One guy's trials and tribulation

[John]

I found this in my yutubes this morning .. interesting commentary by this fellow

[Philip H]

I've been subscribed to his channel for a while.  He's got an interesting modeling arc.

[Dwight in Toronto]

I found this to be remarkably similar to what I went through several years ago. 

The parallels between what we each eventually accepted as “final straws” that culminated in layout tear-down, and the resolutions made as to lessons learned and techniques to be incorporated in a subsequent rebuild, are uncannily comparable.

I don’t know this fellow, but his focus on Union Station, GO commuter trains, The Canadian etc strongly suggest that he is a fellow greater Toronto area resident.  Each of those aspects have always been of great interest to me as well.  For these reasons, I’ve just subscibed to his channel.

John, this was an interesting find - thanks for sharing!

[John]

You are welcome .. I found it educational the way he did things ..  and some things I cringed ..

1) Acid core solder
2) Super glue to glue track down - could be a big reason why his track bowed with expansion .. especially if it was glued to cork
3) The observation on servo's was intersting .. I guess it might be QC on the manufacturer

I wish him lots of luck on the next layout

[randgust]

It's hard to top Jerry Britton's PRR nightmare.

I got to see that layout firsthand just before things went terribly wrong.   Unbelievable.   Most magnificent, complicated, multi-level and accurate PRR N scale layout I've ever seen.   HUGE floor level multi-track staging yard, a two-track helix to bring it up to various levels, a full-size layout of Harrisburg station, horseshoe curve...  big parts of it were running.   I remember watching a Centipede dragging a train up the curve and just marveling.  And it went up to a level just above eye level where you had to 'step up' to see it.   I was awestruck and that doesn't come easy.

If you remember years ago he did the Duncannon bridge module that was spot on and magnificent.  I helped with the backdrop.

And it all failed, on self-disintegrating Atlas code 55 switches, mostly on failed solder joints.   Atlas promised replacements, but he'd already built in so many that were virtually inaccessible for replacement down at the floor level staging yard that he just sorta lost it and junked the entire project.

I couldn't believe it.   And he left N scale, too.    That still stands out as the hardest fall I've ever seen anybody make on a semi-finished project.

I've been lucky.   After several HO layouts, and a 3x6 door layout in N, the layout I started in 1983 for the ATSF 3rd District is still running, and running well, mostly due to extreme overbuilding and design standards from the start.   Yeah, I'm stuck in Code 80, but man, it still runs well, and what few design mistakes I made have been rectifiable.   But nothing has cropped up as a fatal flaw requiring walking away.   I just keep improving it, but the basic track and benchwork, yeah, solid.

I really, really, recommend the following things:   Stranded wire, soldered jumpers to all switchpoints, and 100% solder joints on rail for about 24" sections with soldered drops around rail joiners.  And I designed the entire layout semi-modular in 4' chunks, so it's been in three houses now without getting demolished.   Oh, and paint everything, and get a good dehumidifier.

[John]

And it all failed, on self-disintegrating Atlas code 55 switches, mostly on failed solder joints.   Atlas promised replacements, but he'd already built in so many that were virtually inaccessible for replacement down at the floor level staging yard that he just sorta lost it and junked the entire project.

I couldn't believe it.   And he left N scale, too.    That still stands out as the hardest fall I've ever seen anybody make on a semi-finished project.

Lee @wm3798 and I picked up a lot of his switches for a few bucks each .. most of them are still working ok on my layout .. although the last few years I've been modifying them by adding PC ties in areas where the fails normally happen.  This week I modified a #5 by removing the frog, drilling a small hole on the underside of the frog to allow for solder to catch, then added a #28 stranded piece of hook up wire .. also drill the hole directly underneath in the ties .. after testing for continuity, mix up some epoxy and mount the frog back on the ties .. in addition, add the pc ties where needed ..

[randgust]

That's a relatively happy ending, at least for the material.

I had another good friend - now passed - that managed to build 'two in a row' of enormous layouts with fatal flaws of design and execution.

First was a magnificent, full basement layout of B&O steam era capable of running 100 car hopper trains behind a 2-8-8-2.  We did it.  1% grades.

And it got all the way through scenery, which he was actually pretty good at.   Mostly in the late 70's, early 80's.   But he used that sticky rubber 'ballast' subroadbed instead of cork, and it had two weird quirks.   First, when you curve it, it created reverse superelevation, so everything tilted to the outside....  and then the stuff dried, and shrank.  And with no soldered rail, all flextrack, it started popping track joints.    Pretty much unfixable on a layout that size.  Scenery was dense eastern forest in mountains with garbage bags full of lichen, you couldn't get in to fix much, particularly in the tunnels.   Oh, and laid on homasote that also moved and wouldn't hold spikes or track nails either.   But holy smokes, when it ran, first 100 car train I ever saw in N.

Second one in a new house, basement sized.  He loved scenery, so he decided to charge in really before layout shakedown.   With tons of trees, plaster, everything.   THEN started to test the layout for operation.   OMG, nothing worked.  A basement sized layout, but with all the dampness from scenery, lots of plaster in rails, turnouts (he hadn't taped anything) you couldn't run a train.   And he was developing arthritis so he could get neither on top of the benchwork or under it.   Mostly it just sat there. 

My youngest son, who was about 14, actually got his first 'job' repairing the layout, as he could climb all over and under the thing, knew how to run a meter, and could solder.   It took almost two years, but he got the layout to the point a train could make it around it.   We finally got it to run, and six months later he passed.  My son now does technical electrical repair for a defense contractor with the skills he developed!   Two of my three sons (now adults) have told me the best skill I ever taught them that never came though school was running a soldering iron and a resistance meter.

The survivability of my own layout is at least partially due to watching my friends chronic fails, particularly leading to soldering everything, don't use homosote, and for pete's sake, shake everything down for a year before you put on finished scenery!

[Dave V]

Nothing, nothing, nothing will kill the joy of model railroading more than unreliable trackwork.

I'm far from the best tracklayer out there...but I would sooner sacrifice appearance to ensure that whenever a train runs, it remains running until I decide otherwise.

That's why the Colorado Midland v1.0 was sacrificed in a neighborhood bonfire (after being denuded of toxic materials of course!). Like Jerry I had Atlas track that exploded or wouldn't take solder well.  My current two railroad projects will happily support trains running unattended for hours, and that's the way it has to be for me.

So I totally get it. And this is also why I'm cool with Unitrack for my T-Trak project. It's Dave-proof.

Starting over again--while very frustrating--does, as he points out, present new opportunities.

[wm3798]

I got to see Jerry's layout before he dismantled it as well.  If I'm honest, the problems he had with the turnouts were not all the fault of Atlas.

His design was flawed in that it had way too much inaccessible track.  Great for realistic operation, horrible for installation and maintenance.
He also relied on track nails... that's right, track nails, to secure the flex to the roadbed, and in some instances the turnouts.  On close inspection, you'll find holes poked in the ties of many of the turnouts.  I'm not sure how he applied the nails, whether they were gently pressed in with a nail set, or if he used a 5 bound maul.  But I can tell you, I had pictures where you could see the alignment was less than perfect, with the rails moving back and forth with the nails being the apparent culprit. 

Regardless of how the nails were applied, those turnouts are tightly engineered and pretty fragile.  Whaling away on nails in or around them is going to jostle some things loose.

Properly installed, even those old first run turnouts can be quite durable, witness the Hagerstown/York/Hanover yard complex that was built over 15 years ago and is till holding forth after several moves.  The repairs that @John has done should make them all the more durable.

There's a benefit to slapping down track to get trains running to be sure, but there's a much bigger benefit to taking your time and doing it right.

As my favorite carpenter's adage goes, "If you don't think you have time to do it right, what makes you think you have time to do it over?"
Lee

[Dave V]

Wise per usual Lee.

[wm3798]

Just soaked in the video.  That could have been me in 2011. 

While I didn't have the problems with the track that he did, I definitely found that I had designed some major problems into the layout.  The 10 lbs of layout in a 5 lb room being chief among them.  Certainly the volume of hidden track was a growing concern, and access for maintenance was a hiccup from the beginning owing to the sloped ceiling of the attic space.  (recall that I had built large chunks of the layout out in the garage, where I had access to all four sides, then installed it where it was all I could do to reach one without hitting my head on the ceiling)

Like a dummy, I did the same thing to myself with the HCD by placing the skyboard down the middle and pushing it against the wall into a corner.  I'll give you three guesses as to where most of my track problems are, and the first two guesses don't count.  When I added the yard and a freight siding, I had to do some acrobatics to cut in the switches, and ended up queering the alignment enough that I barely run the outer loop anymore.

The good news is, that as a HCD, I can pull the layout away from the wall and get back there to tidy things up.  But I'm lazy, and also a slob, so it will be a bit of an effort to actually get that "round tuit".  I've also carefully avoided any more electronics than is necessary, although with the yard, I'm finding that while I have the layout pulled out, I may cut in a couple of more blocks to allow the traffic to flow more efficiently.

I don't think I need to tear it down and start over, I just need to fine tune a couple of things to get it back to where I can turn a few trains loose for some JFRTM, which anymore is my primary operating goal.

Lee

[randgust]

Nearly every fatal flaw he hit in there on that video was a near-miss by me.    That old Atlas C80 in my hidden staging yards has held up well.  I made sure that the turnouts were really accessible toward the edges, I've had a couple fail but it's pretty amazing, really.

At the time (1983), the only switch machines that were considered 'bulletproof' were Lamberts - the big HO ones.  I made steel wire and brass tube linkages and junked all the above the table tiny solenoids.   And mounted the Lamberts so I could drop them out from below for maintenance.   Held up well.   Since then I've played with Rix machines that are the same design.  With electrical contacts to show switch position, don't want any more digital electronics.

I've had Tortoises fail on my portable modules, but again, with the screws and bracket designs, they come out fairly easy.  But not as durable as the old Lamberts for sure.

Never realized how much of a bullet I dodged by staying with Radio Shack rosin core solder..... or by doing everything in 4' modules where I HAD to have expansion joints whether I liked it or not.   But the old Atlas and Trix c80 flex has held gauge without issues.  What I'm using now is Peco C55, which is just as durable, anyplace I make changes.  And it's all on cork, on thin plywood, ballasted.  It doesn't move.

I'd mounted my Masonite backdrop panels on furring strips, and yes, the darn things expand and contract seasonally.  Never beat it.  I'd never use Masonite again, even with humidity controls. 

Turnouts?  With the point jumpers, the old Atlas C80's have held up, and where I've improved things, Peco C55 electrofrogs with the springs pulled and manual Caboose Hobbies throws on them. 

So yeah, I look at a lot of stuff that's dated, but darn, it still runs really, really well, this many years later.

[Scottl]

Thanks for sharing the video, it was well done and raises a lot of interesting points that many people planning a layout should consider.

I have a lot of sympathy for his experience and problems.  There is a real culture of (well-meaning) advice and peer pressure out there that encourages us to build overly complicated layouts.  As Lee @wm3798 puts it, 10 lb in a 5 lb space.  But the "operations" thing that is all the rage also drives many layouts to be much more complicated than they probably should be.  Hidden storage, complicated track and benchwork structures are all red flags for me, especially with the attendant scaling of hardware and electronics.   Some people can pull this off with meticulous construction and planning, but I'm not one of them and I had to resist all of these urges when building my new layout.  Having said that, if code 80 was the only way to reliable N scale, I would have moved on to HO or something else.  Once you use finer rail, everything else looks way out of proportion, even if 55 is itself oversized for some uses.

I personally have had very little trouble with Atlas code 55 turnouts and have reused many over 5 layouts spanning 15 years, but yes, a few have failed due to a moment of heavy handedness.  However, in general I find Atlas code 55 pretty reliable and problem free.  I sure don't solder the ends of my rails- that seems like work to me- and it does nothing for reliability that the joiners don't already.  Similarly, I have minimized my soldering of the track in every way possible.  I use the stock turnouts, in most cases have just stopped bothering to to wire the frog, and I have a power feeder to every section of track between turnouts (not every separate piece). My main reason is that soldering track leads looks bad, it is difficult to do on the rail and just adds complication.   I've never had conductivity issues on a length of track and frogs are rarely an issue, but I also keep my track clean with a coat of No-ox and periodic wipes with alcohol.  My slightly sloppy track work has enough small gaps in it that I have never had any kind of deformation due to humidity and temperature changes, but my layouts have always been in stable basement settings.

I have gone the route of servos on my layout and so far like them a lot.  I have had no servos fail and they all work fine.  How they are linked to the turnout is a big deal and even subtle tension on the piano wire will stress the servo, and perhaps even the points themselves.  Also, if you don't carefully define the end limits for the servo travel, you are definitely stressing these components and I expect failure is inevitable.  The thing is, I expect these ebay specials to fail so I plan for it (I mean, they are like $1.50 each).  I used wiring harnesses for all of the electronics and have not had any issues, and if something needs replacing, it is a no-solder job.  I have a few servos that buzz, but that is a minor concern that I rarely notice.  I minimized complexity in the system with a single button to flip the turnouts and no lighting indicators, all of which would add I/O lines to the Arduinos and massive added wiring complexity.  Putting this stuff in hidden trackage where it is difficult to access is a recipe for problems, to say the least.

One observation I have is that I realize now that I treat all my trains with kid gloves.  I handle everything with great care, don't leave s**t all over the place so it causes incidents, and I don't run my trains at high speeds.  For a long time, I assumed everyone else was doing the same.  But after going to some shows and watching the modular operators and a few private layouts, and seeing how trains are stored on workbenches that look like nothing is ever put away, I get the sense that some people are rather (often inadvertently) hard on these expensive, delicate little models we have.  I suspect it was sped up for presentation, but the video showed trains roaring away at track speeds that invite turnout and other issues. 

The last thing I wonder about:  what is acid solder flux for if not electronics?  In my early experiences, it was sold for that purpose.  Now I use acid-free flux but my many uses of rosin core flux for electronics and trackwork in the past have not led to any noticable failures.  It is all rather mysterious to a casual user of a soldering iron.

[C855B]

... The last thing I wonder about:  what is acid solder flux for if not electronics? ...

Stained glass came, for starters.

I've had some trouble with Atlas C55 turnouts, where the jumper from a stock rail loses conductivity to the corresponding through rail. These are usually bad right out of the package, so I (try to) remember to check continuity before laying and do a mechanical fix. Servos have worked flawlessly, even given the screwball mounting system. I've since modified the solid linkage shown there to be a hairpin spring, which doesn't need as much tweaking of servo limit settings.

The layout is constructed with low maintenance in mind. No wood, for one thing, which is where the expansion/contraction happens. It's not the rail. Then I pay close attention to reach. The one area where I knowingly violated my reach rule... wait for it... is the one area where the track is never clean enough and everything stutters to get through, despite multiple passes with the track-cleaning train.

Environment is important, too. A whole-system dehumidifier keeps things at or below 50%RH. Double filtration handles the dust - it's not exactly industrial clean-room quality, but close enough.

[Scottl]

Stained glass came, for starters.

Environment is important, too. A whole-system dehumidifier keeps things at or below 50%RH. Double filtration handles the dust - it's not exactly industrial clean-room quality, but close enough.

Interesting, I guess it is not an issue in that use.

I whole-heartedly agree with the environmental control and conditions.  I keep my layout room door shut, I've sealed off the HVAC and use a heated rug when it gets a bit chilly in the depths of winter.  I also vacuum frequently with an old HEPA filter type that stays out and plugged in.  Compared to the rest of the house, that needs frequent dusting, my room is very free of dust.  Everything stays cleaner, especially the trains and track.  Cleanliness is a big positive factor for keeping the room and layout free of dust.