T-Gauge Point Motor Demonstration

In an old post I explained a very simple method to first modify and then motorise a T-gauge point unit. Well, to prove the concept, below is a video demonstration.

In addition to the modified point unit and servo you will also see a PC and Arduino controller. The Arduino is the eventual controller for the layout but a separate servo controller interface will be needed to power more than a couple servos. Using the pulse-wave modulation (PWM) from Arduino outputs for servo and train motor control will be a feature of future posts. 

T-Gauge Point Motors

Pictured below is your typical T-gauge point unit. You will see it has two tiny switch rails, and these are moved by a gentle push on one of the tiny tabs you see on the outside of the point unit. Magnets integral to the point unit grab the closed switch rail keeping it tight to the nearest stock rail.


T-gauge point unit.

The design of the point unit is less than desirable - the switch detail is quite poor and the large insulated section can leave trains stalled - but to scratch-build would be a challenge too far for me (but evidently not others.)

The biggest drawback is the lack of provision to motorise (actuate) the point unit. Motorising the point unit is a must and here's my take on how to do it.

Firstly, if you flip-over the point unit you will see two small screws on the underside. Undo these and you will see that the switch assembly drops out.

Underside of point unit with switches removed.

On closer inspection of the switch assembly (see the next picture below), you will see the switch rails fit snugly into a bracket plate, and the switch rails can be gently pulled free. The switch rails are moved by a drive switch plate that has lugs between which the switch rails sit; when the drive switch plate is moved the lugs move the switch rails. Finally, a magnet held underneath each stock rail by the bracket plate grab one of the switch rails if it is within 1mm.



Switch assembly showing alterations for fitting with a motor.

The picture above also shows the main alteration needed to fit a point motor. I have drilled a hole in the switch drive plate and cut a corresponding slot into the switch bracket.

The final alteration is to file or cut a small notch into the point unit that corresponds with the slot in the switch bracket, as pictured below.

Point unit with slot cut/filed into it.

Then simply re-assemble. Put the switch drive plate onto the switch bracket. Insert the switch rails back in switch bracket, aligning them with the lugs on the switch drive plate. Place the assembly back in the point unit and replace the screws. Finally, push the switch magnets into place from the underside of the point unit.

And for actuation? A small servo (sub 10g type will do) with a steel rod connected to the servo arm and cut to length to thread through the hole you've just cut in your point unit. The servo can be connected directly beneath the point unit (as shown below) or else the connecting rod can have a 90 degree bend to allow the servo to be mounted to the side.

Servo connection to the point unit.

  

All told, it takes about 15 minutes to make the alteration and then you've got yourself a T-gauge point motor actuating a T-gauge point unit!

The observant among you might be wondering how to control the servo. All will be explained in another post.

Controlling Two or More Trains on the Same Line?

The planned layout has long sections allowing multiple trains per line. Which presents an interesting set of problems:

• How to create block control in T-gauge?
• How to individually control each train on the same line?
• How to supply power to each train?
• How to vary the power supply to each train for speed control and reversing?

T-gauge is just too small for DCC to be feasible. So...

A schematic of the proposed solution is pictured below.



Schematic of train control concept.

The concept solution is as follows:

• Divide the line into blocks, with overlaps.
• Electrically separate the rails of adjacent blocks.
• Fit train detection to each block and overlap.
• Use software and PC to monitor occupany of blocks.
• Assign in software a unique ID to each train.
• Use software to 'follow' a train as at occupies and clears each block.
• Provide one speed controller per train.
• Use pulse-width modulation (PWM) with H-bridge to generate a variable voltage in proportion to each speed controller.
• Use software and hardware to switch using relays the PWM power from block to block, tracking the movement of the relevant train.
• Isolate power to a block by switching off the relevant PWM power if the occupying train is detected at the overlap and the next block is occupied by another train.

The schematic above shows only two speed controllers per line, i.e. two independently controlled trains per line, but more can be added. The complexity of doing so involves adding additional relays per block section to switch the additional power supplies connected to each block.

Finally, for the concept to work, train detection must be able to detect all vehicles of a train are clear of a previous overlap. Thus 'presence' detection is insufficient; a continuous train detection system is required. How to achieve this in T-gauge will be the subject of a future post - but it can be done!