This software calls the Fastclock API and creates logical pulses to drive analogue clocks.
If your clock is not a pulse-driven 12-hour clock, then my software is not the right solution. The alternative is to call the clock API and getting the time.
The whole point with the software is to eliminate the need for any other logic translating time to pulses in other software. This software also automatically fast-forward the pulsing if the analogue time is not same or one minute after master clock time.
This application must be run on a computer with an I/O device that can be connected to analouge clocks in the hall, for example a RUT-socket.
If the fastclock server is running in the cloud, you need an Internet connection, but if you run the fastclock server locally, you can run both this application and the fastclock server on the same computer, for example on a Raspberry Pi.
The main logic of this application makes frequent calls to the Fasclock API and determines when pulses to analogue clocks should be changed. The logical pulses are:
- Positive voltage, sent when moving analouge clock to even minute.
- Negative voltage, sent when moving analouge clock to odd minute.
- Zero voltage, is sent after the configured pulse length duration.
When the time provided by calling the fastclock API differs more that one minute, the pulses are generated at a faster speed to make analouge clocks to sync faster.
When the fastclock is reset to a new session, it sets the time the session start time. This often differs from what is presented on the analogue clocks, so fast-forwarding starts automatically until the analogue clocks also is at the session start time. Fast-forwarding speed should be determined by the slowest fast-forwarding clock.
It is possible to indicate when the clock is stopped and started.
This is a .NET application that can be compiled to run on most platforms. The software is desiged to make it easy to support different ways to control an analogue clock. You only need to implement a piece of software called a sink (see below).
Much of the behaviour is configurable. See this example.
A sink is a component that translates logical clock signals to some form of I/O. There are some ready-made sinks:
- UdpBroadcastPulseSink broadcasts ASCII bytes for pulses '+', '-' and '_', for start/stop '|' and '-' and session completed 'X'.
- SerialPortPulseSink sets DTR high for positive voltage, and RTS high for negative voltage. However, it can be configured to use DTR only for both positive and negative voltage.
- RpiRelayBoardPulseSink using three relays to control pulses and polarities. The Rpi Relay Board is a relay hat for Raspberry Pi that works with this software when running on Linux.
- There is also a LoggingPulseSink that logs to the console.
To run the app locally, see available deployments. Select a folder and click download. It will be downloaded as a ZIP-file. Unzip and transfer to the computer you will run it on.
During March 2022, I started to make Clock Pulse Service for polling the clock API and generate data about pulses to send. I also bought a RPi Relay Board and mounted it om my old Raspberry Pi 2. Then I wrote an implementation of IPulseSink for that relay board, and have now run it on the Raspberry Pi to test it out, and after some adjustments, it seems to work as expected.
With the software, it is also possible to directly use the I/O-pins on the Raspberry Pi without the relay board. This is fine if another device controlling the clock only needs low current I/O.
In this discussion on GitHub, there are many ideas of how to translate a stream of times to pulses for controlling an pulsedriven analogue clock.