From the forum 2017

I haven't been building much of late except track. However I've come up with a very cheap servo powered point motor that I've installed about 10 of so far and I thought that might be interesting. I wanted integrated current switching for the crossing, and I wanted to be able to test and install them without crawling around under boards. I also wanted them to be maintainable to replace parts. They are designed to work with the excellent Megapoints controllers which are probably the most expensive part of the whole system - however the Megapoints products are well produced and worked out components which (in my experience) just work first time if you follow the instructions.

It starts with a bit of old aluminium angle - mine came from B&Q I think roughly 30mmx30mm. I machine or cut a slot where the servo motor will sit, and drill two holes clearance for 8BA bolts to hold the servo in place. In this picture I haven't completely cut out the slot yet. It doesn't need to be pretty: There are also a couple of countersunk holes for attaching some ply to mount the microswitch. …and on the other face (which will be used to attach to the baseboard underside) I mill (or drill/file) a couple of fixing screw slots, and a small 1.0mm hole for the operating wire to pass through: The servo motor is mounted thus, with a couple of bits of ply stuck and screwed on to allow a microswitch to align with the servo operating arm. Easier to do than to describe! The exact location of the various holes and slots is determined by the size and shape of the servos you have. It doesn't matter if the operating wire hole is not precisely in line with the servo spindle centre line, as you can program the servo to suit later on. I've found two small squares of 6mm ply bring the microswitch out to just the right level.

You don't want the servo too far from the baseboard underside as that will limit the throw - my mounting holes are 15mm from the mounting face of the angle piece. You end up chewing lumps off the servo operating arms later on but that doesn't matter either. Installation is a doddle - I'll post a few pics later.

Haines10259 - Dec 5, 2017 at 8:24 PM
The motor is attached to the underside of the baseboard, after attaching an overlength operating wire (spring steel) and extending it up through the hole in the correct tiebar. The servo should be in its 'neutral' position which is normally how it comes, and the blades of the point should be halfway across (wedged with a bit of cardboard). Use the slots to adjust the position of the whole thing - you can see when the operating wire is not under stress and it's in the right place. I use ordinary wood screws just a little oversize for the slots - they cut slightly into the soft aluminium which makes it easy to put it back where it came from.

Don't forget to remove the cardboard wedges from the point blades (!), then use the Megapoints board to program the throw of the servos. They have a dinky little test board that just plugs over the switch inputs, and if you have the baseboard securely mounted on its side then it's easy to see what's happening: The point blades should be lightly sprung against the stock rails at each end of the throw. The Megapoints controller also allows you to select which direction is 'off' and which 'on', so I program 'off' to be the equivalent of the lever in the box being returned to normal (straight road usually), and 'on' being the equivalent of the lever being pulled (turnout road ).

If any part of the servo operating arm comes into contact with the base plate then just snip it off with some cutters. The servo must be free to move throughout its programmed range otherwise it could overheat. Lastly I draw round the base plate with a black marker pen and remove it again to fit and test the microswitch off the layout.

Haines10259 - Dec 5, 2017 at 8:54 PM
You then need to work out which way to wire the microswitch. I use a yellow wire from the crossing opposite the blades, and red and black for the outer and inner rail power. I prefer the roller lever equipped SPDT switches which will generally be rated for 5 or 6 amps which is more than enough. I don't know if all microswitches are the same but I run a quick test just to make sure that I know which contacts are connected when closed and which when open, and I draw a little diagram to make sure I've got the connections right. To keep things simple I generally have the microswitch closed when the point is 'on', as it should spend most of its time 'off' with the switch in its open position. I solder up the wires, double check, and then try mounting it with one screw on the plywood base. You can then apply power and make sure that when the controller switch is on the switch gets closed, and when it's off the switch is open. You can also test the polarity at the crossing before finally installing the motor. Once it's all lined up and working nicely put in a second screw to secure it firmly. Here it is in the 'on' position: and the 'off': If all works properly then push the operating wire back up through the baseboard and tiebar, and screw the whole shebang back onto the baseboard using the marker lines to make sure it's in the same place. The power feed wires then need trimming and attaching to the track feed busbars (yes I am using old mains cable hence red and black). Test once again, figure out at which throw the operating wire extends least above the tiebar then trim it off above the board with a mm or so to spare. It's then very unobtrusive and you won't shred your fingers track cleaning:

This is a double slip with the tiebar operated from one side (to avoid a baseboard cross-brace). Total cost for a point motor is about 50p for materials, £1.73 for the servo and about £1.50 for the microswitch = £3.75 ! Even with £5 on top for the controller board (which will control 12 of them for about £60) it's probably cheaper than wire in tube (at least over any distance) these days. My master plan is to control them all from a mechanically locked lever frame. The controller boards can be networked using just three control wires and two power wires, so the various groups of points and signals can be easily linked up to a central lever frame even though they will be twenty feet or so apart. I just need to get that track finished…..

SimonD - Dec 6, 2017 at 7:25 AM

Dave,
Thanks for posting this, your method of using a “Tortoise-like” spring seems a good way of protecting the servo- I had thought about an eccentric, but it always seems too complicated to make, and thus to detract from the benefits of “cheap” and “simple”.
In place of the Megapoints controller, an Arduino offers some advantages, at the cost of having to program it. It’s very much cheaper, particularly Chinese rip-off copies via eBay. It’s scaleable, in that you can daisy-chain them, you can have interlocking and route setting in software, there are easy ways to connect to DCC, and if you have that hard-to-reach point, you can easily control the frog polarity with a relay.
Thanks again
Simon

Haines10259 - Dec 6, 2017 at 11:04 AM
The spring wire is MegaPoints own suggestion. It works well and it absorbs any twitches from the servos. I actually have an Arduino but I like to keep programming out of my hobby as I earn a living through it - hence mechanical or pre-packaged solutions for me. If I was a mechanical engineer by trade I expect I'd be using computers all over the railway ;). I may use one to control turntables though when I get that far.
Regards,
Dave

Haines10259 - Dec 11, 2017 at 10:22 AM
Postscript - I was finishing testing the board in the photos which has a double slip and five points on it when there was a nasty bang. One of the tiebars had fractured: On investigation it turned out that I'd been a bit eager about trimming back the protruding wire above the tiebar, and because the operating wire is attached close to the servo spindle (following MegaPoints's advice) there is quite a lot of up and down movement in it - it had disengaged completely from the tiebar hole mid-stroke and then pushed up against it snapping it in half. I think with the larger point blade throws in 7mm this up and down motion becomes more significant. I've rejigged all the motors so the operating wire is attached further away from the spindle - this reduces the up and down movement in the wire: I've also reduced the 'spring' with which the blades push against the stock rail to an absolute minimum, as the maximum up and down movement of the wire occurs when there is maximum sideways thrust. You can see the tiebar being deformed by it as the throw begins. A better solution would be to eliminate the up and down movement completely .

One way might be to move the operating wire fulcrum much lower down, and have the operating arm move the wire across using a loop of wire attached to the servo arm, and which can slide up and down the operating wire. Essentially I need to turn rotary motion into as close to linear motion as I can (I imagine that's why Simon was thinking about eccentrics). I also remember the old Protofour turnout operating units which used curtain rail to constrain the movement of the operating wires. So the next batch will be a Mark 2 design. I'll beef up my tiebars too! I used some fifty year old sleeper strip from my 2mm days which may not have been the best choice. I'm wondering if a bit of acrylic would do the job, and where I could get it in strip form. The ideal material would be tough, inflexible, drillable and insulating. Any ideas anyone?

Jim Snowdon - Dec 11, 2017 at 10:37 AM
After going to the lengths of connecting the stretcher bar (tiebars are for holding track to gauge:rolleyes:) to the switch rails through the web and continuing, in prototype fashion, through the stock rail it seems a bit of a backward, not to mention conspicuous, step to then use a bit of sleeper strip as the stretcher bar itself. Granted, the early round bar type stretcher bars are difficult to make for two-rail electrification, given the need to insulate the two ends. The later flat strip stretchers are easier, but if you want something that is inconspicuous, works reliably and is robust, I would suggest a look at Volume 19, Issue 6 of the Gazette -
Jim

Campbell20781 - Dec 11, 2017 at 1:18 PM
Hi Dave
I suggest moving away from the modelling mantra of the hole in the middle of the stretcher bar. Follow the prototype ( what a strange idea! ) and put the drive link at the end of the stretcher bar as shown below. The copperclad stretcher bar has a small brass wire loop soldered at one end beyond the running rail and the drive wire from the point motor rises through that. The stretcher bars shown use Ambis brackets for attachment to the switch rails.

Regards.

Haines10259 - Dec 12, 2017 at 2:24 PM
Food for thought gents. Maybe I can adapt Jim's stretcher bar design to work with my extensions anchored through the stock rail (which in my case perform a real function in restraining the blades from wandering off). It would certainly look better.
I have driven some of my points from the end of the stretcher bar. The Ambis style of stretcher offers better strength vertically which would help me, but really I need round bar stretchers. I need to solve the up/down movement problem in the operating wire as that is the fundamental flaw. Plenty more points to build - I've only just got out of the fiddle yard!

Bob Alderman - Dec 12, 2017 at 4:05 PM
Dave
If you have to drive in the middle add a small wire loop for the drive wire. Like the one on the end of Jim's example. Next time you are at club have a look at the tie bars on Gas Works they are round.
Bob

Jim Snowdon - Dec 12, 2017 at 4:31 PM
And, as Bob pointed out, the loop can be in the middle just as easily as it can be at the end - we have one on our club layout that had to be arranged that way. Wherever the loop is situated it no more than a 360 degree loop in the wire that forms the stretcher bar.
Jim

Paul_l - Dec 16, 2017 at 5:42 PM
Hi Jim
I think Dave means, as the servo passes through the arc, at the mid point the actuator arm is at its highest point, however as it moves to the end of the stroke the actuator bar height lowers below the tie bar - he just cut it too short. I am about to start my first point in O gauge, and was going to have the actuator below the baseboard, with two tubes the correct distance apart, a brass wire, folded back on itself, then a 90 degree bend in one leg soldered to the blade, with the bent wire going under the stock rails, the straight wire inserted into the tube. A piece of plastic rod would then be glued between the actuators to represent the round tie bar.
Paul

Jim Snowdon - Dec 16, 2017 at 7:03 PM
Paul,

I understood that bit, having been there myself. My response was more to do with the consequent breakage of the stretcher bar that Dave was suffering, and a potential alternative design that would both do what was wanted and be more robust.

Getting the mechanical drive from the point machine to the switches has always been something of a challenge and I have probably encountered most of the solutions. The “Tortoise” type method is possibly the simplest in terms of both operation and installability, not least as it can be implemented after the turnout has been installed, thereby avoiding a lot of the hassle of getting the turnout aligned with the hole(s) in the baseboard.
Jim

Haines10259 - Dec 18, 2017 at 9:44 AM
That's an interesting approach Paul - it sounds a bit like the old ProtoFour method which separated the mechanics of driving the blades completely from the cosmetic stretcher bar. I have an article in an old MRC somewhere - it used channel curtain track as a sub-baseboard guide. Good luck with the track building. I'm building a curved crossover at the moment and as happens so often in O gauge I'm taken aback by the sheer size of the thing! You draw it on paper and then when you build it it seems twice the size…

End of Thread