An $80 Open Source Chemical Analyzer
An anonymous reader writes "A group of electrical engineering students at UCSB teamed up with some chemists and built an $80 gadget that can check water for arsenic, measure the level of vitamin C in orange juice, and also do simple DNA biosensor tests. The electronics in a blood sugar meter could do all of those things, but their firmware isn't easily hackable. All of the circuit schematics, gerber files, and software for this project are available on their project website. Another team at Denver Metro College is working to improve upon their design. Eventually, it could be used as a teaching tool in chemistry classrooms, or possibly to do blood and water tests in developing countries."
UCSB EE department does a lot of great research. Personally, I'm a big fan of the work that they do in GaN and AlGaN devices. But I feel that this story might be the least interesting piece of research being done at the entire university.
Is this news for nerds just because it is open source? I mean - a potentiostat? Really?
Come on slashdot - lift your game.
You can never know everything, and part of what you do know will always be wrong. Perhaps even the most important part.
What this group should get credit for is open sourcing a cheap design.
The chemistry and circuit design involved are well-established and taught at the undergraduate level. You can easily find schematics for potentiostats online. It's reaching to say that they've built an $80 chemical analyzer, because a lot of prep work and specialized electrodes (platinum!) are needed to run some of these analyses. This is a cheap lab instrument, not something you take out in the field to make measurements. Ruggedizing and standardizing reagent solutions are what would make a field instrument much more expensive.
I'd bet the group didn't make an exaggerated claim, it's the unfortunate nature of science reporting.
The reason it was published is because it provides an excellent tool for teaching undergraduates about the intricacies of scientific instrument design. It's not ground-breaking, it's not revolutionary, it's simply an experiment to teach sophomores and juniors about voltammetry in a cost-effective way that will hopefully stick with them more than "Here, watch quietly while I use this $75K cyclic voltammeter that we aren't going to let you use because undergrads always screw it up and we can't afford the week it takes to get it calibrated and functioning properly again."
My sister opened a computer store in Hawaii. She sells C shells by the seashore.
It was done a bit better than most such projects- it had a case, display, mini-control joystick, mini-USB jack, basic software and firmware, they tested it in several experiments beyond the basics - the DNA analysis was particularly slick. More effort should have been given to coming up with a useful set of electrodes and reagents and a proper connection from the instrument to the electrodes - alligator clips are a bit below the standard of the rest of the project.
I'm just an armchair engineer, but I suspect the electronic design could be improved a good deal with modest extra cost. A proper electrometer-grade op-amp, using proper guards/shields on the circuit board for the op-amp inputs, preferably not running the working electrode input through a switch (though I'd have to think harder to see if it's really a problem as they are using it, charge injection could be a worry), better insulation and electrode connections on the cables, using a uC with built-in USB, (perhaps an ARM from STMicro similar price, it also has better ADC and DAC, 32bit, faster clock, more RAM and more flash), and maybe a precision voltage reference and a couple of trim pots. This really isn't suitable as a portable instrument as they were using it, though I suppose some people might do so . As a bench instrument with a computer connection the expensive screen could be dispensed with, and DC-DC converter might be replaced with a simple regulator. The case is big enough to stuff full of batteries, and they give a much longer lasting and less noisy power supply than a single-battery & converter setup for a portable device, and for a bench instrument USB power should be sufficient. The ferrite on the analog ground with the double HF/LF bypass caps is a very good thing, but feeding it through a separate, battery analog supply would be even better.
"Is life so dear, or peace so sweet, as to be purchased at the price of chains and slavery?" - Patrick Henry
You can get a spayed or neutered open source self-propelled, neural net controlled chemical sensors from the the humane society for about $50. With a little training of the neural network using some bacon and a chew toy it can detect just about anything airborne. Plus you'll have a hard time building anything with a better low end sensitivity. Put out a little food on the back parch and you can likely snag an open source neural net sensor for free. Before acquiring such a sensor unit be sure to check you landlord's sensor policy. Seriously, how we ever made a living while hunting on the savannah with our snoozes is beyond me.