by Brian Podmore
I bought a set of crossing gates from Crossinggate Models some time ago. They are well made and have particularly good hinge arrangements. I made them up and positioned them insitu on the railway, then nothing for a couple of years. I had every intention of getting them to work but it wasn’t a priority. So last summer, (2014), I revisited the crossing gates with a view to motorising them.
I started by setting up my level crossing gates with a servo driven control system I also bought from Crossinggate Models, but it suffered from severe interference problems, (which they are apparently prone to do and more on that later on in this article), and this resulted in the gates continually juddering. After many fruitless hours asking for advice on various forums and trying all routes to cure this twitching, I decided to take stock of the situation and start again with a different arrangement, this time from GF Controls of Wigan.
Their PCB, (printed circuit board), is quite sophisticated and they do produce a similar model to control signals as well. The use of mini servos and digital rotary encoders to fine tune the position of the servos is a great boon and their kit does not appear to suffer from judder. The ability to set the position of the gates both in the open and closed position is crucial to the correct working of the level crossing, and this kit allows you to do just that by using mode switches on the board. It also has on-board memory so that positions are stored on the board.
The gates I bought from Crossinggate Models as I mentioned were sturdily built and have brass hinges and brackets. To motorise these gates their hinge bar had to be extended to below the baseboard. I used 3mm internal dimension brass tube purchased from Squires to do this. The open end of the tube was threaded to receive a 4BA screw, (I used a tapered tap and tapped down to full depth of thread for about 10 mm. In this way, the screw you use will bite into the half formed thread created below it). The brass tube walls were quite thin so care was needed to accomplish this. The brass tube was then glued to the hinge having been left long enough to come through the baseboard by about 30mm as shown in Fig 1.
The plan was to connect the levers contained in the servo kits directly onto these extensions. In this way I could arrange the servos directly underneath them in what I call 'direct drive' mode instead of building a lever and rodding arrangement as commonly used for model aircraft. A 90 degree rotation was needed and most controls using a combination of levers can only manage 45 degrees. This is more than enough for a signal but not my gates, consequently, I was pleased that the G F Controls kit can actually give 100 degrees of turn. So even if you have offset gates on your crossing, it can accommodate them.
Most servos have small retaining brackets built into each end of their bodies, and the rotating shaft of the servo is often offset to one end of the servo body because of the gearing inside it. With each servo comes a collection of levers, discs, washers and screws that can be put together to ‘turn’ in various situations. Their most popular use is with model aircraft and they pull and push rods connected to these levers to control ailerons and rudders. To set up my intended method of direct drive without the servo levers catching on any surface, I made up special brackets to hold the servos in their correct positions. The top-hat bracket on the left here in Fig 2, holds the end near the shaft, the Z-bracket on the right secures the other end. I made the brackets out of brass strip strengthened with brass fillets. The dimensions I used are shown in Fig 3 below. In fact the two sets had to be different as one extension came out through a bracing spar which meant clearances were different.
In Fig 4, the servos were attached to their brackets and their cables were plugged onto the connections on the PCB. A Hornby on/off switch was connected to test the complete setup before fitting below the baseboard. You can clearly see the adjustment knobs for the digital rotary encoders on the PCB. Unlike traditional potentiometers these will rotate forever in either direction. There are also three slider switches that control programming, bounce and whether you wish to control both gates by one switch. The instructions explain all and my video at: www.youtube.com/ watch?v=8OneinQJjV0 shows this arrangement working.
It was now time to put everything together. So more gymnastics under the baseboard ensued.
There was a small distance between the arm secured to the servo and the upper arm on the gate extension, (marked as B on Fig 5). The pairs of arms are connected together using two thin screws which make up a simple Universal Joint type of arrangement, (anyone who had a Morris 1100 will know what that means!).
I used the levers supplied to attach to the servos and shaft from the crossing gates together, but the ones I used to connect to the tube extensions from the gates did not work as the holes in the levers were almost the same size as the tube. When I screwed them up, it pushed the tubes through the soft polystyrene arms and consequently the levers kept slipping round their tubes. I replaced these with harder levers of a different diameter hole and these gripped the tubes nicely when the screws were tightened up. The lever on the servo was then connected to the lever on the extension tube by two screws as shown in Fig 5. In calculating the size of the bracket these distances need to be measured beforehand. This allows for some adjustment and can raise the gates by a small amount if floor clearance adjustment is necessary.
I decided to keep the control PCB near to the servos. One thing I came to understand with my first failed attempts using servos is that long cables can be a source of interference and servos are prone to being affected by this. It is best to keep cables short and reroute all other cabling nearby that may be a source of electrical interference. If cables have to be long, then they may have to be shielded. I screwed the PCB under the baseboard as shown here in Fig 6. The controlling switch is positioned on its own control box at the side of the railway some way from the crossing as shown in Fig 7 and is a simple Hornby on/off switch.
The instructions from GF Controls explain how to set the gates up and fine tune their position using the controls on the board.
The gates are now fully working. There is little or no interference and the twitching gates are no more. It’s important to make sure that the clearance under the gates is adequate and that the gates don't rub on the floor or catch on brackets as this will interfere with the fine tuning and position of the gates. If this occurs you will have to clear the obstruction, reposition and retune the gates.
This had become an issue with my 'floor' of the level crossing and so I decided to replace the existing epoxy boarding originally supplied with the gate kit with real boarding, (the resin boards had developed a distinct curve and bowed up above the rails in certain places and this was catching the gates moving across the crossing area. I had tried the hot water treatment to straighten them out but they were insistent on reverting to their preferred curvature when cool).
I used strips of lime wood commonly used by model boat builders and although the picture makes it look as if they are single pieces they are indeed made up of strips 7mm wide, (see lead picture).
Finally, I mounted the Hornby switch on a control box. You have the option of using a single switch to control both gates so that one opens first then the second one operates. You can also elect to have the gates bounce when they reach their destination!
The lettering on the control box gives priority to the railway as far as open and closed is concerned, (Fig 7).
A video of the crossing working is available in my channel at: www.youtube.com/watch?v=QX5h04dVmTc
The G F Controls web site is: www.gfcontrols.co.uk
The Crossinggate Models site is: www.crossinggate.co.uk/