Battery Testing

While still only able to use 3 of the 4 motors a general test on battery properties can and was conducted. To give a semi accurate simulation of real flight we ran the three motors with enough lift to hold the quad on it's own weight (70%) of the output. Now in a real flight the fourth motor would draw current so the battery would die faster at that speed however, the motors would need a lower speed in order to maintain hovering. While the actual battery life is likely to be slightly less than the numbers below these are still relatively close and when voltage at which "Safety" becomes an issue would reside around the same point.

time (s)	voltage
0	12.56
75	12
174	11.8
264	11.6
345	11.5
425	11.4
516	11.3
627	11.2
761	11.1
960	11
1122	10.9
1216	10.8
1288	10.5
1298	10.4
1308	10.3
1317	10.2
1324	10.1
1330	10
1336	9.9
1340	9.8
1344	9.7
1348	9.6
1351	9.5
1354	9.4
1360	9.3

After plotting the data which is sadly rather difficult to put into blogger there were two graphs created 1 with the 0s time included one without. With a 3rd degree polynomial fit these points come very close to the curve. When it comes to detecting the voltage of the circuit the simplest and cleanest solution is to follow an idea presented in this webpage. http://www.instructables.com/id/Zener-Diode-Shunt-Regulator/?ALLSTEPS Using an 11V Zener diode which can be found at Jameco.com we could make a circuit that detects when the voltage drops to a point fairly close to the when the voltage becomes starts dropping rapidly. At that point a message will be sent to the Android stating that the quadcopter needs to land but will not take over and force a landing until some time after. Along with this at the 9.3V mark the first motor turned off, for some reason when the voltage goes low individual motors turn off one at a time until they are all dead instead of everything shutting off at once. With this in mind a dead battery would cause a nose dive crash which is even more dangerous then an abrupt shutoff.