Christmas Layout based on Ken H's 5x17... SCARM gaps OK?

Hi Skip, glad you liked the plan.

The version you are looking, was one done for the discussion, and had not been gone over and "perfected".   I took a few minutes and slightly revised the plan, to 1/16" tolerance - .0625" (less than your setting of .079") (the plan is perfect in degrees, you could set your tolerance to 0 degrees).   There is one connection that needs 3/32 to close, (.09375") noted on the diagram.

The revisions I made is in three straight sections that connect the layout.   Here is the slightly revised more perfected layout.

The smallest straight section difference is 1/8", so many plans are difficult to close less than 1/8".   I work the choice of curve diameters and their order/arrangement, in order to get plans to close at smaller tolerance fractions.   Fastrack will tolerate 1/16 and 3/32 quite well in actuality.  There is some variance in manufacturing (not surprising) and sometimes just two straight sections will close leaving a tiny but visible close-up, gap.

Wow.  I can't thank you enough Ken!!!  This is really great!

Going to go through my inventory of track pieces and make an order asap...

Skipdup posted:

Wow.  I can't thank you enough Ken!!!  This is really great!

Going to go through my inventory of track pieces and make an order asap...

Great!  Check the above post updated for the latest version.   Please let us know how it turns out!  -Ken

Best I can do at the moment, is to close that middle straight at 0.09".   I had to make a few other minor changes in different places.

Thank you so much Ken!  I'm replicating in SCARM now and will post the file (for posterity) as soon as I can.

I will definitely post pics and/or video once it's all done...

Well, it's strange.  I've triple checked myself.  But, I'm still getting gaps.  Changing setting to 0.16" "disappears" them.  Wondering if this is just differences in software and if it's close enough...

Attached is the SCARM file & some screen shots...

Skipdup posted:

Well, it's strange.  I've triple checked myself.  But, I'm still getting gaps.  Changing setting to 0.16" "disappears" them.  Wondering if this is just differences in software and if it's close enough...

Attached is the SCARM file & some screen shots...

 

I have not checked your SCARM file.   But a comment from using AnyRail (and perhaps is similar with SCARM).  If you make connections that are not perfect, the computer maintains those variations in connections.   Multiple variances can accumulate into a more significant difference.   After I work out a plan, I take if apart and reconnect it (sometimes needing revisions), to eliminate those accumulated differences.   Sometimes I have to do that more than once.  Perhaps that helps?

Another idea that may, or may not, be helpful:  when I take apart a plan, I also reconnect it in the "constrained" sections.   I build those sections first, then reconnect. This idea:

0.16 is more than 1/8", so I would keep working it to see if you can get closer.

Perhaps you will find this version is better - I found a place to combine small sections into one piece, and I reconnected.   The center straight section is now connecting at a tolerance of 0.039 (smallest AnyRail allows me to set).   So far, after more work, everything seems to be assembling "perfect" or within that 0.039 tolerance.

Ken, I GREATLY appreciate your help.  I sent you an email a couple days ago.  Just saying in case spam caught it.

I'm starting to think there might be some kind of user error on my side.  Or, maybe we're discovering some weird variance in how SCARM & AnyRail perform.  I've been taking this apart and putting together every way I can think of.  This is starting to make me a little whacky feeling.   

The latest version does improve, but I'm still at 0.15" (changing to 0.151" disappears them) in two places - both at the top coming off the turn-outs.  

I'm debating just buying & building off your last AnyRail plan.  As a general rule, if AnyRail says it works, does it work in real life?

Thank again.  I'm sincerely grateful.

- Skip

Well, somehow I stumbled into a "solution"???  It's now within 0.075" on SCARM.  

Ken, I honestly don't know how you do this.  Dang...

I just worked an experiment: I replicated one of the inside reverse loops in SCARM.   It does not close.   So, one of these two track planning software systems is off a bit.   I don't know at this point in time, which is in error.   I am seeing a small rounding-type error with AnyRail in the small straight sections, but not enough to account for this degree of difference, even if accumulated (0.001").   

This may be why you (Skipdup) cannot connect the layout in SCARM.   Will look at some more later.

Some evidence that SCARM has errors:  A length of track that should measure 32.625", is repeatedly measuring at 32.68", measured on SCARM using its measuring tool.  That 0.055 seems small, but errors accumulate, and can add up to a noticeable difference.  [1-3/8, 1-3/4, 4.5, 10, 15" turnout, 5 sections].   Perhaps its a measuring error with the tape measure.  0.055 is more that 1/32 (0.03125).   If this is real, over 20 sections it could add up to 1/8" or more.   IF...   I tried this a few times, and still getting the same measurement, building with new track sections, and multiple measurements.   Hmmm.

Carefully measuring the same set of tracks with AnyRail, measures at exactly 32.625 (had to zoom in to be precise).   

Seems to be measurment errors with the SCARM tape measure, now getting inconclusive results, more like 32.64" (0.015 error).   So this is inconclusive, no reason to suspect SCARM based on this test.  Perhaps one or the other system has errors in the curved sections? - doesn't seem to be, both measure an O48 circle precisely.

So what could be going on?  If the difference is not in the straights, and not in the curves, that would leave the diverging side of a turnout.  I know that an O48 turnouts fits in perfectly in AnyRail into a circle, as it does in SCARM - just tried it.

Attached is my "working" SCARM file...

6This is one of the VERY BEST threads I seen in a long-long time.  All because of a topic that Ken started in January of 2018.

Skip liked it, remembered it and chose it as his Christmas layout -2019.

Ken spots this thread then goes about and puts forth the effort to revise and perfect the plan by reducing some of the alignment issues.  

Results... Skip's job will be easier and Ken, along with the rest of the forum members, will get to see this plan "come alive".

Now that is what Christmas is all about !!!

SANTAFEJIM- AMEN!  Ken's willing to put who knows how much time into helping a stranger is just the coolest.  And his plans are really amazing.  No way I get 10% of this without his help (and that's probably giving myself too much credit).  Very, very cool!

So, I woke up this AM thinking about this.  The two plans are identical, except for a small straight section on the return loops.  The variance between the 2 programs comes down to AnyRail calling for 3 pieces (one 4.5" and two 1.37" totaling 7.26") and SCARM calling for 4 pieces (four 1.75" totaling 7"). 

SCARM measures the actual gap at (approximately???) 7.1".  I'm not sure how to measure a gap in AnyRail.

When building, I'm going to try and remember to experiment with both and see which one is more "right"...  Maybe they're both right???  

Skipdup posted:

So, I woke up this AM thinking about this.  The two plans are identical, except for a small straight section on the return loops.  The variance between the 2 programs comes down to AnyRail calling for 3 pieces (one 4.5" and two 1.37" totaling 7.26") and SCARM calling for 4 pieces (four 1.75" totaling 7"). 

SCARM measures the actual gap at (approximately???) 7.1".  I'm not sure how to measure a gap in AnyRail.

When building, I'm going to try and remember to experiment with both and see which one is more "right"...  Maybe they're both right???  

Good analysis Skip!   My construction confirms your analysis/experiments.   So that implies that one (or as you say, perhaps both) are off on their straight sections.

AnyRail has a nice ruler object that you can place and stretch to measure precisely.   It stays in place, and does not snap back like the tape measure.

So I measured the track sections (curved and straight) in both systems, and both are correct, using their own measurement tools within the software.   Perhaps one system is adding some distance in making the joints?  As I mentioned, the actual track sections don't always butt/join perfectly, sometimes there is a very small gap detectable looking closely.   But so small its hard to imagine that this could accumulate into a measurable difference.

The gap I measure on the track loop on AnyRail is 7.227".

So it makes sense that in AnyRail, 7.26 will bridge the 7.227" gap.  And on SCARM, 7" will span the 7.1 gap.

Likewise, it is sensible that 7" will not span the AnyRail 7.227" gap.   And 7.26 is too big for the 7.1" gap on SCARM.

Interesting, but does not tell us where the variations originate, or which is the most accurate.  I look forward to the results of your experiments with the actual track.

Both SW systems use math to represent the track sections, and there must be a precision (number of decimals) limit in each system.   The rounding errors might accumulate a difference.   It doesn't seem that the location on the screen/grid makes a difference.

BTW, I started a conversation with the author of AnyRail, on their forum, about measurements of track sections and precision.  The author is indicating that the track dimensions and length are accurate to 2 decimal points in millimeters, using the data from the manufacturer.  NO, using actual track measurements from batches of track!

Great info Ken.  Thank you.  Yes, I too am looking forward to seeing what actually happens in real life.   I may build this first in the master bedroom so I don't clutter up the living room before Thanksgiving...  I'm anxious to see this...

Hoping to do my inventory later today (this weekend at the latest) and get my order in asap.

Already been thinking about modifying the layout with an elevated section.  Because, the enemy of good is better...

Interesting that AnyRail actually tests batches of track from Lionel, and uses the actual manufactured track dimensions, which vary slightly due to the manufacturing process.

David from AnyRail:  "For instance, Lionel O gauge part 65505 is in the catalogue as 5.5" (13.97cm), but in reality it is 5.55" (14.1cm)."

That variation (0.05") is between 1/32" (0.03125") and 1/16 (0.0625).   

Unless batches vary dramatically, this implies that AnyRail should be close to the actual real track layout dimensions.

1/8" is 0.125", less than three of these variations of 0.05".   This would account for the difference we are seeing, that is, if SCARM uses the manufacturer published dimensions, and does not test actual track.  (!)

I could not blame any SW system for using the manufacturers published specifications.   And some variation in manufacturing is to be expected (tough to manage quality control in China, as we know).    AnyRail is going above and beyond to measure actual track for its software, I would not have expected that.

One conclusion I come to, is that assembling a long length of track of many sections, and expecting it to close in very small tolerances <1/8", may be an exercise in futility.   One may just have to see what actually works "on the ground".   Second conclusion is that AnyRail should be close to the actual physical track, but with manufacturing variations, perfection is not to be expected.

I have some tiny "Z" scale trains, and one of the popular roadbed track systems has an "expandable track piece":

Would be nice if Lionel were to make a 1-3/4" track section that be stretched between 1.5" and 2".   Would solve a lot of problems!

Just placed my order.    

So, this is probably a silly idea...  I started wondering if I could incorporate a tiny (N scale??) Time Saver inside one of the loops.  I think my kiddos would have fun with that.

Is N gauge even good for something like this?  Searching hasn't really found an answer (which makes me think it might not be good).  

Thanks,
Skip

I suppose the plan could be modified so the two loops also are independent of each other so two trains can be run without worry about them getting in each other's way.

Skip,

Timesavers in N scale are easy enough to build, but not to operate, because it requires a delicate touch to manually uncouple the cars.  And, the very nature of Timesavers makes them impractical for remote control uncoupling.

Chuck

That's a bummer.  For whatever reason, N scale intrigues me.  Or, maybe the thought of it intrigues, since I've never actually seen one...  

Thanks!!

Skipdup posted:

So, this is probably a silly idea...  I started wondering if I could incorporate a tiny (N scale??) Time Saver inside one of the loops.  I think my kiddos would have fun with that.

Is N gauge even good for something like this?  Searching hasn't really found an answer (which makes me think it might not be good).  

Thanks,
Skip

N scale within or on an O scale layout represents the size of an amusement train - so, they look good - gondolas with benches and O scale sitting people

Timesaver may be too complicated - why not an Inglenook?

The library dimension accuracy  conversation is interesting - but, I was able to solve the fitment areas - I zoom when I use the measure tool- I use the default .079" tolerance in the edit settings of SCARM. I use the FasTrack lengths combination chart to find combinations to fill gaps.

I have attached the SCARM file and the combination length pdf.

SCARM only uses dimensions provided by the manufacturer - so, production variances would be an issue - but, I have built many plans from most O gauge track manufacturers and  a SCARM design and have had no fitment issues with the actual track

PRR1950 posted:

Skip,

Timesavers in N scale are easy enough to build, but not to operate, because it requires a delicate touch to manually uncouple the cars.  And, the very nature of Timesavers makes them impractical for remote control uncoupling.

Chuck

Not so!   Microtrains magnetic couplers are reliable and work quite well for switching.   Personal experience.   Plan out the locations for your uncoupler magnets - wherever you would put a remote uncoupler.

Remote couplers on the locos like O scale are not available though, which limits the fun a bit.

Carl- That length chart is VERY helpful.  I looked for one last year but never found.  Thank you!!

Also, really good to know SCARM has worked well with different plans!

Thanks!

Ken-Oscale posted:

I just worked an experiment: I replicated one of the inside reverse loops in SCARM.   It does not close.   So, one of these two track planning software systems is off a bit.   I don't know at this point in time, which is in error.   I am seeing a small rounding-type error with AnyRail in the small straight sections, but not enough to account for this degree of difference, even if accumulated (0.001").   

This may be why you (Skipdup) cannot connect the layout in SCARM.   Will look at some more later.

Some evidence that SCARM has errors:  A length of track that should measure 32.625", is repeatedly measuring at 32.68", measured on SCARM using its measuring tool.  That 0.055 seems small, but errors accumulate, and can add up to a noticeable difference.  [1-3/8, 1-3/4, 4.5, 10, 15" turnout, 5 sections].   Perhaps its a measuring error with the tape measure.  0.055 is more that 1/32 (0.03125).   If this is real, over 20 sections it could add up to 1/8" or more.   IF...   I tried this a few times, and still getting the same measurement, building with new track sections, and multiple measurements.   Hmmm.

Carefully measuring the same set of tracks with AnyRail, measures at exactly 32.625 (had to zoom in to be precise).   

Seems to be measurment errors with the SCARM tape measure, now getting inconclusive results, more like 32.64" (0.015 error).   So this is inconclusive, no reason to suspect SCARM based on this test.  Perhaps one or the other system has errors in the curved sections? - doesn't seem to be, both measure an O48 circle precisely.

So what could be going on?  If the difference is not in the straights, and not in the curves, that would leave the diverging side of a turnout. I know that an O48 turnouts fits in perfectly in AnyRail into a circle, as it does in SCARM - just tried it.

 

O60 turnout is the only one that doesn't work (fit as an arc of a circle). Weird design by Lionel.

I am trying to recall if the Anyrail divergence automatically places the 1 3/8" piece. In SCARM it does not automatically insert those pieces - this is on O60, O72 switch, O72 wye and 22.5° cross. An O72 circle won't join without a 1 3/8" half-roadbed

Skipdup posted:

Carl- That length chart is VERY helpful.  I looked for one last year but never found.  Thank you!!

Also, really good to know SCARM has worked well with different plans!

Thanks!

 

In general, gaps that are stars or slightly misaligned will build ok - however, the perfect joint fitment is needed to run the simulator in SCARM or the sim train will stop at a gap or misalignment.

Some increase the tolerance in the edit settings - it works ok for loose fitting tubular O - not so good for track with roadbed - I leave in default and have adapted to working with the very close tolerance

Moonman posted:

Ken-Oscale posted:

I just worked an experiment: I replicated one of the inside reverse loops in SCARM.   It does not close.   So, one of these two track planning software systems is off a bit.   I don't know at this point in time, which is in error.   I am seeing a small rounding-type error with AnyRail in the small straight sections, but not enough to account for this degree of difference, even if accumulated (0.001").   

This may be why you (Skipdup) cannot connect the layout in SCARM.   Will look at some more later.

Some evidence that SCARM has errors:  A length of track that should measure 32.625", is repeatedly measuring at 32.68", measured on SCARM using its measuring tool.  That 0.055 seems small, but errors accumulate, and can add up to a noticeable difference.  [1-3/8, 1-3/4, 4.5, 10, 15" turnout, 5 sections].   Perhaps its a measuring error with the tape measure.  0.055 is more that 1/32 (0.03125).   If this is real, over 20 sections it could add up to 1/8" or more.   IF...   I tried this a few times, and still getting the same measurement, building with new track sections, and multiple measurements.   Hmmm.

Carefully measuring the same set of tracks with AnyRail, measures at exactly 32.625 (had to zoom in to be precise).   

Seems to be measurment errors with the SCARM tape measure, now getting inconclusive results, more like 32.64" (0.015 error).   So this is inconclusive, no reason to suspect SCARM based on this test.  Perhaps one or the other system has errors in the curved sections? - doesn't seem to be, both measure an O48 circle precisely.

So what could be going on?  If the difference is not in the straights, and not in the curves, that would leave the diverging side of a turnout. I know that an O48 turnouts fits in perfectly in AnyRail into a circle, as it does in SCARM - just tried it.

 

O60 turnout is the only one that doesn't work (fit as an arc of a circle). Weird design by Lionel.

I am trying to recall if the Anyrail divergence automatically places the 1 3/8" piece. In SCARM it does not automatically insert those pieces - this is on O60, O72 switch, O72 wye and 22.5° cross. An O72 circle won't join without a 1 3/8" half-roadbed

Carl, I think we had this conversation before.   AnyRail does not automatically insert the 1-3/8" fitter, but when you add an O60 turnout to the layout plan, you add the fitters that are in the package from Lionel, as well as the turnout.   They are not automatically connected, and one is not required to use them in AnyRail, or in "real-life".

It is possible to use the O60 and O72 turnouts without the fitters, you have to notch the roadbed of the adjoining piece, not hard to do.   I try to include the 1-3/8" fitters in my design so folks don't have to trim.

Finally got time to put the pieces together.  As is typical, had issues...  1st, I didn't order two pieces of track - miss counted somewhere.  2nd, two of the turnouts aren't working.  

I received the two curve pieces I needed over the weekend.  Unfortunately, I am not able to resolve whatever is wrong with the two O48 turnouts.  They won't switch, have sloppy gaps (causing derailments) and short when a train goes over them - I'm at a loss.  Hopefully the other two O48 turnouts don't wind up being bad as well, since I've not been able to throughly test.  Two replacement turnouts are (hopefully) in the mail.

Regarding the "gaps", one side is done how SCARM specified and the other is done how AnyRail specified.  I originally used four 3/4" on both sides - but the right side was "gappy", I changed that side to 4.5" & 3/8.  Frankly, I'm not sure now what I think.  More than anything, I think there's probably far more slop in the track than either AR or SCARM represent. 

Attached are some pictures of the layout, the "gap" sections, and my n-scale layout (the only one running a train so far).

Thanks,
Skip

Skipdup posted:

I received the two curve pieces I needed over the weekend.  Unfortunately, I am not able to resolve whatever is wrong with the two O48 turnouts.  They won't switch, have sloppy gaps (causing derailments) and short when a train goes over them - I'm at a loss.  Hopefully the other two O48 turnouts don't wind up being bad as well, since I've not been able to throughly test.  Two replacement turnouts are (hopefully) in the mail.

Regarding the "gaps", one side is done how SCARM specified and the other is done how AnyRail specified.  I originally used four 3/4" on both sides - but the right side was "gappy", I changed that side to 4.5" & 3/8.  Frankly, I'm not sure now what I think.  More than anything, I think there's probably far more slop in the track than either AR or SCARM represent. 

Attached are some pictures of the layout, the "gap" sections, and my n-scale layout (the only one running a train so far).

Thanks,
Skip

Skip, thanks for keeping us updated.  Very sorry to hear of the problems with your O48 turnouts.   Sometimes I have to work to bend the points to make a better fit against the rails.

You might try the same lengths on both sides of the layout:   (4.5" & 1-3/8"  vs  4 x 1-3/4")  That might reduce the gaps you are seeing some, which look pretty small based on the pictures.

I agree that there are manufacturing variances with FasTrack that will be problematic in trying to achieve a perfect fit.  So far, I have suggested two solutions:

1.  Long-term: Lionel could produce a variable-length track section (as in the Rokuhan z-scale section).   Perhaps between 1-1/2 - 2".  Picture above.
2. Custom-length sections could be created to close a gap, using 1/16" gray plastic shims between track joints and super-glue, requiring that the FasTrack plastic ball-and-socket and alignment whiskers be trimmed.   Multiple shims over multiple section joints.  I have not yet done this to confirm that the idea is practical.   The idea is that we cannot compress FasTrack, and it is difficult to cut-down FasTrack into shorter lengths.   So building-up slightly longer custom track sections of glued short sections may be the next best option.