Following the RoadTest of the CapSense MBR3 Evaluation kit, I thought about creating a simple cursor board with 4 direction buttons and a select button with an I2C interface to a microcontroller host.
Before embarking on any venture, I like to do a bit of research to determine if the project is feasible.
- Are there other products on the market?
- What components will be needed?
- What skills will be needed?
- What tools will be needed?
The purpose of looking at other products to two fold. Firstly to see if there is an identical product already available or if you fit into a gap or can sufficiently differentiate your offering. The second is to see if there is a demand for your product. If you have something completely new then it's looking at more generic products to give yourself an idea.
I'd already looked into mechanical buttons as part of the roadtest so we can discount these from further analysis.
For the capacitive switches there do seem to be some comparative products. At the bottom end of the market there are breakout boards and single click buttons. At the top end there are complete touch solutions with inbuilt microcontrollers.
So there does seem to be a position for my board and the other products also suggest that the price point around $10 to $15 if I chose to sell them.
|Breakout board for the Freescale’s MPR121QR2||Sparkfun||$9.95|
|Breakout board for AT42QT1070||Adafruit||$7.50|
|Single touch sensor board with AT42QT1012||Adafruit||$5.95|
|Makey Makey||Makey Makey||$49.95|
|Bare Conductive Touch Board||Bare Conductive||£55.00|
As I've already a bit of experience with the MBR3 chips I decided that I'd stick with those. There are several in the series but the with it's 10 channels will be able to handle the 5 buttons, host interupt and shield sensor. I took advantage of the Cypress offer and order myself a small batch. As this is a surface mount device it made sense to also use surface mount components for the discrete components needed. I quizzed the Element14 community on the most suitable smd sizes and will use these to design my board.
There are three key skills for this project. Desgining PCBs, SMD soldering and configuring the MBR3 device. I've created PCBs before, firstly by hand in the 1980s using etch resistant tape, then in the 1990s with photo resist using a plugin for AutoCAD to design the layout. The capacitive sensors need a bit of care designing but there's a good design guide for CapSense switches that explains best practice for designing sensors and laying out the board so I'm reading and re-reading that. I've started learning how to use Eagle and created some basic layouts. I should be able to build the board and learn as I go. I've also had some exposure to hand soldering SMD components although I've not tried with a solder paste and a reflow oven.
It should be possible to avoid the need for etching by outsoucing the creating of the PCB. There cheapest option seems to be Dirty Cheap PCBs but there is quite a long lead time for them and any boards they make are Open Source Hardware so they are not for everyone. Another option is Ragworm who are based in the UK so can turn around boards more quickly.
Here's one area where I might need to make some purchases. I'm using the freeware version of Eagle which meets my technical requirements but if I want to make my project commercial then there would be a €62 cost to move to the correct license.
Also my soldering equipment is limited to an accient Antex iron with no temperature control, I have done other surface mount work with that but I also managed to melt one chip so a better iron would be a good idea if I am to hand solder the boards. Related to this some form of magnification would be a good idea. I have also thought about building or buying a reflow oven to assemble the boards, if I plan to make a lot of boards this would definately be a good investment. Again I've quizzed the community for recommended tools for SMT.
For configuring the MBR3 these can be programmed by sending a block of data from a microcontroller. However, I worked out that I could remove the jumpers on the evaluation board and use that and the E-ZClick tools to hook up my new board. The advantage of doing that is that I can tweek the settings and analyse the performance of the switches more easily. So apart from making an adapter this aspect is covered.
Leading on from the components, skills and tools this brings us to costs. My bill of materials is pretty small, the circuit I'm planning is just 12 components costing approx £2 for small quantities.
I've done a very rough estimate below, my low end reflow oven estimate of £50 is a self build solution with the top of the range being an off the shelf model. The soldering iron costs I've just picked a couple of tempoerature controlled models. My costing model is also quite simple in that it assumes that I need to recoup all of the one off costs imediately even though some of the tools/software might have longer term value. Another simplification of the model is that I've not scalled the BOM and PCB costs based on quantity so there could be more savings to be made. However these costs don't include any marketting or logistic/shipping costs so there could be additional costs on top of these.
|Item||Hand soldered (min)||Hand soldered (max)||Reflow (min)||Reflow (max)||Outsourced Assembly|
|Eagle License €62||£50||£50||£50||£50||£50|
|Temperature controlled Iron||£40||£200|
|Jigs, wires etc||£10||£10||£10||£10|
|Build of Materials (BOM)||£2||£2||£2||£2||£2|
|PCBs, outsourced in all cases||£5||£5||£5||£5||£5|
|Consumables such as solder, cleaner (estimate)||£1||£1||£1||£1|
|Total investment based on 10 boards||£180||£340||£190||£1,862||£120|
|Total per board for 10 boards||£18||£34||£19||£186||£32|
|Total per board for 20 boards||£11||£19||£12||£95||£30|
|Total per board for 50 boards||£10||£13||£10||£44||£28|
|Total per board for 100 boards||£9||£11||£9||£26||£28|
There's some really useful points from Anna Lowe in the following video.
As a project this does seem viable, there are other products in this area but not one that's directly competing.
The components are readily available and the skills needed to assemble are within my capability.
The costs for the parts is also good but I need to be aware of the one off costs for equipment I'd need to put these together. I'm likely to start with minimal investiment and ramp up as I need.
I've already been learning to use the PCB tools and reading up on the capacitive switch design requirements. My plan is to draw up the designs for the board and get one made up to test before buying a bigger batch. This will initially be more expensive but will mean I won't end up with a load of unusable boards. I've also been creating Eagle Libraries for the MBR3 chips and hope to release those via the CadSoft website CadSoft EAGLE PCB Design Software - EAGLE part libraries