The McCallum Odyssey

Tell me, O muse, of that ingenious hero who travelled far and wide after he had purchased the famous locomotive McCallum. Many cities did he visit, and many were the nations with whose manners and customs he was acquainted; moreover he suffered much by electronic gremlins while trying to bring his locomotive safely home. Tell me, too, about all these things, O daughter of Jove, from whatsoever source you may know them.

 With apologies to Homer, here begins the tale of my odyssey working on my SMR Trains Locomotive McCallum. I spent the good part of a week tuning and debugging the locomotive to get it to run reliably. Like Odysseus reuniting with his wife, in the end I was able to have McCallum running the way I want it. The breakthrough came when I swapped the McCallum's tender for Whiton's and realized the intermittent problem was in the tender.

 Lets begin  the tale with the decoder installation. Prior to installing the decoder, McCallum was not my best runner. It had problems with shorting. The front pilot truck wiper tended to catch on the turnouts. All in all it was a problem child.  Sometimes installing a decoder helps one find shorts because the sparks are very obvious compared to regular DC running. So I charged ahead, hoping I would have better luck than Patroclus at Troy.

For the McCallum, I decided to install a Soundtraxx Tsumani AT-1000 light steam decoder. The rated current is 1.0 amp. Our testing showed that the SMR locos draw about 0.70 amps at stall. Thus the current rating should be acceptable.

Soundtraxx Tsumani AT-1000 installation in boiler,
with speaker in fire box

The decoder install went rather uneventfully, thanks to my previous experience installing the decoder in Whiton. But even though the decoder worked fine and sounded good, there were problems with running quality. The loco was stalling every couple inches of running, and was shorting at several points.

Scorching on the valve gear from
shorts with the pilot truck

The shorts at the front of the engine were due to the metal wheels touching the valve gear or cylinders on my sharp curves.  I fixed the shorts on the pilot by adding a longer and stiffer spring to the pilot truck post. That raised the front of the loco enough so that the pilot wheels don't touch at cylinder or the valve gear guides. I had scorch marks on those parts so I knew exactly where the shorts were. Now the pilot truck has clearance for my sharp curves, and doesn't look odd at all.

I also spotted a short on the tender where the factory installed wiper was touching the frame near a wheel arch. But these wipers were causing even more trouble. The loco was suffering from intermittent electrical pick-up.


After much trial and error I realized the lack of electrical pick-up was coming from the tender side that is attached to the frame. I could make the loco run by touching the wheel on that side with a hot alligator clip.  I could also make the loco run by tilting the tender to that side.  The eureka moment came when I swapped the McCallum and Whiton tender. The Whiton tender runs well. When it was attached to McCallum the McCallum ran much better. This really narrowed the problem to the McCallum tender and the wheels without the wipers not getting pick-up.

The problem was actually on the other side though. The screws that hold the factory installed wipers on the tender trucks for the non-frame side (on the engineer side of the tender) were rubbing against the tender frame. This was not causing a short because I had insulated the area with nail polish. But it was causing the sprung wheels sets in the tender to lose pick up as the hard frame-on-screw contact was effectively removing the truck equalization. You could actually see the fireman side wheels raising up off the rails.

I completely disassembled the tender. Then I ground off about a 1/16th of an inch from the tender frame wheel arches with my Dremel tool and a grinding stone. Reassembling the engine universal joints is about as much fun as assembling a Microtrains 1015 coupler, but I got the drive to go together. Now the tender wheel sets and their associated gear towers can rock freely in the sprung mount without hitting the frame and the tender wheels track much better.

The locomotives Whiton and Haupt do not have these wiper screws on the tender trucks. I added pickups to those locos using a different design that does not interfere. Only McCaullum came with the factory installed wipers, and the problematic screws. I wanted to keep the factory installed wipers. So I did not remove them, but I did add a second set of wipers using the same design I used on the Whiton and Haupt.

The McCallum came with  a pick- up on only one axle of the front pilot truck.  I added a second pick up to improve electrical contact, and replaced the factory installed one since it didn't work well and tended to droop and get caught on the turnout frogs.  I used small piece of PC board and 0.020 inch brass wire to rub against the flange of each wheel. I also added some sheet lead to weight the pilot to improve it's tracking.

One additional change I plan to do is to replace the drawbar between the tender and engine with an insulated piece instead of the current metal one. The original draw bar is metal. Since the frame on the tender is opposite polarity from the frame on the engine, the draw bar can short if the insulating sleeve on the drawbar stud slips down where gravity seems to want to pull it. My solution will be to laser cut 1/8th inch thick acrylic drawbars that will help insulate.

Below is a short video summarizing the odyssey.



The failed QSI  decoder on Whiton is on its way to Tony's Train Exchange for replacement. I haven't decided which decoder I like best, QSI or Soundtraxx.

I haven't had a chance to program the Tsunami with momentum or sound options, but the QSI decoder seemed to start smoother. I do like the extra sounds it comes with the SOundtraxx like the water spout, steam blow off hiss and the short whistle blast.

Update: I did a bit more tuning on the McCallum drive. I cleaned off some excess epoxy that found its way to the back of one of the drive wheels.  That helped. I also added a couple ounces of sheet lead to the tender. There is room for more lead too. I also oiled the journals on several of my freight cars. All these fixes helped allow the McCallum to handle 8 eight cars across the layout.  It is running real nice. My only complaint is that the "electronic" noise from the motor is louder than I would like. Do you DCC gurus have any suggestions on how to reduce that?

More fun than a barrel of....


 

New freight cars under construction

Finally...some time to work on the layout! I was able to build the chassis for two new flat cars. The flat cars are an experiment using acrylic plastic for the frames. I tried acrylic because I was using Grandt Line Stake pockets, as well as Tichy NBWs for this car. I reasoned that they would be easier to glue to the frame if I used acrylic. The problem arose in that the holes the laser cut for the stake pockets and NBWs were a tad small. So I had to open them up with the drill. When drilling acrylic the bit can foul with melted plastic, making the diameter of the drill too large. So I lost control of the dimensions of the holes. It was painstaking having to clean the melted plastic off the drill after each hole. Next time, enlarge the holes on the laser drawing. The other thing I wanted to test was how the acrylic plastic once painted takes decals. I have been having trouble getting the decal film for completely disappear on my wood sided cars. I just can't seem to get a good gloss surface on the wood to apply the decals.

Some of the barrels I stained and painted this weekend



I also stained and painted about 60 barrels. These are wood barrels that were tuned on a lathe. They are produced by John Rendal Scale Models. Walthers had them on sale, so I picked up a bunch. They took the stain pretty well. I then dry-brushed rust on the hoops and added some stave detail with a pen (not pictured).

One of the nice things about modeling the ACW in O Scale is that if you need cars for your layout, you can build them. One of the bad things about modeling the ACW in O Scaler is that if you need cars for your layout, you must build them. There are no RTR cars available (barring the SMR brass cars which are not very compatible with my wood cars).

A corollary of this phenomena is that trips to the hobby shop are considerably less exciting than they used to be as there rarely are new goodies in the shop that I can use on the layout. Nonetheless, I made a trip to Mainline Hobby Supply in Blue Ridge Summit on Saturday to pick up a new decoder and some other supplies. This is a great model railroad shop, especially for HO and N Scale modelers. Brian and Bonnie also do a great job on mail order items.  The drive up for me offers a treat as it passes through Owens Creek canyon where the Maryland Midland still runs on former WM tracks.



Maryland Midland entering the horseshoe curve at Sabillasville, MD on an earlier trip to Mainline Hobby Supply

Trouble with a decoder

Tonight the gremlins descended into my basement and started trouble with the QSI Revolution U decoder installed in the locomotive Whiton. It was working fine up until tonight. Now it will only run in one direction. All other functions seem to be OK, including sound and ops mode programming.

I tried resetting the decoder several times. I can get the decoder to speak back "reset," but resetting does not fix the problem. Note the QSI decoders have a talk back feature that helps in programming and debugging.

I made a video and audio recording of what the decoder says when I tried to check its speed using function 10. (See theyoutube video below) When I reverse the throttle, the loco doesn't move and it say something like "Vee -et" even though the throttle is 50 percent of max. In the forward direction it says the correct numbers as it speeds up. I think the "vee-et" sound from the decoder is how the QSI decoder pronounces zero.

The only other O scale DCC equipped loco I have runs normally, so it's probably not the Easy DCC system. Anyone that familiar about QSI decoders know what the problem might be?

Update: I spoke with Tony's Train Exchange and they will replace the decoder under warranty. Tony Parisi, the owner and DCC expert, believes that one of the output transistors has failed. But he suggested doing a hard reset by shorting certain output leads as mentioned in the user's manual. I will try that, but I am not hopeful. If the hard reset doesn't work, they will replace the decoder with a new series of QSI decoders called the "Titan." In either case I need to disassemble the loco.

This might be a good time to do a batch installation and get the last three locos converted to DCC.

More tuning and scenery work

Thanks to Linda Dolkos' Book Club, her husband Paul stopped by my layout for a work session. He brought over two Easy DCC throttles to test to see why I've been getting spurious results when I try to use two throttles at the same time. After checking things out, resetting the throttle addresses to null,  and reprogramming the throttle IDs, it appeared that things were running normally. We were able to run the two DCC equipped locos at the same time. Paul was able to run a 6 car train from Falmouth to Brook and back, while I ran the Haupt. Paul also fiddled with the momemtum settings on the Whiton, reducing it a bit to make the engine easier to control. I reminded him that he was the person who programmed the momentum the first time. :) It is cool to listen to the sound chip announced the programming changes through the speaker.

He then went to work on the hillside that he had worked earlier, covering the terrain with kraft paper and reshaping the slope to allow a little more room at the track lpevel.

Meanwhile, I did some more tuning. I re-gauged one of the turnouts at Stoneman's as it was tight. I also adjusted the ride height on one of my box cars. This car was riding high enough, that when it was pushed, the link from the neighboring car created an upward thrust that was derailing the car on some of the turnouts. By filing the laser cut acrylic bolster I was able to get the car to ride lower by about 0.040 inches. That seemed to help.

After that I painted some barrels and finished the windows and doors for the depot at Falmouth. I had to recut the door openings as I had made then too tall. It was a realtively easy fix.  I also cut an access hatch in the floor of the building to allow me to add some interior details. Now I need to come up with some way to detail the interior that is simple and quick. It is very hard to see into the building, but one can look lengthwise through the building as one gable wall is right by the aisle as you enter the room.
Finally I demonstrated the laser cut iPhone tripod adapter I made for holding my iPhone. At some point I'll post some test videos.

HD Video Test

I was able to borrow a Canon EOS 60D from my brother to test its 1080p video capture features. I shot this video in HD, but was unable to upload it to Youtube in HD using the version of iMovie that I own. The highest resolution is 360P.

I subsequently learned how to export an iMovie using QuickTime and was able to upload a HD version to youtube. This one should be viewable in high definition modes. The viewer can select the definition desired when watching the video. At the time I uploaded this video the highest resolution Youtube  allowed was 480P. I am not sure why, as it should be 1080p.I think this is a problem with my version of iMove 7.1.4, as it only allows exports of 920x680, which youtube down samples to 480p. I may have to try a later version of iMovie, as we have a copy on my wife's laptop.