From BEAM Robotics Wiki

Jump to: navigation, search

We apologize for the need to display ads on this wiki. But somehow we must pay for hosting.

The BEAM Glossary says:


[edit] Building & Using BETS1

"What is the best solar engine?" is one of the first questions asked by most BEAMers. But "best" means different things to different people. End use, parts availability, cost and efficiency are all components of "The Best Solar Engine." End use, parts, and cost are easily determined -- but "Efficiency" has been difficult to measure. A number of circuits have been posted (thanks Wilf) showing how to increase efficiency and optimize the design. But hard data is not available documenting circuit efficiency.

The purpose of the BETS1 project has been to develop a way to compare the relative efficiency of solar engine designs, to measure the effects of any 'tweaking', and to optimize circuit design.

The requirements were:

  • Generate a number related to the [SE] performance
  • Easy to use
  • Simple to build
  • Inexpensive (cheap)

Relative performance can be measured by counting the rotations of the motor. The idea being that when more power is sent to the motor, a higher number of rotations results. There are a number of tachometers and counters commercially available -- but they are large, and expensive, and are definitely not BEAMish. To measure the performance of a solar engine we simply hook some kind of solarroller to the test stand.

Fortunately, in most junk boxes, there is usually an old computer mouse. Under the hood, there is a slotted disk that is rotated by the mouse's movement and a photo interrupter circuit to send the motion information to the computer. Software in the computer interprets this information and moves the curser on the screen accordingly.

BETS1 Front View (click to enlarge)
BETS1 Front View (click to enlarge)
BETS1 Side View (click to enlarge)
BETS1 Side View (click to enlarge)

[edit] Building BETS1

  1. Increase the grip between the axle and the slotted wheel by adding a small rubber tire. The rubber tire from the Solarbotics Solar Racer works great.
  2. Mount the motor/ tire so that it will spin the slotted wheel. Glue guns are pretty handy here. Trim away any plastic necessary for a good fit. Make sure the axle spins freely.
  3. Done!

[edit] Using BETS1

  1. Download and run "Mouse Clocker" from the web ( It's a simple and fun program that measures the distance the mouse moves on the screen. As the author, Ryan Dupont, says: "See how long it takes for your mouse to circumnavigate the earth!"
  2. Unplug your mouse and plug in BETS1. Start your solar engine! The revolutions will be displayed as distance traveled.
  3. When finished, unplug BETS1 and plug your mouse back in.

[edit] Hints / Tips

  • Motor positioning is critical. You need good 360 degree contact between the tire and the slotted wheel. But with minimum pressure. Too much pressure can overload the motor and not let it spin.
  • You may need to shield BETS1 from room light. Some photosensors used in mice can be blinded by too much light.
  • Mouse Clocker measures the distance the cursor travels across the screen. You will need to manually move the cursor back to the starting point when it reaches the edge.
  • You can adjust the sensitivity by changing your mouse speed. Slower speeds require more revolutions to travel the same cursor distance. Try it with your regular mouse. Remember, you are measuring relative performance, so keep your mouse speed constant during testing.
  • Since there are two slotted wheels, you can install two different motors.
  • Start your cursor in the middle of one edge of the screen. You can then watch the movement.
  • When testing, discard the first measurement. Then use an average of the next five or ten measurements.

[edit] Summary

BETS1 is cheap to make (do you have a spare mouse handy?) and simple to use. Plug BETS1 into the mouse port, let the beam engine run and get the distance. At last, a hard number for the performance of the beam engine.

[edit] Test Results

I've settled on running 5 tests, each having 5 SE "pops". This gives an average of 25 "pops," which should even out any bumps. Also it is convenient for logging the values -- I generally run 2 sets of 25 "pops."

Just to get my feet wet, I ran a couple of trials. For these, I've created two metrics: distance per "pop" and distance per uF.

The first trial I conducted was of a stock FRED SE (I already had in my box) with varying capacitors; 2200, 4400, 6600, 8800 uF.

Distance (inches) per "pop" of a FRED SE, as a function of storage capacitor size (click to enlarge)
Distance (inches) per "pop" of a FRED SE, as a function of storage capacitor size (click to enlarge)
Distance (inches) per uF of a FRED SE, as a function of storage capacitor size (click to enlarge)
Distance (inches) per uF of a FRED SE, as a function of storage capacitor size (click to enlarge)

I found that on the FRED tests, it appears that there is a non linear relationship in distance travelled to capacitor size. By increasing the capacitance by a factor of 4, you get an 8 fold increase in distance, and almost double the distance per uF (octane?).

The second set of trials I conducted compared FRED SE vs. 1381 SEs using identical capacitors and solar cells.

Distance (inches) per "pop" of two SEs, FRED vs. 1381 (click to enlarge)
Distance (inches) per "pop" of two SEs, FRED vs. 1381 (click to enlarge)
Distance (inches) per uF of two SEs, FRED vs. 1381 (click to enlarge)
Distance (inches) per uF of two SEs, FRED vs. 1381 (click to enlarge)

It was interesting to watch the cursor move across the screen. I noticed that the cursor did not move the same distance each time. This was confirmed when I did the calculations.

So, I set up a 555 timer with a 5 volt regulated supply to pulse the motor (a Solarbotics pager from the solar racer) and ran tests on it. Interestingly enough the distance traveled was very nearly consistant [each within about 5% of the average -- ed.].

Distance (inches) per "pop" of a 555-based circuit (click to enlarge)
Distance (inches) per "pop" of a 555-based circuit (click to enlarge)

Since the 555 can be considered to output pulses at a fairly constant frequency, this means that the BETS1 is measuring SE behavior to within about 5%. Not shabby for something you can build from a spare computer mouse!

This wiki is sponsored and hosted by Interactive Matter
Personal tools
Ads to finance this wiki