LPLC - Low Power, Low Cost Microprocessor Development Board
Graduate from Hobbyist to Professional Microprocessor Development tools.
You may have heard about other microprocessor development boards, and you may be asking: “Why does the world need yet another board?”
Well this board is different. Let me show you how.
While there are plenty of development boards available on the market, none of them quite hit the spot for me, none have that Goldilocks principle. They were either too small, too power hungry, too expensive, or just too limiting.
So how does the LPLC board ( that stands for low power, low cost by the way,) how does this board differ?
Well, my focus has been on the following four areas:
Low power - I want to build circuits that run off a couple of AA batteries for years, or coin cells for months.
Low cost - I want to be able to put the board into a project and not come back a few weeks later to re-use it somewhere else, just because I can't afford a new board.
Powerful - I want to have lots of code space, and use a proper, professional development environment. I write complex code, so I expect a decent debugger.
Small - I like to make tiny projects, as some of my stuff ends up in portable applications and even magic tricks. Plus I'm getting interested in wearable computing, and I see this board as a stepping stone to even smaller, more powerful designs. I don't expect this to be my only Kickstarter campaign!
As you can see from the prototype, the board is compact but still provides over twenty I/O pins. The format allows for plugging into a prototyping board, which is how we would expect initial development to occur. I keep at least one board with pins attached for prototyping work, and the rest don't have header pins fitted – I solder wires directly to the board.
A watch crystal oscillator is included to provide a low power, accurate timekeeping source, but this can be switched off if required to reduce the already low current consumption to virtually nothing – just 9nA. The board only consumes 60nA with a special wake up timer running. ( Which is still “nothing” for all practical purposes. )
A small LED is fitted to the board. Coupled with the template software that will be provided on the LPLC website, this will help you kickstart your own designs – start with the template code, load it up and check that the LED flashes. If it does, you know your hardware and basic code are working and you can quickly progress your own code. Decades of software development has taught me that the initial “board bringup”, getting unknown software working with unknown hardware, can be a major challenge. I've already seen how template software can speed project development, and remove those early frustrations.
The board will of course be professional manufactured, with gold plating, green solder resist and white silkscreen, just like my other board here
which is another PIC18F board, designed specifically for wearable computing applications.
Back to our board, the LPLC board is based on a Microchip PIC18F27J13 processor, which is small yet powerful.
I've been designing embedded systems for over thirty years, using Microchip controllers for 20 years and writing about them for the last ten. I'm quite comfortable with these chips and love the variety - it seems like there is a size, power and peripheral selection for every conceivable project. And with that variety there is very little waste.
I write about the PIC microcontroller in the magazine "Everyday Practical Electronics", and for the past six months in my column I've been designing a small, low cost, low power development board that people can create at home, if they have board etching facilities. Unsurprisingly not everybody does, and the interest in making my board available triggered this Kickstarter. Not only could I reduce the cost of the product through volume purchasing of boards & components, but I could reduce the size of the board by redesigning it using surface mount parts. A double win.
I will be using a PCB manufacturer in the UK who I have used many times over the last five years ( they manufacturered the board shown above too. ) I have secured, in principle, the volume quantities of the key components through a UK distributor who I met with last year.
I have my own reflow oven for soldering the components, but should the demand be very high ( I wish! ) then the PCB manufacturer I am using has the capability and capacity to do the soldering for me.
Here is a quick peek at the schematic:
It's quite simple, but most of the skillful design effort went into the choice of the processor and the minimalist selection of components: enough to be useful, but not so many that the range of uses the board can be put to gets constrained. And all the time while designing the board, I was focusing on “Low power consumption, low cost”.
Here's the board layout
Compact enough to fit into tiny sensor designs, but still possible to manufacture at home if you have the kit.
The board has a wide power supply range of 2.0 to 3.6v which means it can be powered directly and efficiently from batteries. The PIC18F has exceptional low power operation modes – down to 60nA with a wake-up timer running – which enables ultra low power sensor designs that can run from coin cell batteries for months if not years.
It can be powered from two alkaline 1.5v batteries, three high capacity rechargeable batteries for high power, portable applications or a single coin-cell, such as the CR2450. Coupled with a low power Bluetooth transceiver, a tiny sensor based off this board would last for months on a coin-cell.
Here is an example of an LPLC prototype board implementing a simple, battery powered, single channel oscilloscope, driving an equally low cost colour LCD display on a prototype board.
The PIC18F is quite a powerful processor for its size and cost – it runs at 48MHz, and can process 12 million instructions per second, continuously. It's designed to be programmed in C – which is one of the most efficient languages, after assembly. Program storage is 128KB which is massive compared to some of the Arduino – even the 328 version for example ( a bigger version than ours ) runs at 20MHz, has 2KB RAM and only 32KB of program memory.
The growth path for processors within the PIC family is enormous. From six pin processors like this PIC10F, no bigger than a spec of dust
to 32bit, 160MHz parts with 2MB flash and 512KB RAM on board, on chip. And with several hundred devices in between. All of which are accessible and solder-able by hobbyists. All programmed and debugged through the same IDE. This one here has four I/O pins, a timer, 1KB of program memory and 24 bytes of RAM. Small but still quite useful.
Programming the board is done with the PicKit3 programmer, available from electronic component suppliers or on Ebay, where there is a healthy trade in “clone” programmers. Microchip make their hardware designs open, meaning third party vendors can enter the market and help drive costs down. When I started in embedded systems you paid around a thousand pounds – twenty years ago! - for the tools to do this. Now, a programmer & debugger unit will set you back 20 pounds, and last a lifetime.
One of the key advantages to using this board is the free development environment and (also free) range of programming languages made available by Microchip.
The MPLAB X IDE is cross platform ( Windows, MAC and Linux ) and is based on the popular NetBeans IDE. It has an excellent editor that understands the syntax of your programming language, and cleanly integrates code editing, code compiling and debugging all one one environment.
The Microchip debugging environment is fantastic – due to a small amount of additional hardware placed in every processor, the PicKit3 programmer unit can perform hardware level debugging making the debugging of complex, real-time interrupt driven software a breeze. It's a professional development environment, providing the kind of tools that professionals expect and demand. And all available for free, from the Microchip website.
This board will be ideal for people who have outgrown the Arduino platform and would like to enter the realm of professional embedded system development, while maintaining the link to hobbyist oriented tutorials, examples and support. While the Arduino may be great for “show and tell” projects, the LPLC can be used to create practical, complex, useful projects.
A key component of the LPLC will be the the provision of template software project files, enabling you to kickstart your projects. I will provide tutorials for designs based around the board, including the driving of LCD displays, realtime clocks and analog interfaces. Early adopters will be able to help shape the content of those tutorials. The board will will also be compatible with that board created for my magazine column, and therefore any code created for it.
The datasheet for the processor can be found on the Internet; a simple search for “PIC18F27J13 datasheet” will find it. MPLAB X and the xc8 C compiler can be downloaded directly from the Microchip Website.
The board will be made available for £10 inclusive of shipping to Europe, and two pounds more to the rest of the world. The boards will be manufactured in Europe, shipping from the UK and Ireland. The header pins for the boards will be supplied un-soldered to give you the option of soldering wires directly to the pads on the board if you so wish.
The design is open source so you can, if you wish, create your own boards based on it.
If I reach £3000 I will pre-program the boards with a serial bootloader, enabling the board to be programmed from a PC using a simple TTL level serial cable. USB based serial cables are available for less than £5, and are a useful accessory anyway.
Risks and challenges
I've been designing PIC microcontroller products for 20 years, and I have built two prototypes of the LPLC, so I have strong confidence in the design, both of the circuit and the board.
I've done a number of volume run boards, so I am aware of the pitfalls when designing boards for volume manufacture. A good working relationship with suppliers and board manufacturers is key, and I am not changing my processes in any way ( such as using a different, low cost supplier ) to fulfil the kickstarter project.
My main risk would be if I get a significant over-subscription ( which would be very, very surprising. ) In excess of 300 boards, and I would struggle to achieve the delivery timescales stated. If this is the case, I would get my board supplier to assemble the boards too - a service they already supply. There are no difficult or unusual manufacturing techniques called for, so doing that at short notice would not be an issue.
As a columnist for the magazine Everyday Practical Electronics I have a large body of published work using Microchip parts, so I am not taking any risks choosing the PIC as my processor of choice for a low power design.Learn about accountability on Kickstarter
- (45 days)