So I've been working on a robotmeter mkII (robotmeter.com). Currently it uses a liquidware touchscreen, but I'm thinking of doing a bit of a paradigm shift. Right now the code for getting and interpreting the sensors sits on the arduino, which isn't the best thing in the world because the touchscreen (which is programable in the arduino environment) has more computing power. So it should really be the part the interprets the raw signals. This would make the arduino a mere ADC device.
What appears below are my rambling thoughts on this that I emailed to a friend:
This line of thinking started when I began considering buying a pioneer f700bt (http://www.pioneerelectronics.com/PUSA/Products/Navigation/In-Dash/AVIC-F700BT)
I posted this on my MKII robotmeter thread:
I'm terribly indecisive so here goes...
The arduino is a great device for getting analog signals in and posting them to something else for display, logging, consumption. This can be done via usb to a computer, a computer-like device, or via serial to an LCD screen, etc. The arduino is not great at doing any display driving. The 2 line lcd I used in the first version of this project is about all it can do.
The decision here is that now that I'm going to use a real graphical display of some type now and I can't decide between the $174 touchscreen or a $150 device that'll let me display to a headunit/or other screen. (In case you are curious that device is a beagleboard.org board and will allow for lots of other awesome stuff later).
The breakdown is like this:
arduino $35
beagleboard $150
sensors $depends ($24 oilt autometer, $70+ oilt defi)
or
arduino $35
"doublewide" touchscreen $174
sensors $depends
or (and i won't do this but just for comparison)
HKS CAMP II $500 (it does OBD II stuff out of the box that I'll probably never get around to)
boost sensor and harness (the only other sensor i saw was the egt) $135
So, I'm thinking about re-engineering this all so that the arduino part just collects the raw signals from the sensors. This would be one code project and it would already have a version 1 done as soon as I organize the code.
The second project would be to have a device interpret these signals and display them (beagleboard or touchscreen...the touchscreen actually has more computing power on it than the arduino). This might even be forked into multiple projects. And i'd probably only do 1 on my own.
Thoughts?