in place.


Below is a straight on view of the new sections with ties and polyurethane.  Note that there is no track on these sections yet.  All of these pieces are simply laying on the mounts and the mounts are not tightened to the ceiling yet.  Once I put rail on the sections I will be able to tighten everything up.


Finally, after getting the 7th and 8th curve completed, I started cutting the remaining straight sections.  One long 5 foot section goes over the fireplace, and a short 3 foot section goes to the right of the fireplace.  After getting the pieces completed and cut to length, I found that I did not have sufficient ties to complete the work.  So, I am at a standstill till I get more ties made.


Well, since I cannot do anything else, the final thing I did was to cut the metal rail to length and mount both the six foot section and the short 1 foot patch section near the bathroom.  I then ran the cars around the track to confirm that the rail was smooth.


Tomorrow I will go to the store and get more materials for the ties.  I will cut them to length this weekend and get them in place.  With this done, I will have the track finished and will just need to install the rails.  Unfortunately, I am finishing the track to quickly and I doubt that I will have an engine for the train any time soon.  It has surprised me that the project has started moving so quickly once I got the prototypes completed and the final design and methods completed for the system.

Curving around the Stairs

Before I left for work, I glued the ties on the 8 foot section for the long wall and put them in the clamps and press to cure while I was at work.

I had taken a section of the prototype track to work to show people what I was building and I brought it home today. I took the metal rails off of the prototype section and put it against the short section that will be cut for the little patch near the bathroom door to complete that area of the track. The difference is that the prototype section was built with the ties that I dipped in stain and then used a cloth to wipe the excess stain off. The ties that I am using in production are simply dipped and let dry. Another thing to note is that the stain that I bought is both a stain and finish so the stained ties take on a bit of a shine. This is good because it saves me from putting polyurethane on the ties as well as the wood rails.


I started the 5th curve which is the one near the stairs from the basement to the main level. The first step is to figure out places for the ceiling mounts that line up with joists in the ceiling. The second step is to measure the piece to find the amount to cut off from the curve. Remember that I made the curves so that they went beyond 90 degrees to give me room to adjust for sections meeting up without being at perfect 90 degrees. So, I put up the straight section near the stairs and bathroom, then the straight section that is directly in front of the stairs. Finally, I put the curve in place and marked where I needed to cut on the end to connect things up without any sharp irregularities in the track. This may seem complicated, but I cheated by setting the curve in place and then adjusting the straight sections to line up with the curve to make clean transitions. This creates a gap for the straight sections but in both cases, the section already has a gap that will need to be filled. After cutting the curve, I glued the ties on the curve and used one of the other curves to clamp and press the ties to the rail. This forces the glue into the wood so that it is strong. Below is a picture of the curve with the ties in the clamp while in place.


Since the wall to the stairs is slightly longer than 6 feet, I already know that I will have a gap between the end of the curve at the bathroom and the straight section. So, either I have a gap near the bathroom or near the entrance to the stairs. Since the entrance to the stairs will be more visible, I put the patch area nearest to the bathroom.


To create the patch to fill in the gap to the section near the bathroom, I used the scrap that was cut off of the section to fill the gap on the long beam nearest that was completed this past weekend. I put the section on the gap and marked it. Unfortunately, from either end, the place to cut ends directly over a tie that is already glued in place. So, I cut the section and split the tie in half with the saw blade. I was quite surprised at how difficult it was to break the tie from the beam even after being cut in half. After cutting the beam, I glued a tie on the end. This will result in a tie that is not equally spaced with the rest, but it will not be as noticeable as a large gap.


In addition to the patch section, the above picture also shows the 6th curve in the press. The 6th curve was one of my first sections that I put together and the spacing between the rails is not as good as the later ones. So, I had to use the router to cut a little off of the gaps in the ties to accommodate the difference. I also ran out of ties that I had routed out the center section for the cross members so while I had the router hooked up, I routed about 12 more ties for the curves.

Finally, while the 6th curve was in the press, I put the ceiling mounts up for the long wall near the weight room. I then took the 8 foot section that had set all day with the new ties and put the section in place on the new mounts.


Tomorrow I will put a coat of polyurethane on the new curves and the long section and hopefully get a second coat on the sections on Thursday so that I can start putting rail on them this coming weekend.

Halfway point

On Saturday I put the rail on the first half of the S-Curve.  So, this morning I installed the track with the rail.  I was really surprised how well it fit together.


I think what surprised me the most was how well the wood track lined up when the metal rails were connected together.  The metal rails butted tightly together while the wood had a tiny gap.


The only disappointment in this assembly is that the wood rails are not exactly the same thickness.  In the above picture of the rail juoint, I haven’t put the screws into the curve section yet so the curved part is a little high.  But, still, when the wood is at the same level, there is still as light difference.  I think what happened is that the cheap table saw I was using allowed the fence to move under the vibration of the saw and so the wood is not perfectly the same height.  This is OK as long as people are not looking to close at the joints.

The real excitement began when I put the rail on the second part of the S-Curve and installed it.  Note in the picture below that I decided to put in the ceiling mount in the area where there is no joist in the ceiling (right in front of the white car in the picture below).  I looked at it with and without and decided that the symmetry in the curve benefited from the extra mount.  I was quite thrilled when I put the second half of the S-Curve up and the metal track matched up perfectly as did the wood track.  From a distance you almost cannot tell that there is a break and the ties space out nicely at the joint.


So, now that the S-Curve was up and in place, it was time to start on the curve near the bathroom door.  Again, the wood and metal rail matched up perfectly with a tiny gap between the wood and no gap in the metal rail.


Note that the clamp in the picture above is holding the curve tightly to the installed straight piece of track.  This is because I don’t have the steel connectors to force the wood to stay in place.  After putting up the curves and the two straight pieces, I had a 4 foot gap between the sections.  So, I took an assembled 6 foot section without rail and marked it in place.  I cut it with the miter saw and it fit in perfectly if not a little to snug.  I then tacked on the metal rail and put it in place.

The one thing that I didn’t think of was that putting a section of straight track in the middle of two pieces would make joining the track very difficult since the little metal joiners are very tight.  I tried completely opening up the metal joiner but the nickel-silver is very brittle and broke off when I tried to bend it back in place.  So, I took another approach and opened the joiner slightly and tapped it on so that it was completely onto the new piece.  I then installed it in place, and used the nail tap to gently tap the joiner to slide over the end of the other rail.  It worked nicely and I used a pair of pliers to pinch it down on the track.

The only thing missing on this section of track is the power supply.  The order from Whole Sale Trains contained a 75 foot track to power supply wire bundle.  But, the wire has not shipped.  I am thinking of stopping at Lowes and getting some 14 gauge wire or something appropriate and hooking things up so that I can see if I have continuity in the track.  I can test this with the Performer car since it has lights that pick up power off the track.

Now that I have about half of the track installed, I took the cars and pushed them down the track. I was disappointed at how much friction the cars have in their wheel mounts and how much noise they make just riding the track.  I think I can improve on both of these with some lithium grease or something appropriate for metal wheels on plastic mounts.

Steel connectors ordered

I spoke with Pam today from ID Plates with the pricing to add the 2 additional holes was $0.08. Then there was an additional charge for “center to center” drilling. The total for the plates was $2.31 per plate. Pam asked me for a drawing so I used Visio and drew up a engineering picture of what the plate would look like with only the relevant measurements included. And, so we went forward with the order and put it on my Amex. In thinking about it, I hesitated for a moment because I thought that this was an expensive part of the project since these plates cost half of what the engine cost. But, to put it into perspective, each six foot section costs about $49.

First, the Ceiling Mount:

Item Qty Price Total
1x2x7.5in oak board 2 $0.70 $1.40
12×5/16 round oak dowel 4 $0.29 $1.17

Total $2.57

Each 6 foot section needs about 2 Ceiling Mounts (one every 16 inches). So, when you add up all of the parts, you have $49.00 or $8.16 per linear foot. (Note, in comparison, Loco-Boose charges $125 per linear foot.)

Item Qty Price Total
1x2x6 oak board 1 $6.31 $6.31
60 poplar ties 60 $0.18 $10.56
Nickel-Silver Rail 2 $10.50 $21.00
Nickel-Silver rail connectors 2 $0.69 $1.38
Ceiling Mount 2 $2.57 $5.14
Stainless Steel 0.050×1/2×3-1/2 plate 2 $2.31 $4.62

Total $49.01

So, when you look at the total cost for each individual six foot section, you can see that the cost of the Stainless Steel plate is really not that much. The only big difference is that I had to buy a minimum of 100 so I will have 50% waste. But even at double the price, the connector plates are still reasonably similar to the cost of other items and will add to the fit-and-finish of the project.

Bending my first piece of track

I got up late this morning and started the third pass with the router on the ties.  I finished all of them up by about 11am.  With that done, I got out the polyurethane and put the last coat on the ceiling mounts and then started putting it on all of the assembled sections of track.  Then, while waiting for the polyurethane to dry sufficiently to start a second coat, I got out the rail bender and ran a couple of strips of rail through the device.  I was surprised at how well it bent the curve into the track and how consistent it was.  I also experimented with unbending rail to see how difficult it would be if I bent the curve to small.  It turned out to be very easy to unbend it.

After getting all of the ties routed, I glued the ties on the second half of the S-Curve and put it in place to check the look.  Below is a picture of the S-Curve with both sections in place with ties.


Now that I have the curves all covered in polyurethane, the hangers mounted, the seconds checked for proper fit, and most importantly… the rail, I can start putting rail on the curves.  To do this, I got all of my little tools out and started bending rail.  Below is a picture of everything you need to bend and install rail on a curve.  This is the same set of tools needed for straight sections except that it includes the rail bender on the right side above the track.


From left to right, and top to bottom here are the pieces I use:

Item Purpose
Nickel-Silver Rail Guides the wheels of the train
Rail Bender Bends the rail evenly to fit the curve
1/2 inch black spikes Attaches the rail to the track
Hack saw Cuts rail to length
Track Assembled track used to support the rail between ceiling mounts
Curved Nose Needle Pliers Holds 1/2 inch spike in place while hammering
Brass Hammer Heavy head brass hammer for driving spikes in
Small head tap Drives spikes into final position after rail has been threaded onto track
ExactO Knife Scores ties to mark where to put spikes
Code 250 Precision Track Gauge Gauges track to get proper separation between rails

The process of laying the track is really simple but takes a fair amount of time to get it right.  The Precision Track Gauge is the key to the spacing and if you eyeball the rail on the track, you find that it should be about 1/8 inch over the wood rail under the ties.  The reason that you eyeball it is because the wood varies slightly since my saw is a cheap $100 Millwalkee and because the little cross members are pretty close but not exactly the same size.  1/16 of an inch is a big deal when working with small parts. After doing a complete track section, you will figure out how to do it most effectively.  That first track section can now be destroyed or used as scrap because it will have to many errors to be on your final production track.  Well, that is, if you are like me.  I had the benefit of creating prototypes of the various types of wood and got a good bit of experience learning to lay track when doing those sections.  Just be prepared for some waste.

Step 1:  Scoring – So, start by cutting off a small 2 inch piece of rail with the hack saw.  Using a small piece of rail for your measurements and marking is much easier than trying to handle a six foot piece of rail.  Put the small piece of rail in place over the second tie from the end, use the ExactO knife to gently score the tie on both sides of the metal rail.  Move over to the 4th tie from the end and do the same.  Now repeat this same process for every 5th tie until you get to the other end and do the 4th and 2nd from the end.  You don’t want to lock down the last tie because you need to insert the rail connector and it will go between the wood and the rail.

Step 2:  First row of nails – Using the bent nose pliers, take a spike and position it just to the inside of the scored line.  If you put it right on the scored line, the nail will be wider than the line and will push the track over. So, you are adjusting for the width of the nail and positioning outside of the two lines where the metal rail will go. Tap it down until it is the thickness of the plier’s nose.  This will leave enough space to thread the rail under the spike head but tight enough that it won’t move and you can make adjustments before final tapping.

Step 3:  Second row of nails – If you are like me, you will not get every nail exactly where it is suppose to be.  Don’t worry about it.  As long as you are pretty close, you can adjust for it after you thread the rail.  Take your small piece of rail, lay it against the nail on a tie, then with the pliers, position the next spike so that the body will be against the tie when it goes into the wood.  Tap it in to the same depth of the other spike and then remove the rail and do the same for all of the other spikes.

Step 4:  Threading the track – (Bend the rail as needed with the rail bender before threading the track) You should now have two small rows of spikes in your track that are spaced enough to thread the rail through the spikes.  You will be tempted to just lay the rail against the first spike and then start putting the second spike in to avoid the threading.  But, you will find that no matter what you do, you will end up with dings in the rail that are both unattractive and will scratch your car wheels.  Additionally, since the first spike is not on a perfect line with the rail, doing the method of spiking the rail in place will result in a rail that follows the spikes rather than staying with the curve you designed.  So, thread the rail through the spikes so that it is in the position that you want.

Step 5:  Tapping the nails – After you have threaded the rail through your spikes, let the rail determine what the line should be.  If you did it right, you have about 1/16 of an inch of play where the rail can wobble between the spikes and the spikes have about 1/16 variance along the wood rail under the ties.  So, let the rail be your guide and tap in the spikes so that they grip the rail in the line that the rail tries to form.  If you are doing a curve, this is very important.  You have to let the curve of the rail determine the final position so that you don’t have a curve with flat spots.

Step 6 – Cutting excess – Now that you have spiked the rail in place, you simply cut off the end of the rail to be perfectly in line with the end of the wood track.  If you have glued your ties on properly, you have 3/8 inch overhang where the rail is not supported on the end by a tie.  Be as exact as you can because when you put the next section to this one, you don’t want to have to adjust the rail after the fact.

Step 7 – Scoring the second rail – Get your handy dandy Precision Track Gauge out and put the single side on the rail from step 6 and your little 2 inch piece of rail on the end with 2 guides.  Now, use your Exact-O knife and score on the same tie that already has a spike so that you have lines for your nails for the other rail.  Do this same thing for every tie that you put spikes in for the first rail.

Steps 8-12 – Repeat steps 2 through 6 on the second rail.

Step 13 – Install Rail Joiner – Before you mount the rail in place, get your rail joiners (2 of them) and use the pliers to open up the joiner slightly so that you can get it started on the rail.  Use the hammer to tap it on further.  Be careful that you don’t catch the edge of the tie with the sharp underside of the rail joiner.  Leave half of the joiner exposed off the end of the track to receive the other section of track. I tried putting joiners on opposite rails so that I didn’t have to worry about the direction of the track, but found that there are other reasons (like the wood rails are not perfectly symmetrical or the curves only fit one way) that you have to worry about the direction of the track, so it is just as easy to simply put the joiners on the same end of the track.

Well, that is it.  That is how I make a section of track. Assembling it to the ceiling mounts and lining everything up is another story. The nice thing about a custom made track is that you can accommodate construction errors as you go along.  So, if you build a section of track and the metal rail is a little short, you just use the hammer and tap gently on the opposite end until the rail meets exactly with the joining section.  Then, as you build the next section, you increase the length of the rail just a bit to compensate.  This is only really necessary for sections that are not exactly 6 feet long because the rail is exactly as long as your track from the start.  You may also find that it is a nice touch to use a Dremil tool to buff off the end of the rail so that there are no sharp edges to scratch the wheels of your train.

Track, Tags, and more Polyurethane

While stuck in traffic on Rt 606 this morning, I got a page from my housemate that a large box had just arrived by UPS. His description of it being long and thin indicates that it may be the track. That is great because if the track has arrived, I can start putting pieces up permanently!

At noon (lunch time), I followed up with Cascade Laser Corp to get pricing on the tags and talked with Tabitha. She said that the stainless steel tags (0.029×1/2×3) were going to be more than $1 a piece for quantities of 100 to 150. Her price for the brass plaque tags that are stock (0.020×1/2×3) is $0.65. This was about $0.20 higher than ID Plates. Tabitha is still going to work up a quote, but I am thinking that ID Plates may be a better option. Tabitha is working up the quote and will send it by email.

So, I called up ID Plates and talked to Pam. She can get custom tags in either 0.029 or 0.050. Standard steel dog tags are 0.040 and I need something as strong as a dog tag. So, I asked her to get a quote together for both 0.029 and 0.050 in steel in quantities of 100 to 150. She is going to generate the quote and send it by email.

I got quotes back from both vendors. Cascade Laser Corp priced it at $2.98 per tag for 0.029 stainless steel. ID Plates priced it at $1.49 for 0.029. That is half the price of Cascade Laser Corp. Even at 0.050, ID Plates is cheaper. So, I placed the order with Pam at ID Plates for 100 plates, size 0.050×1/2×3-1/2. The holes will be 1/4 inch from the ends and centered on the plate.

I left work a little early and got home at about 5pm. I started routing out the slots on the new ties that were stained by dipping and things look good. Only problem so far is that the new ties are 1/16 to long so they didn’t fit in my jig and I had to adjust. That ate up about 30 minutes of stupidity on my part. Pass one was completed by 7pm. Pass 2 was completed by 9pm. I still need to do one more pass but I did a third pass on about 50 ties so that I could put ties on the second half of the S-Curve to let it dry overnight. So, can you tell the difference between the first set of ties (the two on the left) and the new ones that are dipped before routing (the two on the right)?


The metal rails and all of the accessories that I ordered also arrived. This included the rail bender and the connectors that will be used to join the sections together. After I finish the current completed track, I will start putting some rail on so that I can be ready for the tags to arrive.

The last thing I did for the night was to get out the polyurethane and finish off the rest of the ceiling mounts with a second coat. So, by the time I have started assembling more track, the hangers will be dry and I can continue to lay out the curves and finally the straight pieces.

Gotta do something!

Well, I did nothing Wednesday night because the therapy on my back is killing me and I can’t feel my feet. I just went to sleep at 8pm and felt guilty for not working on the train. So, tonight, I took some pain killers and made myself do something useful.

First, I cleaned the prototypes off the bar and put the USA track that I bought for reference material onto the top of the bar so that the curve simply follows the shape of the L on the bar. It looks very nice. I then put the power supply for the train under the lip of the bar and connected the wires so that the track is powered up and I can test the power supply. Finally, I put the Performer car for the train on the track and fired it up. The lights in the car come on and the little back light on the car comes on. It is pretty. The bar looks much cleaner.

It was raining outside so I put the table sander under the porch and sanded the ends of the new ceiling mounts that I had built. After sanding them down nice and smooth, I got out the Polyurethane and brushed on the first coat of finish to the bottom part. As I brushed it on the little bubbles started appearing and then I realized my the stupidity of my stupidity. Remember in a previous post I said that I had bubbles from brushing on the polyurethane and so I started using an old T-shirt to put the stuff on the track? But, using the T-shirt method is very slow and you don’t get a lot of polyurethane on the wood in the first pass. The wood soaks up just about everything you can get on it. So, for these new ceiling mounts, I decided that I would brush on the polyurethane so that I got a lot on the wood, and then would use the T-shirt to wipe off the excess after it had bubbled. The idea hit me because the little bubbles weren’t coming from the brush. They were coming from the wood dowels. Whenever I brushed across the dowels, they would bubble like crazy and then slowly stop. So, by brushing it on to get a good thick coat, and then wiping the excess after the bubbling had stopped, I got a really smooth and clean finish. They look a lot better than the first set.

Connecting the sections

The last engineering hurdle for this project is to figure out how to tastefully connect the rails together.  I stopped by Lowes and looked through all of their stock to find something that might look OK. My first thought was to cut 1/4 inch Oak stock into 1/2×3 strips that would then be screwed and glued to the ends of a rail to create a large tongue and groove on the side of the rail. So, I got the stock, cut it to length, and then put it up with some test track. It drew to much attention to the joint between the sections and looked to bulky. It is still an option, but I want to look at other options.

After pursuing that idea, I thought about making up small metal plates that are about the thickness of dog tags like what you get engraved for your pet. If I could get a tag that is 1/2×3 inches and 0.020 or 0.040 in thickness, I could join the rails very tastefully with blanks and 4 screws. So, I searched online and found two places that might work out:

Both places have brass tags that are the right shape and length, but they are in bright brass. The ideal tag would be made of nickel-silver and match the color of the track. But, I cannot find anywhere online that makes such a plate. But, I did find a place that makes stainless steel with a brushed finish. While a little to bright, it would be better than brass and the stainless steel would be much stronger than brass.

So, after work today, but before close of business for Oregon, I called Cascade Laser and talked to them about ordering 1/2×3 stainless steel tags with rounded corners.  They said that it would be a special order to get the tag in stainless steel and that they would get me a quote by Friday.  The minimum order is 100 and as long as the cost is less than $1 per tag, I will buy them.  The thickness available is 0.029, 0.031, and 0.040.  They said that 0.029 was standard for pet tags.

DHL delivery is not the engine

Well, as all retailers do, the weight of the box was overestimated. DHL‘s email that the box weighed 5 pounds was an exaggeration and when I picked up the box, it felt like nothing was in it. Well, it certainly didn’t feel like an engine was inside. Instead, it was just the bear/gorilla car.

So, I have all of the cars that I ordered, but I don’t have an engine to pull the train.

The S-Curve is done

While is is very pretty and eye-catching, building the only S-Curve was a bear of a job. The challenges were:

  • The curve traversed an area that ran parallel to the floor joists
  • The curve required making one almost perfect 90 degree curve followed by another 90 degree curve

Since the S-curve ran parallel to the floor joists, it was difficult to locate a secure mounting point for the ceiling hangers. Considering the weight of the track alone, attaching the mounts to drywall is not sufficient and I didn’t want to open up the ceiling to put a cross member between the joists. So, I adjusted the starting point of the curve so that the ceiling mounts zig-zagged in line with one joist and I used only one screw in each mount. Since the curve alternates sides of the mounts, this should be sufficient to avoid a bow. I will put expanding screws into the dry wall to secure the free side of the ceiling mount. The final result of the problem with the joists is that there is a gap in one of the curves where I could not put a ceiling mount. I am considering putting one in just for cosmetic purposes.

The curves are not perfect curves. Even though they came from a jig that was a perfect curve on the work bench, the wood had inconsistent spots in it that caused some areas to bow more than others. When stacked one on the other, the curves are very close to each other, but when cutting the ends to make a perfect 90 degree curve, a difference of an inch can be the difference between a good curve and a ruined one. Case in point, I cut 1.5 inches to much off of one curve, and had to work to rebuild the end so that I would not lose one of the curves.

So, here is the requisite picture. This first picture is straight on looking down the length of the floor joists. The first half of the S-Curve has the ties mounted and the second half is bare. The second half of the curve also does not quite meet up with the next ceiling mount and if you look carefully, you can see that I had to rebuild about 1.5 inches of the curve after cutting off to much. Rebuilding is a strong word…I had to brace and reglue the piece I cut off. Notice that a ceiling mount appears to be missing in the body of the curve closest to the camera because there is no joist in that section of the ceiling to get a strong mount.


This second angle shows the grace of the S-Curve as it makes the bend around the duct work. The first half has ties on it and the second half is bare. Again, from this view it is clear that a ceiling mount is missing. On one hand, the ceiling mount would block a beautiful view of the train as it moves through the curve, but on the other, the lack of the ceiling mount takes away from the symmetry of the curve. After I get the ties completed and am just about to mount the curve permanently, I will decide if I need the extra mount.


More ties, more mounts, and an 8 footer

Now we are into the monotonous part of the project. I have now finished the proof of the design of the ceiling mounts, the ties, and the track. So, now it is time to just repeat the process over and over till I have converted my raw stock into production pieces.

Today was a beautiful day at about 85 degrees and so I started out with the ties. I peeled off the stupid little Lowes SKU stickers on the 3/8 inch square dowels, sanded the glue off, and then stained all of them. So, now I have stained roughly 1000 ties and they are all sitting on my workbench separated to let them continue to cure and dry.

After finishing the ties, I laid out another six foot section of rail and glued the ties to it. This time I did something a little different. It seems that the cross members and ties are of sufficient strength that I don’t need to put screws in the rails to hold everything together. So, I tried this out and built one six foot section without any screws. I sanded it, bounced it around, and it seems to be just as strong as the others that have screws.

Next, I took the second piece of the curve out of the press and lined it up on the ceiling. I then put two more ceiling mounts in place on the curves and screwed them to the ceiling. Only one screw hit a joist as would be expected since the curve forces the mount to be at a 45 degree angle. It doesn’t matter because even one screw is sufficient for the weight of the train. Here is a picture of the entire curve in place with the new mounts. The curve is ready to have a coat of polyurethane put on it and then the rails once they arrive. Here is the required picture of the work.


After finishing up the curve, I built the remaining ceiling mounts. All of the pieces were already cut so all I had to do was apply the glue, set them, and make sure that they were square. So, here is a picture of the second half of the mounts just after glueing.


When I was originally buying the stock for the project, I did not know what length of rail I was going to use and I did not know that I was going to use Llagas rails. So, I bought all manner of lengths in order to be as efficient as I could on the supporting rail system. Well, one of the lengths that I bought was an eight foot piece of stock and I split it to make the rails. So, today I decided that it wouldn’t hurt to go ahead and be efficient in the stock for the long run on the opposite side of the room from the pool table. The immediate problem is that all of the Llagas rail will be in six foot sections and I will have this strange eight foot section. But, that isn’t an issue because I have already cut two six foot sections of rail into two foot lengths for the prototypes. I will just pull that rail off the prototypes and deploy it in production. Here is a pic of the eight foot section under clamp.


Staining the second batch of ties

For the second batch of ties, I have learned a few things.

  • I need to stain the ties before I route out the notch for the rails
  • if I put a piece of wood behind the ties as I cut them, I will avoid the splinters on the back side due to the blade on the soft poplar wood
  • dipping the ties and then letting them dry without wiping them results in a better looking tie as it is consistent and very dark

So, I started the morning by building a jig on the saw that has a thick piece of wood on the fence and then my little odd shaped piece of wood for a stop. I cut four 3/8×3/8×36 inch square poplar dowels at a time and put them into a pile. The last piece always has the stupid SKU from Lowes that gums up the sander and so I separated it into a different pile.

Next, I took two poplar test pieces for the main rails and laid them on top of some old bricks that were pulled up from the patio and are just waiting to be buried or dumped. I then took 3 already routed and stained ties and put one on each end and one in the middle to separate the boards the exact spacing that the normal rails are. Finally, I took a six foot 1×2 poplar board on end and set it at the end of the ties to form a backer. This whole setup was done so that I could dip the ties, brush off the excess, and then lay the ties on the beams that were separated the exact amount for the normal beams. This setup is done because the beam will absorb some of the stain causing a light spot on each tie and I want the light spot to be exactly where I will route when cutting the notches in the tie.

Here is a picture of the newly stained ties on the bricks.


It was about 85 degrees and dry so the ties seemed to dry enough that they aren’t sticky to the touch in about 4 hours. After letting them sit for 4 hours, I took them inside and laid them on the worktable with the light spot from the rails facing upwards. I separated each tie by a small amount for airflow to let them continue drying.


After putting this first set on the work table, I then dipped and set another row of ties on the bricks. But, by the time I set these, the temperature had dropped to 75 and it was a bit more humid so the ties did not dry as fast. I let them sit outside till about 9pm and then took them in because it is suppose to rain tonight.

Laying the first curve

Late last night I figured out how I wanted to put the curves together near the bathroom door.  This is particularly critical because I only have 8 curves and if I damage any one of them, I have to set my jig back up and make another.  While not the end of the world, I would consider this a precious resource for the project.

So, I carefully measured, set the wood in place, and did everything I could to try to make sure that it was going to be correct.  I marked it and then cut the two curves that would make the 180 degree turn.  After cutting them, I took a deep breath, and tried them in place…and they fit!  Perfectly!  So, I then started the process of gluing the ties onto the curves.  Easy enough except that the curves have a bunch of cross members that keep the curve formed and they are irregularly spaced.

So, I had to change bits on the router and route out the center of a bunch of ties.  This is where the original set of ties came into play.  When I was creating the prototypes, I completely stained a bunch of ties before thinking about the issue with the stain preventing the glue from doing its job.  Well, these ties became useless because they wouldn’t stick to the glue.  Anyhow, since the stain didn’t penetrate more than 1/32 of an inch of the wood, I just routed out the center between the beam notches and now I have ties that will fit where the curves have cross members.

With everything ready, I laid the curve on the end of the table, got out the blocks and glue, and started laying the ties.  Everything worked perfectly.  I then took one of the identical curves and put it on top of the glued ties and put 40 pounds of weights on it to make sure everything was flat and that the ties were firmly pressed into the glue.

Here is a picture of one of the curves curing.


And here are a couple of pictures of all 8 curves stacked to make sure that they were consistent.

All8Curves2_full All8Curves_full

And, here is a picture of one of the curves in place near the bathroom door.


While letting the second curve sit with weights, I laid down the rails for another six foot section and glued the cross members in place.

Polyurethane and Latex do not mix

I got home at about 6:30pm, took care of the dogs, and then started on the second coat of Polyurethane on the ceiling mounts.  I would consider myself an idiot for not thinking of this before, but when I put on the first coat of Polyurethane, I used a brush.  I should have used a cloth.  When it hit me that I should have used a cloth, I couldn’t help but remember that we always used a old T-shirt or something soft to apply finishes.  

So, I thought for a moment and decided that an old white T-shirt would work so I dug in the bottom of my closet and found one.  Cotton…nice!  I then put all of the ceiling mounts outside on the stack of old bricks from when we pulled up part of the patio to put in a rock walkway, and put out my can of Polyurethane.  I put on a pair of latex gloves and started wiping on the Polyurethane only to discover that the gloves were disintegrating as I worked on the wood. Was it snagging and ripping the gloves?  OK, so I put on a second glove and was very careful to only tough the T-Shirt material with the glove.  But the fingers that were touching the Polyurethane dissolved once again.

So, now I know…Polyurethane and latex do not mix.

I finished up the job without gloves and my hands were quite sticky.  I washed with soap and most everything came off.  I let the mounts dry outside for a bit, and then took them inside to finish. 

I also got out the sanding cloth and cleaned up the ties wherever there were irregularities in the Polyurethane.  Everything is done and ready to start putting some track on the ceiling.  The only problem that I have not figured out is how I am going to connect the pieces when they are in the air.  I have an idea that I am going to test out on a prototype.  Basically, I will take a small piece of wood and drill through it, then drill similar holes in the end of the mounting rails and use a smaller dowel to loosely hold the block to the wood rails.  The goal is a loose fit that will allow the wood to expand and contract with temperature changes in the room.

First coat of Polyurethane

I got up early this morning and sanded the ceiling mounts to prepare for their first coat of polyurethane. This was simply running them on the table sander to smooth down the ends where the dowels stick out slightly. After completing the sanding, I set them up on a row on some spare bricks outside, and brushed on a coat of polyurethane. I ran into a problem with little bubbles in the brush that would stick to the wood. I have a light cloth sanding pad that is for taking off these little bubbles. There are an awful lot of them. Here is a picture of the ceiling mounts with the polyurethane. A raw mount is sitting on top of them to show the contrast in the color that the polyurethane brings out in the red oak.


After sanding the ceiling mounts, I also sanded down the wood rails that I assembled late last night. After I finished putting the polyurethane on the ceiling mounts, I took the four sections of track and set them outside against the bricks. I then painted on a coat of polyurethane. Again, I had a problem with the little bubbles, but in this case, the dark ties really exposed the bubbles. After the coat dries overnight, I will use the sanding pad to try to take off the bubbles and then put on a second coat. For a test run, I put a second coat on the 2 foot section that I did the original polyurethane test on. There was no noticable color change. So, I am not sure if the second coat is necessary on the track. The reason I question it is that the track has such a contrast between the rails and the ties that the first coat doesn’t seem to need buffing out. Still, I will do a second coat just to make sure that I didn’t miss anything.


I also got a call from California & Oregon Coast Railway saying that they were having a problem with my credit card.  I called them back and the date was wrong on the card.  They said that they had the order ready to ship out?  I am surprised at that as I was expecting them to take my order and send it to Llagas, and then have Llagas ship it directly to me or back to them before to me.  So, maybe they keep this stuff in stock.

Oops and recovery on a Ceiling mount

Well, I discovered just how strong glue is when putting the dowels in the ceiling mounts.  I made a mistake on one of the mounts and needed to remove the dowel bar.  I tried twisting, bending, pulling, pushing, tapping, whacking, and anything else I could think of.  but, it would not budge.  So, I resorted to drilling.  The nice thing about using the jig for aligning the drill bit for the holes is that the bit drilled out the dowel and the piece looks perfect.  It is much easier to replace a dowel than to replace an end piece for a mount.

While at it, I built 4 more mounts for a total of 12 finished mounts, sanded down the rails for another six foot section, and put the ties on the six foot section.  I put the clamps and weights on the new section, took a shower and headed for work. 

Ceiling mount assembly starts

Today was pretty productive.  I started out early by assembling another 6 foot straight section.  Unfortunately, I found that the belt on the sander was to gummed up to really get a good surface so I went to Lowes and got a new belt.  I glued the section together, put the ties on it, and then put the clamps and weights on it to hold it tight.

After finishing the 6 foot section, I clamped a board onto the disk part of the sander so that only 3/4 of an inch of the disk was exposed.  I then sanded a small amount off of each end of the every dowel so that it would fit snugly into the holes of the stock for the ceiling mounts.  A few of the dowels still had some glue on them, so I put the gummed up belt on the sander and cleaned them up.  A close friend came over at about 3pm and I took a break till about 5 chatting with him.  By 7pm, I had assembled 4 of the ceiling mounts.  I found that assembly is a bit slow because I have to get just the right amount of glue on the dowel end, then tap it into the stock with a rubber hammer, glueing the other piece of stock, aligning and inserting the 4 dowels, tapping the second piece on without disturbing the other end, glueing on the spacer that holds the rails in the center of the mount, and finally twisting it while the glue is still pliable to make it lay flat on the table. 

Sounds easy right?  Well, while it is simple, it is not trivial.  Making sure that the stock is assembled with the correct side up is critical and if you make a mistake, you cannot get the dowel out after tapping it in.  Also, after assembly, the whole thing is perfectly parallel, but it is often slightly twisted so you have to do a bit of twisting to get it to lay flat.  So, I can assemble one in about 10 minutes.  I can only assemble 4 at a time as I only have 4 yellow clamps that will hold the track aligning block in place.

I also bought enough 3/8 inch plugs from Lowes for the project.  But, I couldn’t find 1/4 inch Oak plugs.  So, I went online and found the right plugs at Widgit Mfg. Co..  At Lowes, they cost $0.22 each.  Online they cost $0.08 with free shipping. 

Building the ceiling mounts

The prototype ceiling mount took 4 hours to make, but like each part of this project, the prototypes are implementations of ideas and often involve experimenting with how to put the piece together and how to get the geometry correct. I always have the design either on paper or in my head before starting the prototype but I often find that I haven’t thought of some little thing so I have to improvise to get the final prototype. I then think about how to build the production run more efficiently, and sometimes make a second prototype. This is what I did for the ceiling mounts.

So, after thinking about the first and second prototype, I devised a means to produce the ceiling mounts in a production form that would be as efficient as I could make it. This entry details the process for making the ceiling mounts.

The first step is to cut the raw lumber (1×2 Red Oak stock into 7.5 inch sections. The picture below shows the cut wood for 40 ceiling mounts (80 pieces) plus one piece just in case I mess one of the pieces up.


Once the raw stock is cut to length, I sanded it lightly to make sure all sides were flat and to glean off the rough ends. I stacked the raw stock next to my routing table. The router is set up with a detailed bit as shown in the previous posts with the prototypes. Below is a picture of the stock on the left before processing with the router.


During the routing process, I accidentally flipped one piece over and routed the wrong side so that was my one extra piece. So, throughout the rest of the process, I used that piece for test before starting the step. I discovered that the hose to my shop vac fit perfectly to the hole on the back of the fence for the routing table. Previously I did not use the fence because it has a spring loaded guard that gets in the way when routing the end of the wood. But, the vacuum was very nice for keeping the dust down.

Routing each piece was a bit easier with the fence. Without the fence, when I got to the end of the wood, the would would catch on the rough fibers and would kick sideways which would sometimes cause a nick.  But, with the fence, I could run the board to the end without a hickup. After doing both sides of the board, I would then do the ends and put into a pile on my right. The picture below shows the stock for the ceiling mounts after applying the detail.


During the prototype phase, I found that you have to route the wood before you drill the holes for the dowels that connect the two pieces. If you drill first, the wood will be weak on the ends and will splinter near the holes. Yea, that is logical, but I still tried it because it is easier to mark the holes for the drill before routing.

So, after routing all of the stock for the ceiling mounts, I stacked it near the drill. I then used my little jig that I made earlier when I thought I was going to use dowels to hold the curves together and for using as a stop on the miter saw. By using a clamp and setting the jig in the right location, I could do all 4 holes in the stock and get the product to have perfect symmetry. I start with the routed side up, drill a hole, flip the board over, drill the second hole, flip the stock over so the routed side is up again and turn around, drill another hole, flip the stock, and drill the fourth hole. I don’t know why I didn’t think of this when doing the prototypes.

After drilling the end holes that would be used to receive the dowels that connect the pieces, I separated the stock into two piles. One would be for the top pieces and the other for the bottom pieces. The top piece needs a 1/8 inch hole with a 3/8 inch countersink hole half way through the stock exactly 1.75 inches from the edge and dead center in the stock. The holes will be used to receive screws to mount to the ceiling and the 3/8 inch countersink hole will be filled with a 3/8 inch mushroom cap. The bottom needs a 5/32 inch hole with a 1/4 inch countersink hold half way through the stock at exactly 2.25 inches from the edge. The holes in the bottom will be used to accept #4 aluminum screws that will mate up with the track to mount the track to the ceiling mount. Each hole will be filled with a 1/4 inch mushroom cap. The reason that the holes for the caps in the bottom piece are smaller than the top piece is due to the visual impact of the caps with relation to the distance between them.

Below is a picture of the drill press with the little wood jig that I made.  Notice the the jig has a nook for receiving the stock with a long side to keep it flat and it also has another piece of wood on the other side so that I can use the jig as a stop for the miter saw.


After drilling all of the holes, I sanded every side of the stock with a fine sanding belt on the table sander. This took the little burrs off of the edges of the wood where the router bit and drill bit touched the wood. The result is a very soft feeling piece of wood. Below is a picture of the two sets of finished product. The left side is the stack of top pieces, and the right side is the stack of bottom pieces.


The last step in creating the parts for the ceiling mounts is to cut the dowels to length. When doing the prototypes, I cut the dowels with a hack saw. But, for the production run, that would have taken forever and it resulted in uneven lengths. So, for this run, I used my little jig to be a stop on the miter saw and cut all of the 54 dowels into 1 foot sections. Actually, each dowel is short by 1/8 inch which is the thickness of the saw blade. I have plenty of clearance for the tallest car so this little bit won’t matter. What does matter is that every dowel is exactly the same length.

One little problem that I run into with the dowels is that Lowes (and Home Depot) put their computer SKU tag on the dowel on the end by wrapping it around the dowel. The tag is difficult to get off and it leaves a sticky side on the wood that gums up sanding belts and also prevents the wood from taking stain or finish. So, I separated the ones with labels into a different pile (in picture below on top of right group of stock in the picture below).  I then peeled off each little sticker, then rubbed with vinegar, let sit for about 2 minutes, then rubbed with dish soap and water.  The gum didn’t completely come off, but what didn’t come off isn’t very sticky at all.  So, I then took the dowels down to the sanding belt and held the dowel so that the full length of the dowel was on the belt parallel to the direction of the belt movement and I quickly rolled the dowel back and forth so as not to produce a flat spot.  The remainder of the glue came off, partly on the belt, but did not create gummy areas on the belt.  Below is a picture of all of the pieces (except the mushroom caps) necessary to build the ceiling mounts.


The remaining work to be done is to sand 3/4 inch of the ends dowels down just a millimeter so that they will fit snugly into the holes of the stock. This little sanding bit is necessary because the holes are 5/16 and the dowels are 5/16 so the fit is to tight to reliably be able to put the dowels in with glue and have time to position them. If you don’t sand, within a few seconds of putting the dowel in the hole with glue, you cannot turn it, pull it out, or push it in further. So, taking a little off of the dowel gives you time to work on adjusting the position in the stock.

First Straight Section Production Run

So, I started the morning early since I couldn’t sleep. I assembled the first production 6 foot section of track. I decided that the maximum length of any section would be 6 feet since the metal rail is 6 feet in length. By approaching the construction of the straight sections this way, I will be able to take down a section if it gets damaged and I also minimize the number of rail cuts I need to make.

What I decided to do for the sections is to start each run with a tie offset by 1 tie width. This will make it so that the begin and end of the section meets with the next section and has a transparent seam between the sections.

To make sure that the work bench stays flat since I am going to be putting weights on the ties to press them into the glue, I took a poplar 6 foot section and glued it to the bottom of the plywood work bench. I then drilled and made countersunk holes so that I could screw through the plywood to the poplar beam and get a really tight fit.

Then, I put two 6 foot beams on the table and screwed down a spacer at the head of the table and laid out the ties with spacers all the way down the beam till I got to around half way (3 feet). I then screwed another spacer in the position that was the right space to fit between two ties. I continued down the end and put a spacer at the end. The spacers are just cross members that are permanently attached to the work bench. After getting these in place, I marked the beams at 2 feet and 4 feet. I then glued a cross member between the beams so that it fit between the ties. Finally, I took the ties off the board, marked where the permanent cross members are, clamped the cross members in, lifted the board, drilled pilot holes in the wood with counter sunk holes, and screwed the cross members in. I cut 5/16 dowels and glued them in over the screws and let it sit for about an hour to set. The last step was then to use the hack saw and cut the ends of the 5/16 dowels close to the edge of the beam. I then took the naked beam with cross members and put it on the table sander and smoothed everything down just like I did for the curves.

After getting everything cleaned up, I cut a 6 foot poplar board into 3 two foot sections to be used with clamps and weights to press the ties into the glue to get a good bond. I then started laying the ties one by one. I put a dab of glue on the tie and pressed if firmly onto the rail so that the rail fit into the channel that I routed into the tie. When I got to the first cross member, I put one of the 2 foot poplar boards on the group, clamped it, and continued with the rest of the rail. Once everything was clamped down and all ties were on the board, I put 10 pound weights on each 2 foot section and let it sit for about 5 hours. After the 5 hours, I lifted the unit up, pulled out all of the little spacers, and has my first production rail. Here is a picture of the rail during the 5 hour setting period.


The first section of track goes up

I laid my first 6 foot section of straight track. Man…it is time consuming to put those little nails in.  It is good that I use poplar wood for the ties as the little nails bend if you try to put them through the oak. I put a nail on every 5th tie and then ran the gage over the rail to make sure that the spacing was correct. I tried to build a jig that would fit between the ties and the rails to set the spacing on the rails, but the nailing always results in a slight misadjustment of the track. So, I resorted back to the previous method of setting the track by eye and it looks just fine. When the track is mounted on the ceiling hangers, the 1/16 inch imperfection is invisible. Below is a picture of 3 of the six foot sections mounted on the hangers.


Below is a picture of the mounted wood rail with the metal rail attached and with two of the train cars on the rail.


I also got an email from DHL that a 5 pound package has been shipped from Wholesale Trains. This could be the engine!


My back was hurting so bad last night that I couldn’t sleep a wink.  So, I went ahead and got out of bed at 4am and went to the basement.  I lifted a bit and tried to stretch out my back, then pulled the ties from the tape that was trying to protect the wood from the stain, and just messed around.  I left for work pretty early and so I went to Lowes at 7am.  They had just opened and the place was empty.  I had gone to Lowes in Chantilly before my visit to the doc about my back, but the Lowes in Chantilly was completely out of 5/16 Red Oak round dowels.  But the Lowes in Sterling near where I work is often better stocked.  So, at 7am, I went back to the isle where the dowels are and they had a full bin of them.  So, I grabbed every one of them.  They are 3 feet long, so it takes 1 and 1/3 of a dowel to make a ceiling hanger.  I think I will need 48 ceiling hangers so I will need 64 dowels.  Well, they only had 54 so I took the lot of them.  I also got some round 3/8 Red Oak mushroom plugs that will fit into the countersunk holes for the screws that will put the mounts up on the ceiling.  We will have to see how those look.

Ordered the track

OK…so I ordered the track and nails to start assembling the system.  It will be a couple of weeks before it arrives and by that time I should have a bit of the system mounted to the ceiling.  When I bought the sample track, I bought it directly from Llagas Creek Railways.  But, they have changed their order form so that you now must purchase through their dealers.  Problem is that the dealers are tiny little shops that don’t have online stores and whenever I called a store, I either got a recording or no answer.  But, even though I couldn’t get through to California & Oregon Coast Railway, I was able to place an order online.  Here is the order:

Code Name Qty Price/Ea. Total
LCPKNSRNJ Nickle Silver Rail Joiners (20) 2 $13.75 $27.50
LCPKLS Large Spikes (500) 2 $11.00 $22.00
LCPKPTG Precision Track Gauge
(Code 250/215)
2 $16.95 $33.90
LCPK250NS Code 250 NS rail, 6′ 36 $10.50 $378.00
Track and Rail Shipping
1 $11.00 $11.00
LCPKLRB Lindsay Rail Bender
Rail Size: 11250
1 $84.95 $84.95
PB- 250-
SJC- 5_0mm
Rail clamps
1 $20.00 $20.00
  Shipping: UPS Ground     $39.65

  Total     $617.00


Polyurethane 2 foot section

Below is a picture of a completely assembled 2 foot strip of track with polyurethane.  The ties are the original ones that I dipped and rubbed so they are a bit lighter than the ones that I have finished up a brush.  I purposefully put the finished section on top of the unfinished curves to show how the polyurethane changes the color of the wood.  It takes on a beautiful honey red color. 

This was the last step in the prototyping in that I now have my method of assembly, method of stain, and have verified that the colors look good. 


The prototype Ceiling mount

I have the curves for the area near the bathroom cut to length, the ties are stained, and I have validated the colors after putting on the finish.  So, now it is time to finalize how I am going to attach the track to the wall or ceiling.  A good portion of the track runs near the walls but I have been thinking about the overall look.  I could use square wood like the rails against the wall.  Or, I am thinking that if I use round dowels and a nice routed edge on the mounting wood, I could achieve a look that is slightly greek.  So, I need dowels that are not to large to upstage the mounting wood or to large for the train but large enough that they can support the weight of the train. 

So, I chose 5/16 inch dowels and mounted two dowels on each side of the ceiling mount.  Rather than using a different mount for the track near the wall and the track that is suspended from the ceiling, I decided that all of the mounts will be identical. Below is a view of the track mount.


This mount is 7 inches wide and has about 1/4 inch clearance when a car goes around the tightest curve.  So, the production mount will be 7.5 inches wide to make sure that I have more clearance as cars go around curves.It is a bit hard to see the details of the routing but here is a closeup view of the top.  The two holes in the middle of the mount are for screws that will hold the top to the ceiling.  The holes are drilled for 1/8 inch screws with countersunk holes.  After mounting everything, I can use a plug to cover the screw heads.


With this design for the ceiling mounts, I gained an extra 1/2 inch over the original plans for the mounts.  At least now it won’t look like the Giraffe’s head isn’t about to get cut off.  And, just to make sure that the tallest car will clear the mounting assembly, I put the sample track through the mount and then the tallest car. 


My curve versus retail curve

I assembled all 12 pieces of the retail track that I purchased online.  The goal of buying the track is that it was suppose to be the same size as what I had built.  I was disappointed to discover that the measure of the diameter of the track appears to be to the inside edge of the ties. I had interpreted the drawings from Aristo to indicate that the radius of the track was to the centerpoint between the two rails.  So, as you can see from the picture, my centerpoint in my curve is the same as the inner edge of the ties for the retail track.

So, my track ends up being slightly tighter than 2.5 foot radius.  In the picture below, I have put my track on top of the retail track.