Some beginner help

Welcome to the Forum, Highlander! I think you've got enough of everyone else's two penn'orths without me chipping in; I run live steam and battery power. Have a look through the threads that interest you on here, pinch the ideas you like and decide what you don't want to do, too. Everyone has a slightly different approach.
 
This is what I have done About the Railway, but I am a bit of an outlier being in the land down under.
I am dabbling in the dark arts of raspberry Pi, Picaxe microcontrollers and Arduino with the aid of the dreaded and feared AI.
Automatic feed back of a signal setting to a transmitter would be theorectically possible but how to do it I don't know.

Enter the dark Lord; I put this into Chat GPT, "I have a garden railway that uses 2.4GHz RC system to control loco speed and direction. I would like to be able to change a signal remotely and when it changes the loco speed corresponds accordingly ie at stop it slows to a stop and at clear it accelerates away. Is this possible?"

The all knowing one returned this (bear in mind it is talking about what I use for my electronics)
Yes — it is absolutely possible, and it fits very well with the sort of systems you’ve already been working with
The key idea is to let a signal command influence the throttle input rather than directly cutting power.
Below is a clear, practical way to think about it, from simplest to more advanced.

The basic principle (important)​

Your 2.4 GHz RC system already controls:
  • Speed → throttle channel
  • Direction → direction channel
What you want is:
Signal aspect → modifies throttle behaviour automatically

So instead of: throttle → ESC → motor

You insert a logic layer: Handset throttle → controller → ESC → motor
↑
Signal state

Option 1 – Most practical & flexible (recommended)​

RC receiver + microcontroller in the locomotive

How it works
  • RC receiver goes into a small microcontroller (Arduino Nano / ESP32 / PICAXE).
  • The controller:
    • Reads the throttle channel
    • Knows the current signal state (Stop / Caution / Clear)
    • Outputs a modified throttle signal to the ESC
Behaviour example
  • Signal = STOP
    • Ignore further throttle increases
    • Gradually ramp speed down to zero
  • Signal = CLEAR
    • Allow throttle again
    • Ramp speed up smoothly

✔ Smooth prototypical behaviour
✔ No power interruption
✔ Works with any ESC
✔ Very realistic braking and acceleration

How does the signal talk to the loco?​

You have several choices:

A) RC channel for the signal
  • Use a spare channel on the same transmitter
  • One switch = signal state
  • Receiver already in the loco
Simplest if you have spare channels

B) Separate 2.4 GHz link (ESP-NOW / NRF24 / RC relay)​

  • Signal mast has a small transmitter
  • Loco listens for signal commands
  • Works even with multiple loco
Good if you want trackside signals controlling trains automatically

C) Trackside sensor + wireless command​

  • Signal changes when a train occupies a block
  • That signal broadcasts STOP to approaching trains
This is how full block signalling is done.

Option 2 – No microcontroller (limited but possible)​


RC relay or electronic switch​

You could:
  • Use a relay to:
    • Reduce throttle
    • Switch to a preset low-speed signal
  • Or cut motor power

⚠ Downsides:
  • Abrupt stops
  • No smooth deceleration
  • Feels unrealistic

This is usually not recommended for garden railways.

What most garden railway modellers do​


The modern approach is:
  • Battery + ESC + RC receiver in loco
  • Microcontroller between receiver and ESC
  • Signals send commands wirelessly
  • Throttle is interpreted, not passed straight through
This gives:
  • Smooth braking into a stop signal
  • Automatic restart on clear
  • Speed limits for caution aspects
  • Future expansion (block control, station stops, etc.)

Is this worth doing?​

✔ Yes — and it’s very achievable
✔ It does not require replacing your RC system
✔ It scales nicely from:
  • One signal → one loco
    to
  • Multiple signals → multiple trains

Hardware you would typically need​

In the loco:
  • Existing RC receiver
  • ESC
  • Arduino Nano / ESP32 / PICAXE
  • Optional: voltage regulator
At the signal:
  • Servo or LEDs for aspect
  • Small controller or RC switch
  • Wireless sender (or RC channel)
This should give your grey matter to chew over
Oh and welcome to the forum where everyday is an adventure

Now I am interested in this so maybe another thing to add to the list
 
GAP said:
This is what I have done About the Railway, but I am a bit of an outlier being in the land down under.
I am dabbling in the dark arts of raspberry Pi, Picaxe microcontrollers and Arduino with the aid of the dreaded and feared AI.
Automatic feed back of a signal setting to a transmitter would be theorectically possible but how to do it I don't know.

Enter the dark Lord; I put this into Chat GPT, "I have a garden railway that uses 2.4GHz RC system to control loco speed and direction. I would like to be able to change a signal remotely and when it changes the loco speed corresponds accordingly ie at stop it slows to a stop and at clear it accelerates away. Is this possible?"

The all knowing one returned this (bear in mind it is talking about what I use for my electronics)
Yes — it is absolutely possible, and it fits very well with the sort of systems you’ve already been working with
The key idea is to let a signal command influence the throttle input rather than directly cutting power.
Below is a clear, practical way to think about it, from simplest to more advanced.

The basic principle (important)​

Your 2.4 GHz RC system already controls:
  • Speed → throttle channel
  • Direction → direction channel
What you want is:
Signal aspect → modifies throttle behaviour automatically

So instead of: throttle → ESC → motor

You insert a logic layer: Handset throttle → controller → ESC → motor
↑
Signal state

Option 1 – Most practical & flexible (recommended)​

RC receiver + microcontroller in the locomotive

How it works
  • RC receiver goes into a small microcontroller (Arduino Nano / ESP32 / PICAXE).
  • The controller:
    • Reads the throttle channel
    • Knows the current signal state (Stop / Caution / Clear)
    • Outputs a modified throttle signal to the ESC
Behaviour example
  • Signal = STOP
    • Ignore further throttle increases
    • Gradually ramp speed down to zero
  • Signal = CLEAR
    • Allow throttle again
    • Ramp speed up smoothly

✔ Smooth prototypical behaviour
✔ No power interruption
✔ Works with any ESC
✔ Very realistic braking and acceleration

How does the signal talk to the loco?​

You have several choices:

A) RC channel for the signal
  • Use a spare channel on the same transmitter
  • One switch = signal state
  • Receiver already in the loco
Simplest if you have spare channels

B) Separate 2.4 GHz link (ESP-NOW / NRF24 / RC relay)​

  • Signal mast has a small transmitter
  • Loco listens for signal commands
  • Works even with multiple loco
Good if you want trackside signals controlling trains automatically

C) Trackside sensor + wireless command​

  • Signal changes when a train occupies a block
  • That signal broadcasts STOP to approaching trains
This is how full block signalling is done.

Option 2 – No microcontroller (limited but possible)​


RC relay or electronic switch​

You could:
  • Use a relay to:
    • Reduce throttle
    • Switch to a preset low-speed signal
  • Or cut motor power

⚠ Downsides:
  • Abrupt stops
  • No smooth deceleration
  • Feels unrealistic

This is usually not recommended for garden railways.

What most garden railway modellers do​


The modern approach is:
  • Battery + ESC + RC receiver in loco
  • Microcontroller between receiver and ESC
  • Signals send commands wirelessly
  • Throttle is interpreted, not passed straight through
This gives:
  • Smooth braking into a stop signal
  • Automatic restart on clear
  • Speed limits for caution aspects
  • Future expansion (block control, station stops, etc.)

Is this worth doing?​

✔ Yes — and it’s very achievable
✔ It does not require replacing your RC system
✔ It scales nicely from:
  • One signal → one loco
    to
  • Multiple signals → multiple trains

Hardware you would typically need​

In the loco:
  • Existing RC receiver
  • ESC
  • Arduino Nano / ESP32 / PICAXE
  • Optional: voltage regulator
At the signal:
  • Servo or LEDs for aspect
  • Small controller or RC switch
  • Wireless sender (or RC channel)
This should give your grey matter to chew over
Oh and welcome to the forum where everyday is an adventure

Now I am interested in this so maybe another thing to add to the list
Click to expand...
Have to say I am somewhat sceptical about this but I would follow advice from outpr resident expert on all things RC PhilP PhilP . Plus there would be the issue of fitting in an Arduino (other similar devices available) plus circuitry into whatever part of the system decided upon. Oh and if permanently outside weatherproofing that device.
 
I am somewhat distracted at the moment, with keep going over to the horse-spital..
However, the AI has got this spookily close to things some of us have been looking at and talking about.

This may be a lot easier in the climate of parts of Australia, where it is drier (more arid?) than the green and pleasant (read damp) lands of the UK.

IR detection is not the best, outside.. Dappled shade is a problem, though using an alternating current to drive the emitter, means you can look for a narrow band of signal. - Less noise.

Everything will need a source of power. - We have looked at single Lithium cells for signals, to make them autonomous units.
These could be brought in between running sessions, and there are multi-output charges that can be used to charge a number at once.
But, how many would want the regime of charging a dozen or more signals, after every session?

I have got as far as drawing up a housing that would look a little like an equipment cabinet. This would house a small microcontroller / radio board, battery and psu, servo. These could be made to 'clip' between a pair of sleepers for location. Possibly a sub-unit connected for train detection on approach?

It all adds complexity, and how many people would want this?
What 'style' of signal should we produce?
Upper or lower quadrant? - easy to alter.
Which country (or company) signalling rules do we work to?
And the biggy: Could we make any money out of it?

We can give people something very generic (we are playing trains, after all) but building a 'proper' signalling system, would be down to the individual.
If you are prepared to learn how to code, it is not too difficult to get 'something' running, but to create a 'product' robust enough for the 'average' modeller, is a different matter.

PhilP.
 
PhilP said:
I am somewhat distracted at the moment, with keep going over to the horse-spital..
However, the AI has got this spookily close to things some of us have been looking at and talking about.

This may be a lot easier in the climate of parts of Australia, where it is drier (more arid?) than the green and pleasant (read damp) lands of the UK.

IR detection is not the best, outside.. Dappled shade is a problem, though using an alternating current to drive the emitter, means you can look for a narrow band of signal. - Less noise.

Everything will need a source of power. - We have looked at single Lithium cells for signals, to make them autonomous units.
These could be brought in between running sessions, and there are multi-output charges that can be used to charge a number at once.
But, how many would want the regime of charging a dozen or more signals, after every session?

I have got as far as drawing up a housing that would look a little like an equipment cabinet. This would house a small microcontroller / radio board, battery and psu, servo. These could be made to 'clip' between a pair of sleepers for location. Possibly a sub-unit connected for train detection on approach?

It all adds complexity, and how many people would want this?
What 'style' of signal should we produce?
Upper or lower quadrant? - easy to alter.
Which country (or company) signalling rules do we work to?
And the biggy: Could we make any money out of it?

We can give people something very generic (we are playing trains, after all) but building a 'proper' signalling system, would be down to the individual.
If you are prepared to learn how to code, it is not too difficult to get 'something' running, but to create a 'product' robust enough for the 'average' modeller, is a different matter.

PhilP.
Click to expand...
Sorry to put you in the frame Philip, I know you are very busy just now I just could not think of anyone else with sufficient nouse in the forum. Not to decry any in the forum, but RC, Dead Rail and such linking is complex as Philip has said above, at least it is being thought about.
 
No problem...
Needy teenager, but I do know it hurts after back surgery. Been there, got the scars to prove it. :D

PhilP.
 
Good point about detection.
I was not planning on using any detection but rely on the signal transmitter sending state and the loco signal but that would entail setting the stopping distance to match the range.
I am planning an occupancy detection system using RFID so I would use that for loco location.

As for power I could just use a sealed acid battery charged by a solar panel for all my signals, but LI-ion cell holds a lot of charge and the power drain would be low.

I do not think it would be be commercially viable it would be a gimmick that each individual could use depending on electronic expertise.

As far as coding goes I couldn't write a program to save my self that is what I keep AI for. I just ask the questions and evaluate the answers and when it writes my code I run it and if I like the look of it or it does what I want I expand on it.

It would be a lot of mucking around but as with my point setting indicator board would be a fun project.

But I am so far down the rabbit hole now it all seems wonderful in my land.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom