2mm Found in the Attic!! - FS160 Switching Layout.

[JimG]

Well it's been about three weeks since I last wrote anything and that time has been taken up in designing a servo point motor which could work with micro-switches for polarity changing instead of using frog juicers. The MERG servo mounts worked well but the combination of the small throw of FS160 tiebars and the lever lengths in the MERG design meant that there was very little movement to operate the micro-switch which could have made adjustment quite difficult. Also, the pitch of the micro-switch mounting holes on the MERG design didn't match the pitch of the holes in my stock of Cherry switches.

So the 3D CAD was fired up and I started a new design of mount. It was also an opportunity for me to get my head round the developments in Solid Edge. In their 2023 edition, they had completely re-designed their user interface. I had downloaded that version but hadn't really got my head round the different new look of the software and hadn't persisted since I hadn't had a requirement to use it in anger. So I downloaded the current 2024 version and spent about a week making sense of the new look then started on the mount.

And here's the final result :-



There's just over 6mm of movement on the innermost hole on the servo horn and I was looking for about 1mm movement at the tiebar, so the six to one ratio required is provided by two levers, the first horizontal one providing 3:1 and the second vertical one 2:1. There is then about 2mm movement at the lower end of the second lever to drive the micro-switch.





These are the extreme positions for a full 90 degree movement of the servo.



This shows the external bearing arrangements for the two arms - 9mm diameter for the second arm and 6mm diameter for the first arm. The bearing for the second arm was situated as close to the base of the mount as possible to get the 2:1 ratio with the end of the operating wire about 7mm beyond the base.



The connection between the two arms is the short stub on the first arm locating in the slot in the second arm.



The holes in the servo arm are pretty well spot on for tapping 12BA, so the peg on the arm is a 12BA screw with a brass tube collar (1.3mm ID; 1.5mm OD) which allow it to be locked on.



The operating wire is 0.7mm nickel silver and is held in the slot in the second arm. The end of the wire has a very short right angle bend and the wire is fitted by entering it in the slot from the left in this view and pushing it home until the bent portion springs into a hole in the arm, I'm going with the 0.7 NS wire at the moment but I've also got some 0.5mm piano wire and I might try some of that as well. That will mean just a small adjustment in the CAD drawing to narrow the slot.



And here's some of the parts as the design developed. The light grey one in the top left corner was the first try where I was intending putting the micro-switch next to the servo. But I changed my mind and the present design started with the next light grey base, one down on the left. I then changed to another resin which meant a day of testing to get the best settings, and then all the next parts were the darker grey. Most of the adjustments made through all these parts were relatively minor - mainly fine tuning bearing and slot fits to get a good running fit with minimum play. The Solid Edge software allowed these adjustments to be made quite easily, so I tended to keep fiddling until I got things just right.

You can just see evidence of the white powder generated when the resin bearings are operated so I've applied a small amount of Labelle 106 grease to all moving surfaces and that seems to be working well on the working model. Next job now is to fit this prototype and see that it works well on the job. If it does, I'll then have to knock out another twenty. One of the design requirements for this mount was that it could be printed in good sized multiples - e.g. I can print six bases at a time on the rather small plate on my Phrozen Sonic Mini 4K printer. I haven't yet worked out the best way of packing the plate to get the best production output.

Jim.

 

[simond]

Jim,

I did something similar about 18 months back, but I reckon your approach is better than my effort. Nice job!

 

[JimG]

Nearly a month and a half has passed before I can show any progress. The main reason for the holdup was obtaining 8mm long screws to hold the point motors to the MDF baseboard covering. I couldn't find any locally (i.e. the UK) so I had to order from China through Ebay and delivery was over two weeks - and I ordered the wrong diameter on the frst order, so I had to re-order with another two weeks plus delivery time.

So I started fitting the new servo point motors to the end baseboard, which had a requirement for three. Two were fitted with no problems except for having to adjust the throw on the servo arms from the first hole to the second to give me the correct travel at the tiebar. But the third point tiebar was a problem. It was very close to the baseboard edge and the point motor could not be placed.



You can see the slot top right, very close to the end of the board and the tiebar is at an angle which would mean that the motor could not fit.

So I opted to use one of the servo motor setups from the previous layout along with tube in wire to operate the tiebar.



This was the servo end with a 3mm styrene baseplate to clamp the tube and allow the motor to be mounted using its slots to provide adjustment.



The other end of the tube had a form of tiebar to mount on the underside of the baseboard with the bar projecting from the end. This actually took a bit of time to complete since I had to experiment with sizes of the the clamps so that the tube would be clamped firmly but not too much that would grip the wire inside.

This all looked fine, but it didn't work!!! The tube was about eight inches long and formed a 180 degree bend to the nearest point where I could easily fit the motor. There was much too much give in the tube. I built a new support for the motor which would half the length of the tube and only require a slight bend in it, but it was still useless. :-(

So the thinking cap went on and the tube in wire was dispensed with and I rebuilt the sub-baseboard tiebar...



... so that the servo motor could operate it directly.



A short strip wood base was provided adjacent to the tiebar and the servo motor mounted on this to connect with the tiebar. And it still didn't work!!!! - with too much play, even with the pin on the servo horn on the outermost fourth hole. So the problem wasn't all down to the tube in wire. I haven't really dig deep into the reasons but I assume that the combination of two spring arms in the setup, friction on the sub baseboard tiebar and possibly some flexibility in this tiebar contributed to the loss of movement.

But I then had the eureka moment and decided to make a mirror image of my new point motor which would allow the spring arm to engage directly with the turnout's tiebar.



The mirror image motor is on the left, and it only took minutes to design, using the mirror function in Solid Edge.


... and here it is in position at the bottom - it took about fifteen minutes to fit and worked first time. I was fortunate to be able to remove the stripwood base for the previous effort without too much damage to the MDF (it was glued). You can also see the screw holes lower left and upper right for the earlier abortive efforts.

Why didn't I think of making a mirrored version of the motor in the first place you may well ask? I think old age might be affecting my ability to do a bit of lateral thinking. I think I was probably fixated on only having one design of motor which could be swapped in and out without worrying about which motor it should be. But I was willing to use an older design of servo motor ?????

Jim.

 

[JimG]

The ten MERG servo motor mounts have been removed from the central baseboard and I've made up motors to replace them.



After my saga with the 3D printer I only had enough parts for nine of the original design mounts, so another mirrored version (bottom right) was completed to make the ten required.

The only change I have made to the design is replacing the 0.7mm nickel silver wire on the arms with piano wire which measures 0.65mm diameter. I had noticed that one or two of the nickel silver wires had bent on the general rough and tumble on the workbench so I opted for a stiffer component.

So the servo motors should be fitted tomorrow - too much distraction with the Scotland football match on this evening.

Jim.

 

[JimG]

Back after almost a month with not much to show. But I had to break in a new printer or, more accurately, break me into a new make of printer. I had always used Phrozen printers up till now but decided to try another make and went with an Elegoo Mars 4 Ultra. The choice of printer was really decided by the cabinet which had housed the previous Phrozen Sonic printer and the Mars 4 Ultra fitted this quite neatly.

Then followed a day or two getting used to the printer and setting up the best settings for the Siraya Fast Navy Gray resin. I use a test called "Cones of Calibration" and that lets me home in quite accurately on the best exposure.

Then it was back to the servo mounts. There was one close grouping of turnouts in the centre of the main yard board which was giving me problems in fitting three servo mounts adjacent to each other even with the mirroring of the mount as detailed earlier. I had got a further supply of smaller microswitches and I then drew up a more compact version of the servo mount with the microswitch above the arms rather than alongside. But even that wasn't enough and I've now designed yet another version to get round the problem of this particular situation.



Here are the designs so far. The original and its mirrored version on the right fitted with the Cherry microswitches. The more compact version, with the smaller microswitch on top, is top left, and the newest, long version with just one arm is bottom left. The reason for the longer version is to get the bulk of the actual servo body out of the way of fouling the other servo mounts. I had thought of completely re-designing the mount with the servo oriented by 90 degrees to cut down its bulk, but opted to go for a stretched version.

So much for a standardised servo mount, which was my original intent. But I thought it through and convinced myself that the part most likely needing replaced would be the servo, so removing the whole mount, then changing the servo on it, then replacing the mount would not be too much of a sweat rather than keeping complete replacement mounts for all the variants.

The Solid Edge Community Edition software has been excellent for all this messing around in 3D CAD. Basically you just grab bits and pull them around until you get what you want when using it in its synchronous mode.

And I had also missed the facility in the Chitubox slicer to mirror parts with one click, which is even easier than doing it in Solid Edge.

Jim.

 

[JimG]

I've got there at last - all the new turnout servos are fitted on the main, yard baseboard. It seems to have gone on forever, from the original fitting of the MERG servo mounts way back at the end of March, then changing my mind and designing my own servo mounts for 3D printing through three versions with their mirrors, then the 3D printer breaking down to be replaced with a new printer during all the design work.



So here are all the ten mounts fitted - two are hidden by the baseboard framework from this angle...



...and here's the need for the long version to fit around one of the original designs.

Apologies for the rough looking edges to the baseboard framework. The B&Q three ply sheet is actually five ply with a very thin veneer finish on each side which tears at the sight of a jigsaw blade.

You may have noticed another piece of equipment fitted to the baseboard on the first view.



Here's the view from the side which becomes...



...the MERG Ezybus main unit with control panel. You have to have access to this panel to set up servos, so the board is mounted on some short drawer runners which allows it to be pulled out for the setting up but otherwise the board is pushed home for safe storage during running or transportation. I'll neaten it up with a front plate for the drawer with a wee handle.

Next now is the DCC and the Ezybus wiring and I might at last see something moving.

Jim.

 

[JimG]

Things have moved on very slowly over the past month or so - the garden and other things tending to take up a lot of time. But I eventually worked up the interest to start wiring the layout. Wiring has never been my favourite task - more a necessary evil.

So I started on the end board with the seven yard roads. The first thing was to get the droppers from the rails done. I use 1.5mm twin and earth for DCC bus wiring and I use the thinner earth wire for the droppers from the rails.



The line and neutral cables from the twin and earth are used for the bus - I assign the brown wire to rails to the front of the layout and blue for rails to the back. The thinner earth wire is soldered to the appropriate bus wires.



My old Bib wire strippers are excellent for cutting the insulation of the bus wire in place...



...and the short section of insulation is removed with help from my craft knife and tweezers...



...then the dropper is wrapped with a half turn round the bus wire and soldered. I strip insulation from the line and neutral wires to slip onto the bare droppers to insulate them.



Here are all the droppers for the seven roads in the yard attached to both bus wires. The bus wires are held in place with self adhesive bases for tie wraps.



This shows the bus running along the rear of the baseboard and dropping down behind the end board to provide for some more droppers.

Work moved on to the the main baseboard where the droppers and bus wires were fitted.



Next came getting the track power dealt with. On the left is the back of the PCP board for the NCE PowerCab. The output of that feeds a MERG cutout board which then feeds a short tag board connected to a bus feed to the bus wires along the rear of the baseboard. The tag board also feed three flexible cables to feed the other three baseboards. One of these cables is visible cutting across the PCP board and you can just see the other two top of picture with their bared leads being fed through holes in the support to get to the tag board.



Here's the power feed to the end baseboard, clamped to the baseboard side, and with its three pin XLR plug fitted.



And the first loco movement between the main baseboard and the end board. And it does move.

I did find a couple of dry joints, but otherwise everything worked. Maybe not the neatest of wiring compared to some seen on other layouts but I've been fighting to allow access to point motors, etc., and small rather cramped boards don't help. But these small, light boards - 900mm x 450mm - allow me to position them easily on my workbench to get the best angle for whatever job is required and makes life easy. There is no way I could work under a baseboard given the state of my back, hips and legs.

Next job is to wire up the turnout polarity switches and get the Ezybus system working.

Jim.

 

[JimG]

It's been nearly five months since I last added to this thread but another event got a bit in the way. I got severe abdominal pain one day at the end of November. After a discussion with a doctor on 911, I was admitted to hospital same day and a CT scan showed that I had big problems in my colon. It turned out to be a perforated sigmoid cancer - not very funny at 82. So it required two operations, a week in ICU and a week in a general ward before I got home. But it had taken a fair bit out of me and I spent December and January being a bit of a couch potato. But I've started to get back to normal during February and I'm now well used to my wee stoma bag.

So it took a while to get my brain in gear as to where I had got with the wiring. After getting the servos installed I had started wiring up the polarity microswitches and the Ezybus system and had been pretty well designing it as I went along. So after a couple of months away from the job, I had to try and work out what I had been doing, and where I had got to.



This was the wiring for the polarity microswitches. The white twin cable brings the DCC from the bus and is terminated in the small tag block. The three leads from the microswitch on the servo mount go to a three pin connector. The two outer wires in the connector are the DCC feed and the centre wire is the feed to the crossing nose. On the other side of the connector, the two DCC wires got to the tag block and the white central lead breaks out of the black spiral wrapping to go to the crossing nose. This was done so that a servo mount could be removed and replaced easily. I found that trying to solder wires to the micro switches in place under the baseboard could be quite difficult with cramped conditions, so the three-pin plug/sockets were fitted. It also means that if you get the polarity wrong, you just reverse the three-pin connection.



I also set up a bus for all the Ezybus connections. I opted to use 8 pin PCB headers soldered to a strip board. The bus has four ways and they are attached to every second pin. This means that if a socket fitted to the pins is fitted wrong way round, no circuit is made with any of the ways. Black and red are 12V negative and positive respectively. The I2C bus SCL feed is the blue and the SDA feed is green.



The receptacles to fit on the bus pins are made from eight way sockets. And I've just noticed that I've fitted them the wrong way round for the picture. These feeds are from the two external sockets for control panels (closest) and the 12V feed for the bus.

That was the work I had done up to my hospital visit.




...and this is the work I have now done to date.

The two circuit boards upper/centre are Ezybus output boards and four servos are actually plugged into the right hand one. The white cables from the bus board bottom right are connected to the two output boards and also to the Ezybus control panel on its drawer, lower centre.



The bus board has been modified since shown in the picture earlier. I found that some of the 8-pin headers were too close together and making plugging or removing the sockets quite difficult, so they were all reset at a wider pitch.



A closer shot of one of the Ezybus output boards with four servos connected.



The control panel was originally made for the other layout using two Ezybus input boards but I will have to add another input board to deal with this layout. I thought I could manage with the two boards but once I had given a bit of thought to it, three boards are necessary. The Ezybus system is quite basic and routing of inputs to outputs is quite rigid. This layout will require five output boards spread over the four baseboards and every two output boards require to be driven by one input board. The MERG kit for the third board arrived this morning, so I'll get it built tomorrow. I also have to re-make the innards of the control panel to accommodate this board.



And when I powered everthing up, this appeared - most heartening.



...and when I queried the system it showed this - the top line shows that Input boards 1 and 2 are connected. The bottom line shows that two output boards are connected. It should be 2 and 3, but I got the address pins wrong on what should be Board 3.




Here's the shot of the control board showing the system readout. For anyone interested in using Ezybus, the input boards shown previously and the control board above are Davy Dick's original product where he supplied the PCBs and the builder had to source the components. All the modules have now been updated and MERG supply complete kits for them but the latest products still run with the early products.

Jim.

 

[jonte]

Welcome back, Jim, and here’s to improved future health

The wiring and electrics are not only neat but most impressive.

Jon

 

[oldravendale]

So sorry to hear about your health issues, Jim. We hope you'll continue on the path to full recovery.

Brian

 

[JimG]

The wiring and electrics are not only neat but most impressive.

Well, sorta neat.

I'm constrained in making things really neat since I have had to allow for parts to be easily removeable. My servos are bulk bought from Ebay so I have to be prepared for duds, especially in service. I've already got a few rejects, some of which were faulty out of the wrapper.

I use zip ties to hold cables in place and it is just a case of cutting tie(s) to move a cable if required, then use new ties afterwards. I have also made use of spiral wire wrap to keep runs of fine cables under control with the danger of them getting entangled in the adjacent moving parts in the servo mounts. This is really handy since you can easily remove it to get access to wires and re-fit it afterwards.

And it's not only servos. I had to remove both Ezybus output cards to sort out a few dry joints, so I'm now going to make a test lead for the cards to test before they are installed.

Jim.

 

[JimG]

So sorry to hear about your health issues, Jim. We hope you'll continue on the path to full recovery.

Brian

Brian,

Many thanks. I think I am now pretty well back to normal, well as normal as you can be at 83. The continuing post operative support from colorectal and stoma staff in North Bristol is excellent.

My only regret is that my emergency treatment probably used up two elective operation slots and caused more postponements.

Jim.