MECoB stands for Modular Embedded Control Board. It’s an initiative by Electrodesigns.Net to build a small, general purpose, configurable and flexible control board based on this idea. At this stage it is just an idea which may be discussed to get it matured. The idea, in brief, is to have a single carrier board with many IO boards/modules, stacked in a rack model over the carrier board and inter-connect them through a common and very simple interface bus like SPI or I2C. I am currently considering I2C because it is very simple, available in every microcontroller and does not need an additional signal like select pin as in SPI.
The MECoB may contain these modules:
- Processor (PRO-MOD)
- Power Supply (POW-MOD)
- Digital Inputs (DIO-MOD)
- Analog Inputs (AIO-MOD)
- Relays (REL-MOD)
- PWM (PWM-MOD)
- …Some other Board
The following lines may run trough the bus:
- 24VAC/DC: As it is an industrial standard to use 24V as power source.
- +5V: Some ICs or controllers run on +5v
- +3V3: The standard VDD line to power ICs
- SPI lines: DI, DO, CK
- I2C lines: SDA, SCL
- GPIO: Around 4-6 for SPI-CS or general purpose usage.
The MECoB may not look like a traditional PLC, rather simple enough to be adopted by open source community. But is should serve the same flexibility with an open source spirit. It should be easy enough to be handled by a hobbyist at the same time useful enough to be used by professionals.
A possible configuration may be a carrier board with a common bus on it. The bus connects different IDC connectors. While the processor board as well as the IO or peripheral board snap into the IDC headers. The IDC headers on the carrier may be male while all those on IO boards may be female.
Above images show the connectors and possible arrangement of the carrier and IO boards. This is simple scheme yet very flexible to incorporate different combinations and number of boards to be used according to individual requirement. The concept can be started with a simple hardware, later the individual modules can be upgrades with the time. Another advantage is that different developer can work on different modules so the development is easier and fast. The user is free to choose among which modules he chooses to use.
The bottleneck here seems to me the carrier board which I think needs to have some limited number of slots to snap the IO boards upon it. However, if there is a possibility of using a simple scheme to integrate the IO boards without using a carrier, it removes the possible bottleneck. One possibility may be to use din rail receptacles on each board and snapping them on common din rail. Although this is a common practice in the automation field where PLC modules with other equipment are installed in this way, connected together with ribbon wire, but is this workable idea in the embedded world?