Single direction points

[RW James]

I actually saw this done on a very nice display layout and he got around the mainline polarity problem with a neat little trick - he reversed the wiring on one of the locomotives so it always ran opposite the other.

 

True - he had to use magnetic reed switches to start and stop the trains, but it worked flawlessly and virtually unattended. He had two sidings on opposite sides of a convoluted loop and the trains would wait at each station while the other train made its journey. It was always a big hit at the shows.

[Prof Klyzlr]

Dear Rick,

 

Tis true that reversing the motor polarity in a loco will achieve "2 locos run in opposite directions".

 

However, given that running at an exhibition can be stressful enough as it is,

having to ensure that one puts the right loco on the right train first-time, everytime,

for operation/reliabilities sake,

is IMHO "making an (unnecessary) rod for one's own back"...

 

Furthur, lets say your loco fleet "dies" mid show,

(most exhibitors have at least one "horror story" that goes along these lines... )

 

if you _must_ have a loco "wired reverse" for the layout to work "as intended",

then you've just made it impossible to solve the "all locos dead" scenario by simply heading to the nearest trader,

buying 2 "off the shelf locos" of vaguely suitable type,

and putting them both straight to work...

(giving you a chance to calmly sit down backstage, and work out what went wrong with the "stars of the show"...)

 

As the great Live Event Producer said "the show must go on..."

 

I personally avoid detection systems such as magnets/hall effect sensors for the same reason...

(Not that I often do, but I'd like the comfort of knowing I can drop any "foreign" equipment onto any of my layouts,

and they won't immediately tech-fail due to not having magnets or other "mandatory mods" the layout requires...)

 

Just my $0.05c

 

Happy Modelling,

Aim to Improve,

Prof Klyzlr

[shortliner]

Hi, John - and welcome! When you have had a chance to poke about a bit, you'll find several places where your expertise will be welcomed, including "Narrow Gauge" and "overseas" in the special Interest section. Perhaps a few postings of "Brooklyn - 3 AM" might interest our members interested in foreign layouts

Best

Jack

[Suzie]

It is all about looking at what you are trying to do, then finding the best solution. Unfortunately compromises have to be made in the smaller scales, but you usually have a choice of which compromise you make whether it be dead frog, wrong polarity wing rail or tricky rail cut. The great thing about a forum is that all the options become available.

 

I think that if I was using the point in a station area with bidirectional running on the straight bit, I would probably think about motorising it! Much less work and the electronics is pretty straightforward to automate it.

 

 

 

 

 

[Kenton]

Prof K - welcome to RMWeb and thanks for the additional emails.

 

Though obviously I do not wish to stifle the debate on the use of this point in other layout scenarios - I do not have to worry about bridge rectifiers, changing polarity or even detection systems in this layout application.

The issue of which way a loco runs is small beer as I can put it on the layout and if it runs the wrong way use the 'hand-of-god' to lift it and turn it. In the vast majority of cases the loco can run facing or backwards without too many objectors. The loco+ length breaks in the simple loop are sufficient to enable the current running train to stall within the loop (both parts of which are intended to be hidden from view so the sudden stop is of no concern) As the intention is completely operator free (well ignoring derailment resets :D ) the only big hurdle with the electrics is the timer-relay circuit.

 

Suzie, you are correct there is a wire on the OO9 points at the frog, but it is not too obvious just what effect cutting it would have. Simply gapping the switch rails in the usual way - I use a piercing ("jewelery" as in video above) saw, standard issue equipment for a kit builder is probably workable even if I have to super glue the rails into the sleepers.

[Huw Griffiths]

Tis true that reversing the motor polarity in a loco will achieve "2 locos run in opposite directions".

 

However, given that running at an exhibition can be stressful enough as it is, having to ensure that one puts the right loco on the right train first-time, everytime, for operation/reliabilities sake, is IMHO "making an (unnecessary) rod for one's own back"...

 

Furthur, lets say your loco fleet "dies" mid show, (most exhibitors have at least one "horror story" that goes along these lines... ) if you _must_ have a loco "wired reverse" for the layout to work "as intended", then you've just made it impossible to solve the "all locos dead" scenario by simply heading to the nearest trader, buying 2 "off the shelf locos" of vaguely suitable type, and putting them both straight to work... (giving you a chance to calmly sit down backstage, and work out what went wrong with the "stars of the show"...)

 

As the great Live Event Producer said "the show must go on..."

 

I personally avoid detection systems such as magnets/hall effect sensors for the same reason... (Not that I often do, but I'd like the comfort of knowing I can drop any "foreign" equipment onto any of my layouts, and they won't immediately tech-fail due to not having magnets or other "mandatory mods" the layout requires...)

 

Just my $0.05c

Hello - and welcome to RMweb.

 

I've enjoyed reading your contributions on a number of other websites - so I'm also looking forward to seeing what you come out with here.

 

I agree with your emphasis on keeping everything on layouts as simple and electrically / mechanically standard as possible. As an electrical engineer (and one time university "lab rat"), I know good practice when I see it. I also know why it's important - much of this stuff is ultimately based on experience.

 

With your references to "backstage" and live events, would I be right in assuming that you have a background in sound engineering or stage lighting? Certainly, this is a very tough environment for any equipment - every bit as demanding as the strain gauges and other civils instrumentation I dealt with at the university. I'd imagine there might have been loads of occasions when you've needed to make instant substitutions of equipment.

 

 

I agree with your comments about locos wired "wrong way round" - sooner or later, dodges like this are bound to catch people out - this is why it surprises me that some people continue to resort to them. Although I've never exhibited layouts or models (I haven't built enough - and what I have built isn't of a standard I'd want to show off), I've known loads of people who have. Despite loads of work building and testing things, quite a few of them have had things break down during shows - when there isn't time to do proper repairs.

 

 

 

I think it's fair to say that, when repairs need to be done mid-show, they often look terrible and run even worse - it doesn't matter if it's locos, wagons, track or wiring.

 

I also think that the only really effective mid-event repair is likely to be a drop-in substitution of a complete working unit.

 

If repairs are needed, they're probably best done back at base (when you've got all the equipment and time you need - and you haven't got people breathing down your neck, asking "is it working yet?").

 

These observations don't just seem to apply to exhibitions - from a work perspective, I'm also aware of occasions when they've applied to lab tests and demonstrations at college open days. I won't go into any more detail here - this has happened with lots of people (most of whom aren't currently in a position to defend themselves).

 

 

I sometimes wonder if some people have "wrong way running" issues with DCC - and if this leads to "fun" with consisting. Fear of "wrong way running" problems is one of a number of things which scare me off DCC. I'm used to "old school" analogue wiring and I like the certainty of the NMRA S9 convention.

 

Saying that, even though I'm not about to go digital, the On30 "critters" I'm currently building will actually be DCC ready! I can assure you that there's madness in my method (or something along those lines). The colour coded wiring - and plug / socket arrangement - make it very easy to trace all the internal wires and "switch out" different bits when fault finding.

 

 

Returning to this thread, I actually bookmarked your points article when I saw it last year - so I'm glad it's resurfaced (this time, I've printed it out). I think I'll also recheck Lynn Zelmer's website, to see if there's anything else that interests me.

 

 

Anyway, that's more than enough waffle from me.

 

 

All the best,

 

Huw.

[Prof Klyzlr]

Dear RMWeb Team,

 

Apologies for the aspect ratio issue, but just uploaded a YT video which shows Dynamite Canyon Tramway in action...

 

 

Happy Modelling,

Aim to Improve,

Prof Klyzlr

[Huw Griffiths]

Many thanks for the video - very interesting, informative and watchable - in short, excellent.

 

Regards,

 

Huw.

[Prof Klyzlr]

Dear Huw,

 

Glad to hear the YT helped explain the concepts some

 

I just managed to sort out a correct 16:9 version, so if I have to report, I'll adjust the link in the thread...

 

Yes, I've been involved in Live, Studio, Festival, Nightclub, Pub, and Sound Design/production sound for quite a while, and currently work for FairlightAU http://www.fairlightau.com

 

Dealing with total amp failure at 3AM teaches you things,

(Inc keep a Maglite and roll of Nashua 357 gaffa tape on you at all times... ),

 

but I'm forever reminded that the lessons from the rough-and-tumble world of transporting/setting-up/operating/tearing-down/getting-home-again sound, lighting, and stage gear are imminently/inherrently applicable to all facets of building/running an exhibition layout...

 

Again, glad to hear the YT helped,

Love to hear if it helps get something happening on the workbench...

 

Happy Modelling,

Aim to Improve,

Prof Klyzlr

[Huw Griffiths]

Dealing with total amp failure at 3AM teaches you things, (Inc keep a Maglite and roll of Nashua 357 gaffa tape on you at all times... ), but I'm forever reminded that the lessons from the rough-and-tumble world of transporting/setting-up/operating/tearing-down/getting-home-again sound, lighting, and stage gear are imminently/inherently applicable to all facets of building/running an exhibition layout...

This sounds about right.

 

The comment about the torch and gaffa tape reminds me of my time at the university. I often needed to tape wires to the floor, or inside (sometimes over) test rigs, prior to using hydraulic jacks to smash specimens. Other times, I needed to trace wires - often in dark, inaccessible places - or find ball bearings that had fallen through a grating into a pit. Just a shame I don't particularly like heights or confined spaces.

 

 

Saying that, most of what I've been able to transfer into the real world model making actually related to my strain gauging work - working from dimensioned diagrams - accurate placement - neat wiring and soldering (often in places where there isn't really space for wires) - firmly anchoring wires, while allowing for movement.

 

I also learnt a lot more than I wanted to about glues - how (not) to use them - and surface preparation (very tedious, because strain gauging places all sorts of demands on glues). This also explains why I go out of my way to avoid using glues - I don't trust them as much as some manufacturers would like me to.

 

A number of years back, I read somewhere that some of the early "superglues" had actually been designed to stick skin together (patching up some types of battlefield wounds, for just long enough to get someone to a field hospital). This mirrors my experience - some of them don't seem exactly "super" for anything else!

 

 

Thanks for the prompt about the workbench. Unfortunately, family commitments mean that I'm not likely to get chance to make much for a while.

 

When I do get chance, I'll probably also want to do some more experimental "part-models" before I actually build something recognisable (if I can ever say that about freelance stuff).

 

 

Anyway, I think I'd better stop waffling.

 

 

Regards,

 

Huw.

[Kenton]

So what was the outcome?

 

Well I'm now the less than proud owner of one "random route selection point"

 

Something obviously amiss. Although not yet tested connected to any serious lengths of lead-in and follow-through track and only by "hand-of-god" pushing of an OO9 truck, the results are:

 

The truck passed through the point from the curved direction toe to heel without any problem

The truck passes through the point from the straight direction toe to heel without any problem

However from the heel of the point it appears to be random which direction it travels - possibly favouring the curve

I'll test with a loco tomorrow ... but

I fear something is wrong

[Prof Klyzlr]

Dear Kenton,

 

For reference, the following text assumes the TOE is the end of the point/turnout with the throwbar,

a la

 

 

The "Heel" would be furthur to the RIGHT of the "frog/Vee" as shown...

 

 

 

 

I'm not sure, but I think the 2nd pic tells the story...

(Out of interest, what did you use to cut the flangeways?

remember, "the faster the tool works, the more critical it is that you are _sure_ you're cutting off/into the piece you actually want to!" )

 

To wit:

 

In the 2nd pic:

 

The flangeway for the _straight_ route (bottom cut flangeway)

 

does not appear to perfectly follow the inside railhead edge of the straight stockrail

 

 

The Purple line shows were I _think_ I see the flangeway has been cut.

(above the purple line is a definite black shadow/line, which I believe gives away the upper edge of the flangeway "trench" as carved)

 

Rather, it appears to angle back towards the stockrail, once it's already started diverging...

(IE when travelling heel <--- toe as per the above pic orientation,

the wheels are being pushed slightly sideways (UP as seen in the pic) by the remaining toe-end portion of switchrail --->,

instead of being let travel "straight ahead" as their natural geometry wants them to).

 

In contrast, check this pic of the DCT turnout

(Rotated to match the turnout above)

 

 

Notice how:

 

- The flangeway is dead straight and true, following the inside railhead edge of the straight stockrail

 

- There is _no_ portion of switchrail which intrudes on the natural "straight" flangeway

(no possibility of the wheels being "nudged off course" as they attempt to travel Heel <--- toe along the straight route, as shown)

 

- It's totally a fluke and was NOT consciously built this way,

but the Straight Route flangeway also appears to start closer to the TOE end ---> of the point.

 

This _may_ work as a self-guarding guardrail.

 

IE by the time the flanges pass over the Curved flangeway at top of the image,

they are already "locked into" into the Straight route flangeway,

and thus simply cannot "drift sideways" far enough to catch or derail on the Curved flangeway/switchrail/etc

 

I think your example turnout _may_ be able to be saved, IF:

 

Step 1 - the "straight route" flangeway is carefully carved back such that it presents a "smooth inside running face" on the straight stockrail,

(and may need building back up with solder on the Heel end so as to form a definite "straight route flangeway/trench")

 

Step 2 - the "Curved route" flangeway may need some strategic building back up with solder,

(regular 60/40 will do),

 

if only so that the flangeway gap the wheels have to jump when running heel <--- toe, or "always straight" mode,

isn't _sooo_ big as to cause a very obvious "wheel drop", and "body rock"...

 

Given the (Bemo?) HOn30 test truck/wheelsets shown, Step 1 may be all that's needed to resolve the immediate mechanical issue,

Step 2 will just refine the assembly, and provide a smoother path thru the point...

 

 

 

BTW, If it's any kind of rule to go by, the procedure that's worked for me on a number of these:

 

- remove the throwbar

 

- solder both switchrails to their stockrails

 

- carve the _Straight_Route_ flangeway such that it's arrow-straight and follows the inner edge of the straight stockrail

(razor saw worked for me, gave a straight cut, which was nice and thin in cut width.

Oh, and it didn't cut thru material faster than I could think about what it is I needed to cut! )

 

- once a truck rolls thru that route Toe ---> heel without any hint of 'wiggle' (horizontal movement),

'bump' (horizontal or vertical), or flange 'hump-over' (vertical, because flangway is not deep enough at a given point),

 

It is perfect, and should NOT be changed or modified unless Absolutely necessary

(see below, better to introduce slop to the "Curved Route" flangeway than to mess with a perfectly-adjusted _Straight_Route_ flangeway!)

 

- _then_ carve the _Curved_Route_ flangeway, testing Curved Route Heel ---> Toe.

 

NB that the Curved Route flangway can and will possibly need to be more "sloppy" than the straight route flangeway.

 

Kenton, looks like you're on the right track, just gotta refine it a bit...

 

Happy Modelling,

Aim to Improve,

Prof Klyzlr

[Kenton]

Prof Klyzir

Thank you for the analysis of the problem - and terminology aside I am in complete agreement with your conclusions.

 

The cutting method was initially a razor saw. However with the Peco 009 point this was proving to be difficult as:

1. I came to realise that the "straight" stock rail is most certainly not straight

2. It was cutting deep in to chairs+sleepers either side of the channel (its blade is longer than the point)

3. The resultant channel was too narrow to take the wheel flanges of just about all the stock - especially the Bemo HOe trucks.

 

I then resorted to the Dremel cutting disc. Not for speed but for channel width.

I was happy with the channel on the curved side but rather disappointed with the straight.

 

With your additional explanation - and hindsight - I now realise that it is this straight channel that is the most important to get right.

 

I think that this is going to be a bit of a problem on the Peco OO9 point - as it starts off with known bad geometry. The definite curve in the straight stock rail as well as the other known problem of the check rails at the V (simply fixed)

 

I shall have a go at filing back the remaining part of the switch blade soldered to the straight stock, I cannot see that filling with solder and re-cutting is going to be easy, but may attempt that as a last resort. In this case I think filling even more of the area with solder will be a good idea. The point is well made that the channel on the straight needs to go beyond the left of the red circle so that there is a checking of the wheel set on the straight route beyond the gap in the curved rail.

 

The following images just show the faulty geometry of the Peco O9 point. Although these were taken of the butchered point, it is just as pronounced on a new one.

 

Back to the bench... its going to get messy.

[Prof Klyzlr]

Dear Kenton,

 

I must admit I didn't realise that the PECO points were so malformed,

Is this a "known issue"? I've never seen a PECO point soo far out-of-whack "out of the box"...

 

 

Maybe the straight rail can be "more/re-straightened" before the cutting occurs,

(after the throwbar is removed)??

 

with the straight stockrail tradjectory as-is, the wheels are being naturally 'led' towards the diverging route from the get-go,

which obviously is not what we're aiming for...

 

 

As a minor tip, I find that when the curved switchrail is soldered to the straight stockrail,

and it's time to cut the 'straight route' flangeway,

the natural 'V' shape thus formed works as a great guide-in to start the cut,

dragging the razor saw _towards_ the toe of the point,

(the tip of the razor saw will naturally 'centre"/nestle up against the inner face of the stockrail as it drags up and over the soldered joint...)

 

As you know, the DCT points started life as PECO N-gauge "setrack" points,

so I have confidence that you can make the OO9/HOn30 versions work...

 

Happy Modelling,

Aim to Improve,

Prof Klyzlr

[buffalo]

Dear Kenton,

 

I must admit I didn't realise that the PECO points were so malformed,

Is this a "known issue"? I've never seen a PECO point soo far out-of-whack "out of the box"...

 

 

Maybe the straight rail can be "more/re-straightened" before the cutting occurs,

(after the throwbar is removed)??...

 

That, I think, is the problem as Kenton has already removed some of the timbering and tiebar. The 'straight' rail has started to curve because the curved rail is trying to straighten itself out. Out of the box, the straight rail is quite straight but as soon as you start to modify them in any way the straight rail starts to bend. To an extent, the problem can be overcome when the turnout is stuck down on the board, but it would be good to be able to get it right before that is done.

 

Nick

[Kenton]

I must admit I didn't realise that the PECO points were so malformed,

Is this a "known issue"? I've never seen a PECO point soo far out-of-whack "out of the box"...

I think I first saw it mentioned on RMWeb - but cannot find that reference to it ATM

However, a search just brought up this (see last para) on another forum although the topic is not directly relevant.

[shortliner]

Not OO9 but HO, this was made years ago from a tatty old Joueff turnout from the scrapbox at an exhibition, when the Prof first described it to me. I have to admit I was dubious, but the sense of wonder when it did exactly as he described, through a turnout that I knew didn't move because I'd soldered it up solid, was profound. I just dug it out of my spare track box, and tried it out - it is like the Ronseal advert "Does exactly what it says on the tin!" No hesitation, no derailing - it just works. BTW. Kenton - your link doesn't work unless you are a member

[Kenton]

BTW. Kenton - your link doesn't work unless you are a member

Sorry folk, didn't realise the following is a direct copy of the relevant paragraph of the above link to "Narrow Gauge Railway Modelling Online"

I remember the introduction of the 12†Peco 009 points in the 1980’s and recall my disappointment at their poor running characteristics at the time. I still have my notes on the subject from back then and in many ways the old 9†points were better! As is now commonly known there is a basic geometry error in the curved rail of all three 009 points that Peco produce, basically the point of the V in the frog is narrow to gauge and this results in the flange of the leading wheel hiting it. The check rails also have to abrupt a lead in and as produced the straight rail is actually curved so the left and right hand points are actually slightly Y points. The rail head throughout the point also varies in height. As John has said this can all be fettled to stop this for most wheel profiles and once you know what to do it does not take long to do each point before it is laid, and a large file across the top after it has been laid sorts out the railhead problem.

 

Brian Meldon

[buffalo]

Interesting. Do you know when that was written? As a newcomer to HOe I have some new, as yet untouched, examples of the SL-E491 and 492 turnouts that I have just been measuring. They are far from perfect, but don't exhibit most of these flaws. The gauge varies a bit between 8.94mm (near the blade hinge on the curved route) to 9.24mm at the exit of the curved route, but the straight stock rail is straight and a small sample of Bemo and Liliput stock (untested b2b) has no problems with the check rails or frog. However, as I said above, as soon as you start to hack them about they do start curving.

 

Nick

 

edit: having said that, the couple that I have stuck down after modifying (electrofrog style with insulated blade hinges) are not exactly the smoothest running turnouts I've ever seen, especially with bogie stock.

[Kenton]

Interesting. Do you know when that was written?

Reply #4 on: November 19, 2009, 07:12:52 PM

 

So not exactly historical

 

As a newcomer to HOe I have some new, as yet untouched, examples of the SL-E491 and 492 turnouts that I have just been measuring. They are far from perfect, but don't exhibit most of these flaws. The gauge varies a bit between 8.94mm (near the blade hinge on the curved route) to 9.24mm at the exit of the curved route, but the straight stock rail is straight

Me too, and they do seem to vary quite a bit.

 

I took your note of the tendency of the curved stock to force the straight into a slight curve but do note the first photo is when I first spotted it and at that time no heat had been applied and only the tie-bar (loose anyway) and the switch rails (and we all know how they are hinged ) had been removed. The sleepers were still in place.

 

I compared the modified point by laying over and under a new point (bought at the same time) fresh out of the box. The rail ends all line up by eye though that is not a perfect test. Doing some measurements with the vernier and you get quite a wide variation within the point and between points as you have seen. I guess some of the gauge widening around the V and switch rails is deliberate to ease flow through the point or simply an accident of the moving parts.

 

In all 3 examples I tested the curved rail gauge at the V was smaller than the straight V - but only within ± 0.1mm

 

Prof K has already highlighted the lump on the inside of the straight stock - the gauge at this point is 8.86mm significantly narrower than at (what I call the 'heel') 9.03mm and in the middle of the somewhat flexible switch blade at 9.1~mm

 

a small sample of Bemo and Liliput stock (untested b2b) has no problems with the check rails or frog. However, as I said above, as soon as you start to hack them about they do start curving.

edit: having said that, the couple that I have stuck down after modifying (electrofrog style with insulated blade hinges) are not exactly the smoothest running turnouts I've ever seen, especially with bogie stock.

 

I know that the Bemo truck is possibly not the finest of wheel-sets but their election was deliberate. I have had experience of them jumping on the check rail/V on even unmodified points. A problem that no end of tweaking of the B2B seems to resolve. Though to be fair they are a bit lumpy and unpredictable. It was what was at hand and what will be selected to run in both trains and I thought that if the point can cope with the wide/deep flanges on them then the finer Parkside sets should prove to be no problem.

[buffalo]

...

I took your note of the tendency of the curved stock to force the straight into a slight curve but do note the first photo is when I first spotted it and at that time no heat had been applied and only the tie-bar (loose anyway) and the switch rails (and we all know how they are hinged ) had been removed. The sleepers were still in place.

When I modified mine, the bending took place as soon as I severed the link between the two large timbers either side of and above the tiebar. This link is all that maintains the integrity of the plastic base between the main part and the three timbers at the end. I found that I had to remove this link in order to make sufficient room to detach the blades from their 'hinges' so that I could insert insulating hinges. It appears that the rail is not curved before assembly, as one would do if handbuilding, but just sprung into the moulding jig so it has enough spring in it to start distorting things as soon as the rigidity of the plastic base is compromised. Presumably this is a manufacturing short-cut that would normally have no effect on the shape of the turnout.

 

I compared the modified point by laying over and under a new point (bought at the same time) fresh out of the box. The rail ends all line up by eye though that is not a perfect test. Doing some measurements with the vernier and you get quite a wide variation within the point and between points as you have seen. I guess some of the gauge widening around the V and switch rails is deliberate to ease flow through the point or simply an accident of the moving parts.

 

In all 3 examples I tested the curved rail gauge at the V was smaller than the straight V - but only within ± 0.1mm

Yes, in the ones I measured, it is the curved route that has the narrower gauge. If anything, I would have expected this to be wider (if not the same as) the straight route.

[Kenton]

When I modified mine, the bending took place as soon as I severed the link between the two large timbers either side of and above the tiebar. This link is all that maintains the integrity of the plastic base between the main part and the three timbers at the end. I found that I had to remove this link in order to make sufficient room to detach the blades from their 'hinges' so that I could insert insulating hinges.

OK I'm going to have another go at a new one this evening (I need 2 for the layout anyway)

I think I can remove the tiebar/switch assembly without cutting the webbing.

However when it comes to soldering they are going to be right at the solder hot spot - so the end result may just be the same.

 

I guess another option would be to do some selective sleeper replacement with pcb sleepers to lock up the geometry ... but the I guess I might as well be building hand-laid points and that defeats the simplicity and goes against my grain.

[shortliner]

The other alternative might be to hack about an N gauge one.

[Kenton]

The other alternative might be to hack about an N gauge one.

True, but I do not happen to have any setrack N points and the long points I do have certainly do not look anything like appropriate to narrow gauge (not that they would fit the layout either).

But I guess if nothing works with the 3 OO9 points I have/had before the weekend and there are some on show at Aborfield or ALSRM (very unlikely) then then that may be a viable option.

[Adrian Wintle]

Kato Unitrack #4 switches will do this out of the box - the frog can be set dead and the switch rails are sprung lightly enough that even Microtrains low-profile delrin wheels have no problem pushing them aside on exit. I've used them to make hands-off reversing loops in N-scale (DCC, so the reversing loop module takes care of the polarity and the train never stops)

 

Adrian