Splashboard V 2.0 IoT






Last Saturday’s embedded systems meetup was super awesome. 9 very active participants, 4 hours of great learning and a tangible output.

We even had participants who’d traveled from Dedan Kimathi University of Technology in Nyeri specifically for the meetup!

The team reviewed Splashboard V 1.0, prepared requirements for V 2.0 and drafted specifications to guide online specifications refinement, design and a subsequent build on 8th October, 2016 at Gearbox. It’s going to be a month of intense but fun online working and networking.




IoT – The Internet of Things. As mundane as
it sounds, we are on the cusp of experiencing the next all-encompassing wave in
technology. Simply put, IoT potentially means ‘everything’ could be connected
in one way or another. Applications range from the easy but novel (chilling
your beer to the exact desired temperature and having it brought to your couch
laden self) to the far more complex, and possibly lifesaving e.g. in
engineering and construction fields. Detecting individual stresses in a bridge
pillar long before a crack is seen comes in handy.

Here at Gearbox, we’ve caught the IoT wave.
So in true Gearbox form we went ahead and innovated! Introducing the Splashboard – an IoT device used to keep
track of the quality of reagents used in Splash Center. The Splash
Center is a machine
used in the wet processing of PCBs. Wet processing includes developing,
etching, and tinning. This requires the use of reagents whose pH, conductivity
and colour must be monitored since the chemicals wear off with consistent use
which diminishes their quality. The current Splash Center has no such
functionality, hence the invention of the Splashboard. It collects data for
processing and display which tells the user the current state of the etching,
developing and tinning chemicals. The data collected will enable the user to
ensure that the reagents are always in tiptop condition.

This past weekend we had a lovely meetup
(with tea and cookies to boot!) to initiate the Splashboard v2.0. Version 1 works, the least you could ask of a
first prototype, yes? But, it’s riddled with a few undesirables especially in
its clunky mechanical structure. The
positioning of some components makes it cumbersome to use: the LCD is not
conveniently set- it requires pillars to prevent it resting on the components
plus it covers some of the user buttons, the barrel jack is too deep inside,
and the mounting holes aren’t even on the edges. The board is constructed using
through-hole technology which makes it too large.

what will Splashboardv2.0 look like?

4 relays for power isolation to ensure that the
conductivity probe and pH probe don’t give false data when dipped in the same

EEprom and cloud to eliminate the need for an SD
Card reader

A Serial Peripheral Interphase port, SPI which
enables the shift register to take in serial data and an I2C (I-Squared-C)
adapter for low-level communication between components

A Wi-Fi
module to convey data readings
wirelessly through the internet which
someone can view through an internet enabled devices

AVR programming to enhance our product design

Use of Surface Mount Technology to make the
board more compact

A 20 character per 4 lines LCD

Real-Time Computing, RTC, for time stamps

ATmega644 - the main microcontroller

InterProcess Communication, IPC level 1

IRC580 to eliminate the need for two voltage

This will be a good challenge as far as design techniques
are concerned. PCB drawing shall be done usig KiCad. The board will be
double-layer with through-hole plating.

After we’ve made our IoT device, and everything else
imaginable is on the internet of things, doesn’t this conjure up the inevitable
big brother scenario? What happens when everything is ‘online’?

See you at the next meetup!



Hi. I’d like to attach an estimate of power requirements for the Splashboard. How do I do this, if possible. Thanks



Hi. I cannot find the data sheet for IRC580. Please provide a link, Thanks




try searching optocoupler 4n35. It’s the IRC580 equivalent.



Here’s the code file: https://drive.google.com/open?id=0BwpTqo-zXGAdVkhHWHpfMDZFLXNhYUZpVnZOU1k4Zzg0aTVJ



Hi. I’ve attached draft specs for the power subsystem. 1 of 2. Comments?



2 /




I forgot to add that the power LED is not shown in this schematic, but is in the UX schematic. Ideally, the switch should be in that too. I’m not able to add other documents because I’m a new user to this forum. Ed., any other way to post my sketches for comments and review?



In lieu of digital napkins to sketch on, I have created tentative schematics. Note that the indicated Buses and Nets will not compile, but I find that it is easier to conceptualize with more descriptive names on the pins. Please feel free to comment. :slight_smile:
To start: This overview shows how specific data protocols are used between different parts of the system.



Page 2 of 12: I redrew the Power Supply schematic because I needed to include the Power LED to help with board layout. I haven’t determined the LED to use, so the 1K resistor is simply a placeholder value.



Page 3 of 12: The Microcontroller. Although bigger than the ATMega328, it’s still somewhat constrained by the number of pins available for the UX and Sensor IO. I assumed that the use of analog mux/demux would help. Illustrated are the compositions of the data bus, and the signals that go through them.



Page 4 of 12: Every MCU needs a programming interface. I added an option for USB vs FTDI. Thoughts?



Page 5 and 6 of 12: Multiplexers. Incoming channels come in from the left, and output channels are to the right. I forgot to specify the chip. It’s a 74HC4052. Ideally, a single chip with two input and four output channels would be used, but I haven’t found one yet. Both pages are similar, but are included to complete the overview.



Page 7 of 12: The ESP8266 WiFi module shares a hardware UART with the programming headers, but instead of a jumper, we could use software switching through the multiplexer 1.

1 Like


Pages 8 and 9 of 12: Basic I2C connections to the EEPROM and Clock chip.



Page 10 of 12: I assumed that we can find Double-Pole relays. That’s because without them, we need to double the number of relays if we want to isolate both Power and Ground Channels for 4 effective switches. I think that 8 single pole relays would be too many for the board. Thoughts? Furthermore, Solid State relays are great, but only seem to be available for AC switching. Maybe my GoogleFu is not powerful enough?



Page 11 of 12: Sensor Input and Output Channels, and Protocols available per connector. All three analog channels are connected to the ADC circuits of the MCU. 3 of the 4 Digital pins are PWM pins.