Automating the BrickTracks & FX switches

[Ludo]

Hi community,

The BrickTracks R104 switches are already a while available on the market and the FX switches will follow soon.

I ordered some FX switches, (I have only 9V track) and was wondering if someone already automated his switches using any kind of automation, using servos or anything else.

I am completely aware that the driving mechanism of the BrickTracks switches are different than the FX switches (Rotate versus slide motion), so any idea is welcome.

My switches will be placed on MILS modules, so there is some space to use a servo, just not knowing how to do so.

The experience I have with the servo's on my level crossing controlled by an Arduino UNO are a bit strange.

When the controller gets initialized, the servo's are 'flying' full in one direction (using the servo library), and If not proper controlled I guess that the driving mechanism of the switch will be destroyed if this happen.

Knowing this behaviour, I guess also that it is recommended to set the switches in a kind of 'start' position before the servos are initialized, this should be executed before the end of an event.

Anyone out there who have more experience with servo's and their behaviour who are willing to share it here?

The do's and dont's, the pitfalls, ...

Thanks in advance.

Ludo

 

 

 

[Moz]

I much prefer mechanical safety over relying on the software. So design the linkages (etc) to use the full range of servo motion, or at least tolerate it. In the past I've used a really short servo arm to bump the switch from side to side so that the full servo travel wasn't necessary but also didn't matter. That made setting them up easy, as long as the arm didn't hit the switch mechanism when it was going in everything was fine (the arm can't be on the wrong side or in the wrong position). That can be set up to make the official shifting mechanism unnecessary (I was using official Lego switches so I had to remove the manual lever and grind down the friction mechanism).

With BrickTracks you absolutely will break them if you over-rotate, so I would be very tempted to not use their mechanism at all. A quick play with one suggests that putting a post through the hole that the rotary mechanism attaches to should give a servo arm something to bat back and fort as it rotates past. The other option would be gearing the servo down so it only has as much rotation as the BrickTracks mechanism wants. But it would likely be easier to 3D print a variant of their mechanism that has a servo drive hole in the middle and requires/allows as much rotation as the servo provides. Then you just have to get the alignment right when joining the two.

(edit) mechanism sketch  (paint.net sketch of the mechanism)

Edited July 1, 2023 by Moz

[Phoxtane]

The last time I looked into this, I used a hobby servo but set it up such that there was a lot of mechanical slop (see below).

Nowadays I'd look into a belt drive; that way you only need a DC motor as opposed to a servo, letting you use plentiful GPIOs over limited analog pins. (You still need two pins to set the direction and turn the motor on/off with most driver boards, but you can do some work with shift registers etc. to expand what a given microcontroller can do for you.)

 

 

 

[Ludo]

Thanks @Moz& @Phoxtane for the reply's.

Did read the reply's on the provided original article too, interesting.