April 29, 2009 AT 1:05 pm

Tweet-a-watt in the news…

3466103321 Dc4D548Dab
3467207864 0Ca17973Df
The little power tweeter is making the engineering rounds…

MAKE Volume 18: ReMake America

ReMake America! These challenging times have presented us with a rare chance to try out new ways of doing things. The opportunities for makers are terrific — we can start at home to remake manufacturing, education, food production, transportation, and recreation. In this volume you’ll learn how to make an automatic garden, heat your water with the sun, monitor and share your home energy usage, and more in MAKE Vol 18.

Tweet-a-Watt Successfully Combines Personal Fabrication, Social Messaging, and Green Electronics by Matthew Traum @ Design News.

…the Tweet-a-Watt, monitors and reports energy consumed by appliances and electronics plugged into it. As reported in “Show of Power” in Mechanical Engineering Magazine, electricity from a capacitor runs an embedded XBee wireless module within the Tweet-a-Watt just long enough to send daily data through a computer to the Internet. There a social messaging service like Twitter displays the results.

In his 2005 book “FAB: The Coming Revolution on Your Desktop – From Personal Computers to Personal Fabrication” Neil Gershenfeld describes the future of engineering design as moving away from mass production to personal fabrication. According to Gershenfeld, driven by the desire for personalized products, people will begin modifying technology by adapting commercial products for unique personal applications. A look at the Adafruit Industries Web site suggests the first wave of Gershenfeld’s personal fabrication future is already here.

SHOW OF POWER by Jeffrey Winters @ Mechanical Engineering magazine, ASME Int’l (the American Society of Mechanical Engineers).

For most people, the only time they are aware of their power consumption is when the utility bill arrives. Then, it’s tidily totaled for the month. But a gadget that would enable users to publish their daily electricity usage won a green industrial design competition in February…

Kits available!

Related

3 Comments

  1. I applaud the innovation. I doubt the accuracy.

    One quick thing to make it more accurate very quickly. Take a sample of the voltage and hold it, then multiply it by the current sample taken 2.78ms later. For the European folks, multiply it by the current sample taken 3.3ms later. If you are taking samples once every 1ms, then just go with 3ms delay between the samples before you multiply to get power.

    This will not truly adjust for capacitive power factor variations, but it will be closer to reality than taking simultaneous samples.

    It is a very good idea to use this device for such a monitor, but it is not very accurate. In fact, the current lag problem is probably less of a problem than the inherent inaccuracy of the op-amp that Kill-A-Watt is using. Beyond even this, the current sense resistor they use, as well as component tolerances reduce the overall accuracy.

    At best, this is a good estimate of power usage at any given time. Still, it is better than not knowing anything. Maybe it is science that makes me want this to be a perfect measurement. I must face the fact that having a somewhat accurate reading per minute is still better than having one reading per month in my electrics bill.

  2. its not that accurate at all. its not intended to be very accurate and we dont claim it is, especially as its only taking data once every 1 ms and has to be calibrated by hand. this adds an error of perhaps 10%
    the sense resistor and opamp and component error (they’re all 1% metal film) pale in comparison.
    however, we feel that 10% is worth it to keep the project simple and microcontroller-free

    do you have any references that show adjusting for capacitive power factor corrections by not taking simultaneous samples would make it more accurate?

  3. JorgenFlueror

    First of all, I hope you did not think I was criticizing. This is a very innovative idea and it is well executed/documented. I don’t have any good references on how to adjust for power factor that would apply here. Delaying the current measurement by about 3ms would put it more in line with standard AC power usage, wherein current lags voltage by 60-degrees in the cycle. Highly capacitive loads manifest themselves through alterations in this lag-time.

    Revenue grade meters use specialized ICs to track that lag and adjust for it in their measurements. Using those ICs isn’t very hard, but would add much complexity to a design like this.

    Although power factor is a problem to utility companies, like you point out, it is not a huge factor in accuracy for the tweet-a-watt.

    A very innovative design you have done here.

Sorry, the comment form is closed at this time.