Making the insulating gap

Once in a while a while we all get a light bulb idea. For many years now I've always cut a shallow groove in copper-clad sleepers/ties when building track, turnouts, etc, in order to isolate one rail from the other, placing it as close to one rail as practically possible. And always at 90° because that's how it's done matey (and we are creatures of habit when it comes down to it). We've had a bit of an going discussion on the site about whether a cut with a file introduces more weakness when compared with using a grinding wheel in a Dremel (or similar) to remove the copper layer. Personally I've found that both weaken the copper-clad, so it's one of those "yer pays yer money and makes yer choice" scenarios where everybody is right (or wrong, depending on which side of the fence you are).

Earlier today I was cutting some pieces of styrene strips at 90° and 45° to make some cosmetic sleepers/ties, and was simply scoring the surface with the cutter blade and snapping between my pinkies as I couldn't find the NWSL Chopper. "Old on, sez moi", when said pinkies started to protest, it's a darn sight harder to snap at 45° than at 90° (quite a lot, try it and see).  And even more so at 30° (next to impossible, the styrene just twisted). Which when you think about it makes sense.

Then the bulb went on - why am I not doing this with copper-clad ties. So I did a quick (empirical) test. It's relatively easy to snap copper-clad after a light insulating cut at 90° made with the edge of a triangular file, it was next to impossible with a 45° cut and impossible with a 30° cut. Bingo.  :thumbs :thumbs :thumbs (Hold on, I've only got 2 of these).

Unfortunately it means I'm scrapping the double crossover I'm building which has 90° isolating cuts, and starting with fresh copper-clad with angled cuts. Picture shows what it looks like. Same depth on both, as I built a jig to make sure.



Nigel 
Save

Thinking out of the box, very good.  I suppose it's a bit like why tank sloped armour works so much better than vertical.  Question is, is the strength of the strip with the 90 slot adequate for the job?  Most timbers just sit there and aren't under any significant stress.  You know my thoughts on tiebars.

John

Neat idea but I wouldn't strip the crossover Nigel just to replace them - tiebars yes if they will be stressed out but I never found that happened. More often than not, it was the soldered blade to the tiebar that came adrift so I also solder a right angle piece of thin brass rod to the tiebar & blade web for extra strength.

Hi John,

It's the tie bars I had in mind, most of the sleepers/ties in a turnout get very little stress, and in straight track zilch. The problem with the double crossover is that it's one big set of frogs held together with long bits of copper-clad. No tie bars, but I've a few RH turnouts to build. The stress areas there are the tie bars and the junction twixt the blades and the closure rails. Unless there is an inordinate amount of slop there will be tension at both ends. If the tie bar is tight the closure junction needs to be loose. Vice versa.

Hi Ron,

What side are you soldering the angle to - the inside? As I mentioned, the issue with the tie bar is that it does not move in a straight line because the rails are following arcs - 2 separate and distinct ones, so the rails are alternately being stressed on different sides, and the side where the blades meet the stock rails needs to be tight.. John is using a pivot (the brass pin) through the tie bar to allow the blade to turn independently, which will I think need a bit of slop to accommodate the fact that the ends of the blades are not moving parallel to the tie bar (which will be restrained by the motor wire). I've yet to have a tie bar fail, but I have read that it can happen. Probably because of lack of attention to the geometry rather than intrinsically bad soldering or a weak solder junction. I'm starting over because a 30° cut is a lot less noticeable than a 90° cut. I played around with scale diagrams just to make sure.

Nigel

Yes Nigel, on the inside, you can just make out the  right angle rod out


the blades are not hinged but flex for 4"/100mm without problems. On my 3 ways units, flexing is down to 3"/ 75mm

and yes, I have never filled in the gaps, let paint and ballast dust do it !

Nigel, for a giggle I compared three turnouts to see what the blade throw is.

Peco code 75 small OO turnout:  2.5mm

Peco code 124 O turnout: 3mm

My turnout:  2mm (I used my flangeway gauge) and it remains to be seen whether wheels will have problems.  My limited testing seems to indicate that they're good.

Point is that the stress on the hinge isn't large whatever you do.  However, a solid solder joint between rail and tiebar will see greater stress than a hinged one.

Empiricism rules.

John

Hi John,

Yes, it's a small distance. However, if you want another giggle, solder up the connection between the closure rails and the blades, and solder the points to the tie rod. Then see how much resistance you have when you move the blades across. Especially with code 124 rail. Having that pivot is going to make a world of difference, and I'm definitely going to give it a try with the turnouts.

Hi Ron, Thanks for the confirmation. As I said, rigid at one end means slop at the other.

Nigel 

All of this is interesting for me, guys.

I have 12" long point rails in Code 143 which are 10" from the last attachment; with no hinges.

At rest (disconnected from the motor), they sit midway and deflect 1/16" each way when thrown.

They are soldered to the throw bar, right across underneath the rail foot but now I'm leaving hatches in the infill so I can do the thing with the rod, later on, if required.

So far, so good. 

Max, I was wondering whether to use a loose heel (hinge) on the switch blades or leave the rail in one piece.  For code 75 BH, the Tortoise will move the blades easily if they are one piece.  I wasn't certain about code 124 so I did what Peco did and put in a loose heel.  I suppose its a quasi loose heel because I soldered the code 100 rail joiner on opposite sides to the closure rail and switch rail.  The switch rails move easily so I think I did the right thing.

Nigel, I think that intuitively, the brass pin hinge is better although it takes a bit more work to set up.

John

I'm only using model aero-servos, John.  They can easily hold against the Code 143.  My Tam Valley decoder switches off once the end point is reached, so that part of it is OK.

I'm mainly concerned about the connection to the throw bar, but I can always go back and beef up the joint if necessary.

I'll leave it for now, but keep a close eye on it.   :shock:

I don't have any experience with servos but I'm thinking Tortoise have a lot of torque. 

John

Hi Max,

When I rebuilt the Shinohara turnouts (code 100) I used a semi-floating heel, with the rail joiner soldered to the blades but not the closure rails (as original). Shinohara opened the ends of the joiners up a bit to allow the blades to move laterally. Just soldering a connecting wire from the bottom of the joiner to the closure rail stiffened things up dramatically. I stuck with the blade-stock rail gap throw that Shinohara had used, 2.8mm (around 1/8"). Same distance on my Peco Code 83 turnouts. Which is way too generous for modern stock. NMRA site is down at the moment, so I can't check there. Real life was 5", which in 1:87 is 1.46mm (1/16"). The templates I normally use from Fast Tracks have a throw of 2.25mm and are supposed to be NMRA compliant. Some adjustment is probably called for, the less throw on the point blade the less resistance.

Nigel

RTP turnouts tend to allow WAY too much clearance between stock rail and switch blade.  I was looking at the Peco code 124 again and it really looks silly.  In OO there was never any solid standard for wheels so Peco had to design to the worst case.  That was decades ago of course, but things don't seem to have changed despite improved wheels on modern stock.

Even 2.25mm seems large (given that Peco code 75 is 2.5mm).  Finescale track has 1mm flangeway so the answer probably lies somewhere in between.  If you've controlled your wheels standard you can probably build tighter, but if you have to play with others, you might need to stick to NMRA.

(Aside:  I read that Peco are mulling a new range of track and turnouts with code 75 BH rail.  Track is supposed to be first and, subject to decent sales, Peco will release turnouts.  It's a chicken and egg thing I think because those people who care about such things won't want to mix BH track and FB turnouts.)

There is less excuse for O gauge I think because wheels tend to be more of a constant standard.

John

Hi John and Max

Correction, the 5" I quoted is I believe the NMRA  standard (to allow for over-scale wheel flanges), real life is often nearer  3.5", which comes down to a minimum of 1mm in HO, 1.12mm in OO, and 2mm in UK O  scale. I just checked the flanges on some old HO diesels from the 1970's  and 1980's, maximum I came up with was 0.95mm on an Atlas/Kato RS-1 diesel.  Freight cars and passengers cars were between 0.8mm and 0.9mm (and some of the old ones waiting new wheels had pizza cutters).

The  Shinohara turnouts I used have excessively wide throws to allow for polarity changes (both point blades are connected, power routing approach), something I  should have checked. I suspect this is a legacy with the Peco turnouts. So now I have an excuse to go back and refit with angled isolation cut tie-bars and a throw of 1.46mm as the polarity now  follows that of the stock rails.

It might also be worthwhile replacing the currently separate closure rails and point rails with one piece and getting rid of the rail joiner or using a brass-strip hinge in the web. I'm not sure how a throw of 1.46mm will work with  those Caboose Industries ground throws, which have a throw of 0.165"  (4.2mm). Which begs the question, would they have worked properly with a throw of 2.8mm? I may have to go to manual throws underneath from Fast Tracks  (which I want to try anyway).

If the turnout is properly wired for DCC (and DC) short circuits should not be an issue, which of course they are with the original power routing.

I  can control my wheel standards, this exercise is for the modular club  track, where there are some old clunkers that come out every now and again with their wagon wheel treads and flanges. Mind you, if I stick to NMRA  standards and the club specifications follow NMRA standards…

Nigel

I checked both GOG and S4 Society and can't find a specific dimension for switchblade throw.  If your 3.5" throw holds then that scales to 2mm in 7mm scale so that's comforting.  I may go back and tweak my tiebar to add another 0.5mm of throw.

Of course flanges are probably wider than the real thing.  In OO/4mm flanges are typically up to 0.8mm thick, whereas finescale is 0.5mm, that makes quite a difference in acceptable flangeway.  Tyres on the real thing are about 6", so 2mm in 4mm scale - this agrees with P4 (EM is 2.2mm).

I was thinking of your having to cater to your colleagues with their ancient stock.  Wicked of me to tempt you to deviate from the good book.

Probably you can get away without a loose heel, although I guess it doesn't hurt.  Rail is pretty flexible.

It's all good fun though.

John

All good, John.

I have a further complication - particularly on the turnout nearest the gauntlet.  To negotiate the slipway cut out, the curve facing the heel end of number 1 turnout is a very tight radius.  I haven't measured it, but it's very tight.

The others aren't much better, as I wanted to fit all six on to the centre module.

The wheelbase of my loco is 10 1/2" so I need the full 1/8" clear space on the inside point - otherwise, the loco climbs on to the point rail. 

It's not very pretty, but once the infill goes in, it's not so noticeable.

A bit difficult to visualize Max, maybe a picture?  :It's a no no

John

Heh Heh.  I asked for that.   

Stand by . . .

Here we go . . .



You can see the tight curve leading into the gauntlet on the left.

Allowance for the foot of the far point rail has caused the large gap on the infill.  Once the outer filler pieces are installed, it shouldn't look too bad.

I just re-noticed the crappy soldering.  I must fix that.   :oops:

It doesn't all that tight Max.  Min. rad. is 4'.  If stock negotiates the turnout without derailing, I guess it's OK.  I wouldn't be happy with wheels climbing the rails though.

It's a short transition from the curve to the switch blade, that could be problematic.

Amazing what photos show up.

John

Just a thought, the switch blades look a bit thick, is the gauge OK at that point?  Should be 29mm IIRC for F7.