Problem with geometry in Roco 2.5 mm

[soNak]

It seems as if there is a small bug in the geometry of the Roco 2.5 mm track lib:

If you use the correct parts according to Roco's old geometry plan, the geometry does not work out as expected. It looks like the radius eather for the turnouts (42313/42315) or for the curve (42233) is not correct:


According to the original Roco catalog:


Can anyone confirm this or help me out if I am completly wrong here? 
Thanks!

[Ferdinand Bogman]

Hello soNak,

Today I downloaded AnyRail, started designing a layout using the "HO Roco 2.5mm" library, and ran into the exact same problem! There seems to be no curve piece with which you can make the turnout track curve back to run exactly parallel again to the straight track of the switch!

I thought I might be wrong, but you found proof in the Roco catalog, that such a curve piece exists. It must be a bug then.

Let's tell them!

Kind regards,

Ferdinand.

[David]

The turnout has an angle of 12º50', and the curve has an angle of 12º10'.

In the case you use the curve to extend the diverging track, you'll end up with an angle of 25º.
In the case you use it as a return curve, you'll be left with a slight angle of 40' (0.67º). That's what you see here.

So in theory (and in AnyRail), it doesn't fit perfectly, but in reality it does, because of the play (wiggle room) in the track.

This happens quite a lot when using rigid sectional track.

To solve this, you can create imperfect connections (so at a slight angle) in AnyRail.
How this is done is shown here:
https://youtu.be/20hDeHXfAdo


I've attached a solution for this particular case.

[Pete]

exactly the same situation is occuring in Roco H0e-
the adjustment track <32205> when used with either 32409 or 32411 point, will bring the two tracks back to equi-distant - for proof, look at the track diagram displayed with any Roco H0e digital starter set.
with the anyrail6 H0e libuary, the two tracks show a distinct divergance and will not reconnect if planning a passing loop.

[VictorPM]

The turnout has an angle of 12º50', and the curve has an angle of 12º10'.
In the case you use the curve to extend the diverging track, you'll end up with an angle of 25º.

One could wonder what on earth is the use of a 25° curve when there are no curves to complement to a 45° or 90° curve... so that's a poor excuse. I think the real reason they made this 12°10' curve is so that the total length of the 42204-42233-42204 combo fits in the scheme shown in the catalog. You still need to flex that combo a little, which in reality is no problem but digitally it is ;-)