10, Sep 2012 / rebound


It’s hard to believe, but summer is coming to an end. If our lack of blogging shows anything, it’s that it was a busy period for REbound. And while development efforts are only escalating here in Boulder, the team wanted to take a quick break to share some lab experiences. That’s right, REbound started testing this summer!

When we started this effort in June we weren’t sure how much could be accomplished on our nascent startup budget. Thankfully we’ve discovered the power of open-source hardware. You may not be aware, but a vast group of individuals has been working for the past 10 years to bring easy to use, low cost mircrocontrollers to market. One of the best, Arduino, is now so prolific and ubiquitous, that we had no issues building our own test equipment.

Like everything we do, we’re approaching testing in an iterative manner. We started with some basic thermal analysis of a potential storage material. To accomplish this, we needed a high accuracy calorimeter with low heat losses. There are various options out there, but none gave us the accuracy needed at a cost within an order of magnitude of our budget. The data acquisition (DAQ) alone was cost prohibitive. So, for a low-cost solution, we turned to the open hardware movement.

Using an Arduino Uno and a Adafruit data acquisition board, we were able to put together our own DAQ system for less than 5% of a commercial system cost. Using other off-the-shelf components, we were also able to build a calorimeter fitting our needs. The final machine is equipped with:

  1. Automatic heat loss compensation
  2. Fully controllable/independently measured 0-300W heat input
  3. 3 independent crucible temperature measurements
  4. Temperature measurement error < ± 0.5°C
  5. Remote SD or direct USB data collection
  6. On-board clock for real-time data collection
  7. Integrated and isolated magnetic mixing

When we first started down the open hardware path, we were admitting highly skeptical. Our main concern: would open source hardware compete with vastly more expensive DAQ systems? The answer to that question is a resounding ‘yes.’ Sure, if you need 100kHz samples taken from hundreds of independent accelerometers, go with the $10k system. But for accurate, easy-to-control temperature measurements, we found the available open-source hardware options met or exceeded all our needs.

First, we successfully calibrated the system with excellent match to expected values. With the sensors calibrated and the system performing as expected, we ran over 30 trials analyzing how our storage material performed under different conditions. These tests, although fairly basic, indicated the material performed as anticipated.

These tests wrapped up in September and construction of a lab-scale system prototype has begun. Sticking with our open source success, we upgraded to an Arudino Mega to manage the prototype’s additional sensors and control points.

Getting this thing up and running by the end of the year will require a tons of work, but one thing is for sure: we wont be held back by the inability to get low cost, high quality, microcontrollers.


Leave a Reply

Your email address will not be published. Required fields are marked *