Firepop
So first off our robot is named firepop because pop is another name for soda and we use soda to put out the candle. We decided to use mentos and diet coke to put out the candle. After numerous experimentation we found that the best mechanism for getting the mentos to work was to drop in mentos into the diet coke through a small hole in the bottle cap (made using the smoldering iron) and then the jet would be directed through the hole. The main limiter to the mentos diet coke fountain was the amount of carbonation in the diet coke and secondarily the covering around the mentos.
We had lots of trouble with basic mechanical aspects of the car. First because of the way we attached the diet coke bottle (we used the 12 oz bottles) we had to weight the back of the car with 2 9 volt batteries (we latter added on another 3 9 volt batteries and a fishing weight lure). This meant that the front of the car didn't put too much weight on the front wheel and that the car didn't bounce a ton (or look like a low rider). Also throughout this project we replaced the batteries twice, the breadboard 3 times, and the logochips 3 times (we burned two of them). We also used three different lego motors for the wheels.
We tried at least 6 different tires and realized that the tires we had were the best but that we had to have a power offset of 500 between the left and right tires for the straight function (cause our car won't go straight otherwise!). We mopped the floors multiple times (which helped a fair amount), and always cleaned the tires of grit (again helped). We even tried velcro on the tires to increase tread...didn't really work though! Basically what we discovered was that chance STRONGLY influenced how well the car would position itself and work overall.
Our code was basic in the idea -- three parts.
1. Spin using the coarse wide angle sensors to find the candle and then point in said direction (stop when the candle (a set value of say 250) is visible in both sensors).
2. Go straight using the difference found in both wide sensors until a certain maximum value, or certain distance, or if a max value is found in any sensor (this uses bang-bang to go straight -- works pretty well).
3. Nudge -- goes straight using the difference found in both fine sensors (aka the same as the go straight but uses fine sensors to drive instead of coarse sensors) using bang-bang
There is also a spin that uses the fine sensors instead of the coarse sensors to position itself.
We had fun playing around with recursion but of course pico blocks takes issue with recursion. But we instituted a part where based on a specific amount of time (using timer) or a specific number of turns completed the candle value that sensors must find would decrease by 20.
We tried to make the nudge call itself recursively but that made the car turn off a LOT! Urgh....lots of problems with this....but we tried to make the code as modular as possible by giving in variables that we change each time we call the methods.
So in the end we were able to make the car get to the candle and position correctly about 70% of time. However, we can't get the mentos and coke to spurt out because by the time the car gets to the candle all of the carbonation is gone (URGH). However, we do have videos of the candle burning the string and releasing the mentos and then candle being put out by the spurt. We couldn't get both of these things to work at the same time (at least not in time) but we did find some ways to improve the probability of this. First off, keep the coke very cold! minimize the time the car takes to get to the candle, decrease spins ect in the code.
But what we realized was that our mechanism for burning the string and the delrin actually is INSANELY adaptable! So you can easily attach a balloon to the delrin piece we made. And since the car finds the candle pretty reliably, this would work very easily. Or, you can use the string to tie up a styrofoam cup in the middle of the delrin piece. Now, if you then get the car positioned correctly the candle burns very quickly through the cup and whatever is in the cup dumps on the candle. I used washing powder and it worked perfectly!
Thus our car and mechanism for putting out a candle is very changeable depending on your wishes. We like to point out that even though out car didn't work, we put in a lot of time and effort and tried a ton of different things. I will post the pictures and videos tomorrow.
We had lots of trouble with basic mechanical aspects of the car. First because of the way we attached the diet coke bottle (we used the 12 oz bottles) we had to weight the back of the car with 2 9 volt batteries (we latter added on another 3 9 volt batteries and a fishing weight lure). This meant that the front of the car didn't put too much weight on the front wheel and that the car didn't bounce a ton (or look like a low rider). Also throughout this project we replaced the batteries twice, the breadboard 3 times, and the logochips 3 times (we burned two of them). We also used three different lego motors for the wheels.
We tried at least 6 different tires and realized that the tires we had were the best but that we had to have a power offset of 500 between the left and right tires for the straight function (cause our car won't go straight otherwise!). We mopped the floors multiple times (which helped a fair amount), and always cleaned the tires of grit (again helped). We even tried velcro on the tires to increase tread...didn't really work though! Basically what we discovered was that chance STRONGLY influenced how well the car would position itself and work overall.
Our code was basic in the idea -- three parts.
1. Spin using the coarse wide angle sensors to find the candle and then point in said direction (stop when the candle (a set value of say 250) is visible in both sensors).
2. Go straight using the difference found in both wide sensors until a certain maximum value, or certain distance, or if a max value is found in any sensor (this uses bang-bang to go straight -- works pretty well).
3. Nudge -- goes straight using the difference found in both fine sensors (aka the same as the go straight but uses fine sensors to drive instead of coarse sensors) using bang-bang
There is also a spin that uses the fine sensors instead of the coarse sensors to position itself.
We had fun playing around with recursion but of course pico blocks takes issue with recursion. But we instituted a part where based on a specific amount of time (using timer) or a specific number of turns completed the candle value that sensors must find would decrease by 20.
We tried to make the nudge call itself recursively but that made the car turn off a LOT! Urgh....lots of problems with this....but we tried to make the code as modular as possible by giving in variables that we change each time we call the methods.
So in the end we were able to make the car get to the candle and position correctly about 70% of time. However, we can't get the mentos and coke to spurt out because by the time the car gets to the candle all of the carbonation is gone (URGH). However, we do have videos of the candle burning the string and releasing the mentos and then candle being put out by the spurt. We couldn't get both of these things to work at the same time (at least not in time) but we did find some ways to improve the probability of this. First off, keep the coke very cold! minimize the time the car takes to get to the candle, decrease spins ect in the code.
But what we realized was that our mechanism for burning the string and the delrin actually is INSANELY adaptable! So you can easily attach a balloon to the delrin piece we made. And since the car finds the candle pretty reliably, this would work very easily. Or, you can use the string to tie up a styrofoam cup in the middle of the delrin piece. Now, if you then get the car positioned correctly the candle burns very quickly through the cup and whatever is in the cup dumps on the candle. I used washing powder and it worked perfectly!
Thus our car and mechanism for putting out a candle is very changeable depending on your wishes. We like to point out that even though out car didn't work, we put in a lot of time and effort and tried a ton of different things. I will post the pictures and videos tomorrow.
Comments
Post a Comment