Last minute tragedy…

So it was race day and all was calm in camp renegade. Our robot was working and well built. The only issue we had was that the front sensor wasn’t forward enough. With it being 9 in the morning all of us in the group assumed that 2 sensors wouldn’t take 6 hours to mount and get working…………..WRONG!!!!!

As we began to mount the front senor after much testing it was apparent that this wasn’t going to work with the switch in the way and that color sensors were to be put on each side at the front. Doing this would cause a lot of issues as our steel ramp was in the way. After many efforts they fit and we began testing again. Using the voltmeter we were getting a great range out of the signal, ranging from .05V – 4.3V. But when it was on the sumo table it didn’t work………there was something wrong with the code. Once the code was adjusted there was no range on the signal staying at 0.55V. This continued back and forth for 4 hours and the fight was getting close. Eventually we got everything working and were ready to go.

Match:

  1. Our Robot ran right off the table
  2. Wouldn’t run – just completely stop
  3. Forfeit
  4. Forfeit

These results were so disheartening. We had tested the robot out several times before the match even beating everyone we faced. Still an unknown reason our robot ran right off the table and stopped working. The most basic of code was put in after the match and still nothing. All the connections were checked, so the only thing I could think of was that the chip blew.

Hindsight is a wonderful thing. What I’ve learned from my experience in doing this project is that the basics and fundamentals should be done to 100% perfection before even thinking on making improvements. I believe as a team we all wanted to do our best and by that we came up with all different ideas. But what was needed was high torque and lots of weight as after all it is Robo sumo. If I was to do it again I would have delegated a sensor each per person and for them to be an expert on each sensor and share an over view with the team. Then the design would be heavy, high torque with a simple design. Combining that with precision sensors I felt we would have done better and enjoyed it more.

All in all I really enjoyed the atmosphere for this project. It was enjoyable to walk into class on a Friday morning for Johns comedy show and to grab a coffee and then go down to play with robots. In college we tend to read a lot about applications but very rarely see them. The first time I really was amazed was when I got how a color sensor worked. It really blew my mind of how such a small trick with voltage drop could have such a diverse application. Robosumo was my favorite part of the week and would like to thank Dave, John, Ted and Damon for all there help

Advertisements

Final blog before becoming victorious………

Due to the Easter holidays little work was done and due to this no blog was made. Fortunately though I spent the entire weekend before the competition building the robot and mounting the sensors. This is the finished product

20150421_114951
The chassis is made out of a thick perspects that was laser cut. My first issue, was having to mount the 2 motors onto the chassis. The main issues with this task was that the terminals for each motor were so tight that they might be touching, how to hold them down and what space would be left inside for the breadboard. To combat this I just soldered the connections and then carefully wrapped each connection in thin strips of insulation tape. I then drilled two holes of each side of the motors in the perspects and matching holes on a small sheet of metal. I then ran bolts through all the holes and a nut was tightened on the other side of the perspects. The motor anchor can be seen below

20150421_114143

I then left the full length on the bolts so the 9V battery can slid perfectly in between them and then the breadboard sits on top so as not to be touching the wheels.

20150421_114302
The next task was putting a front bumper in with a switch in it so when we make contact with another victim the switch activates causing our motors to go into beast mode. Mounting the motor was straight forward with a hole drilled out in the middle for the switch to stick through which was screwed to the perspects. This in turn caused a problem as the switch was intruding on the front colour sensors space. To counter this I tapped the switches terminals from exposure and pushed the colour sensor beside the switch as close as possible and facing it at an angle.

20150421_11432120150421_114341

Only having two wheels meant that the front bumper was always touching the ground meaning it’s unlikely that another robot will be able to get under it but just in case a lip was mounted in the back extremely tight to the ground

20150421_114353
Due to the tight space in between the bolts and the rear bumper, the rear colour sensor was constructed on copper sheet rather than using connector blocks.
I then bolted two separate sheets of aluminium onto the front and back for the body and carefully bent them until them met in the middle and bolted them together. The final sensor was to be the range finder. A hole was cut out of the front of the body and a bracket mounted inside to hold it down. This was quite trick as the space was limited causing me to mount it upside down.

20150421_114412
The bot was almost complete, only for the sides. Both sides were made out of the aluminium but were put on differently. One side was permanent and the other with a hinge to access the battery and breadboard.

All in all I’m very happy with the outcome but once again I really underestimated how easy problems arise from such simple tasks. But in saying that I really do think we have a fighting chance as our robot has a lot of good aspects going for it. It’s a brilliant design that can’t be flipped, excellent wheels and with Ronan (hardest worker in the group) working on code. The finishing touches should be done this evening with all the sensors being tested again before the big day. Just in case you missed the opening photo, I give you
SHES FAIRLY TIGHT TED
20150421_114951

The Eagle has Landed

Last Friday there was no blog input as the race to the wall had just finished and our robots were taken apart and finished with. Ronan found a great website called finger tech which as of the 5th of March I ordered 2 motors and 2 gears from the company in Canada. The wheels were made of the highest traction polyurethane material in the market paired with 16mm gold spark gear motors, leaving us as a force to be reckoned with.
So a week has past and we are still waiting for the gear to arrive. The next sensor we have to understand and learn its concept is the range finder. John was kind enough to show us in the 1st 1 hour’s class and that the key trick was that it needed to be mounted 2cms back to get the full range of the entire board. Our designed is as though the robot itself is a ramp so what will be needed is that the range finder has to be flush and not to stick out which will sure to give us difficulty as it’s a quite compact design. The following too photos were given to us to fully understand the rangefinder and how to wire it, although our pin set up might slightly vary.Final complete circuit                                                              Range Finder Circuit

Once this was given to us we all went down to the lab to dish out tasks and try tackling the very last sensor. Sean was doing some research; Nathan was trying to see if he could design the body for it to be lazier cut, while Ronan and I worked on the range finder. The sensor itself is quite simple but doesn’t sound as so. It has four terminals, a live, a ground, a trigger and an echo. To put it simply the trigger sounds out a signal and the echo catches any of the bounce back cause a potential difference of around 5V. Unfortunately this is too high for the processer and will damage it. This was why a diode was placed on the echo in reverse bias and a pull down resistor leaving around 3V at the input pin. Ronan and I decided that the best range for our Robot would be 80cm as the table is only 100cm in diameter. So Ronan and I set out marks on the table up to 80cm. We then tested span of voltage at different distances and using different resistors and found 3kohms to be the best suited resistor. Everything seemed ready and rearing to go……….

But wait………………………………………………………….…………………………..THE EAGLE HAS LANDED!!!!!!!!!!!!

As I get home the packet has been delivered on the Friday evening. I opened it up and the motors and wheels have been delivered.

Wheels and motors connectedSo after a bit of building I put this little beauty together down below. The idea is to have only 2 wheels with a razor sharp front and a metal wedge at the rear with the colour sensor mounted on it and so it cant be flipped. The front colour sensor will be through a hole at the front. I am yet to design the  the cover but should be like a ramp with the heavy batteries at the front under the roof

Side of base Rear with hole to screw colour sensor

Week 5 Physically building the robot.

Set the mood, tensions are high, 2 days before the race and everyone is eager to build the final robot. So on Wednesday evening our team decided to go to the work shop after class. We had a design in mind. The minute we set foot in the work shop and tried to apply our design there was a problem with the front axle. The second that happened I knew we under estimated the difficulty that would come with physically building the robot. The issue arose when the front axle couldn’t be pushed back far enough leaving our 10 by 10 design in peril. After several other attempts mounting the axle and a flat plate it was already closing time. This really began to worry me as we only had 1 more day. So being the only one in the group with I took the project home and said I’ll stay up tonight and try work something out.
So when I got home I tried a few more designs and the closest I got to mounting a main plate and an axle was when I spaced the back and front of the plate of the motor with nuts and place the axle in between the plate. This looked promising but then I seen the motor itself was dragging on the ground leaving the wheels hanging in mid-air.
It was 2a.m and I finally had enough of trying. I decided to go to bed and as I was falling asleep I had a brain wave that brackets with holes should be hanging from t the plate and the axle goes through the holes under the motor itself meaning we could push the wheels as far back as possible. I got up and it only took me an hour to do. I was then back up at 7 and had only 2 classes on the Thursday so I spent more time finishing the robot where I mounted the colour sensor at the back and the switch at the front which gave me no problems. I then met up with my group and we fitted a metal box strapped on the plate, in that was the battery pack and on top was the breadboard. The design was finished!!!!!
The above notion sounds great but it was actually 5pm the day before the race and nothing was tested. We didn’t start off very well as there was no voltage coming out of the pin for one side of the motor. We eventually got it working by simple putting in a 10k resistor to ground. The robot was finally running and the colour sensor, switch and code were working perfectly but after many trials we still couldn’t get the robot under 20 seconds. We finally got it on 20secs give or take. We then went back to basics adding in another capacitor on the 6volt side to stop noise and cleaned up the circuit as the chip was only giving out 2.5 volts at best causing such a slow speed hoping it was just a loose connection. We then thought it was down to the chip constantly resetting due to the high pitched screaming from the motor. We then asked for the assistance off Ted and David as we felt like we did everything and after 4 hours of trouble shooting we called it a day on 20seconds. Ted diagnosed the motor being damaged……………
This news was quite a disappointment as the amount of effort was for nothing and that our whole design was based on a broken motor. But I will say I learnt a lot though and that was a good lesson on how much work it takes for the bare minimum to build a robot and to get started straight away on the next robot task. I’m hoping our next design will be a bigger success. Below was our abomination

image-4 image image-3

Week 4: Just around the corner

On Friday our class was cancelled. But due to the tight schedule with the battle to the wall and back was next Friday a 2 hour workshop was held in the afternoon. This actually turned out to be a better for our team as very few people showed up which gave us more one on one time with the lecturers to ask questions and to trouble shoot.

So as to keep up with last week’s new habits we started off with assigning each of us tasks. It was Sean first, it was his job to look up the plan of attack for the drawing design of the robo, which as of right now is still quite in the junior stage of development being, two wheels, a front roller and a bread board mounted on a flat plate that’s attached to the twin motor. Ronan was writing and testing code. Unfortunately our team was having a lot of trouble with the code and after a while trouble-shooting it with our lecturer it was decided that Ronans code was perfect. Our analysis found that there was 4 volts on the singular motor but when attached to our twin motor that we are using for the race there was a huge volt drop leaving only 2.2V on the motor. To counter act this we thought it would be best to use a 9V battery. This problem turned out to be a blessing as I believe nobody thinks they can use anything over 5V giving us a huge advantage. To be honest if it was up to me the robo would be running off enriched uranium haha. The 9V battery was also was a lot more compact as well as leaving more space and less weight. Also there was the problem with contradicting voltages which left us at a loss. After 10 minutes of deliberation with David we just put it down to charged capacitors on the Arduino board. I myself was working on the actual practical design of the robo by taking it apart and trying to see where it can be improved. I found that the board that was mounted on the roller was far too heavy and unnecessary and that the roller on the front would not do as it would be far too unstable when it hit the wall. To counter act this I decided we should put front wheels on it and that the wheels should be of a similar size to add stability. Our robo is now looking very compact and stable and with the 9V battery I’m positive we will win.

Week 3: A More balanced team

After last week our group felt as though there wasn’t enough structure through our delegation and work loads. So to streamline our project and make our team more productive we decided to sit down at the start of every session and brainstorm ideas and delegate tasks for all of those ideas.Our group is made up of four people: Ronan, Nathan, Sean and myself. The tasks were given out to each member and were as follows:

My duties for the day was to build and get the color sensor (fig.1) working. I first began by drawing out the circuit to gain a better insight and overview. The factors that I knew straight away were that the maximum current allowed through the LED was 60mA, there was a 3.5 volt supply that would have a volt drop of 1.25 volts across the LED leaving 2.25V on the resistor connecting the LED to ground. I began to think that 60mA was far to bright and would also reflect the colour black. After using equations[ I=V/R ]and trial and error I found that 120Ω was the perfect resistance value. The next issue with the colour sensor was with determining the value of the resistor on the pull down switch. This was to be crucial for the output signal. After using equations [Vo=R1(Vin)R1/(R1+R2)] and trial and error a resistor of 10kΩ was used. With this now being applied, when the LED was reflecting on white the output signal was 3.3V and when it was hitting black the most voltage in the signal line was 0.3V. Meaning by using c code (fig. 2) when the signal voltage exceed 3V it knew it was on white.

I was also in charge of recording and analyzing the steps taken today with everyone’s work so their would be an ease of information when writing our blogs. The steps where usually: task – problems – solutions. At the beginning it was all done by addressing the obstacles ahead of us and turning them into tasks. I would then return in the middle of the class and record the issues they had with each of there tasks and finally at the end we had a debriefing to see how we each resolved our problems. With this way of recording our information I felt it gave everyone in the group an overview of what was going on in the project and our development.

Ronan was to work on the limit switch. His first issue was that he found a floating voltage on the pin of 0.8 volts. To resolve this Ronan put the floating voltage on the ground wire by switching around the floating voltage pin. Ronan was also in charge of coding the color sensor by distinguishing between black and white. The only issue Ronan had been that the LED on the sensor wasn’t working, but shortly afterward we found it was as simply to be the resistor on the Vcc was connected to the incorrect pin. Nathan was mostly part of our design team. He was researching for the design of the Robosimo and the material and came to the conclusion that perspects would be the lightest for speed and the most cost-effective. He was also designing our Facebook page where we could have a forum to share ideas when we’re not together. Before finishing the day he also neatened up the components on the bread board. Sean was to research future problems that lay ahead of our project and to print out the data sheets for the chips that we are working with. I felt this week we really made a lot of progress and I really do believe next week could be our week to have the first robot to the wall and back.

Figure.1: Colour sensor.     Figure.2: C.Code

Colour sensorc code

Week2

We started the day by focusing on getting the switch to work. We began by by putting a volt meter across the switch and seeing if there was a difference in the open and closed state of the switch. After installing the switch we moved onto our main goal of the day, of making the motor go forward and reverse with a delay with C code so we can determine the “On” and “Off” rather than constantly switching between the two states as its 6000 cycles a second. The first piece of code we applied didn’t seem to have any effect, nor did the second small adjustment. We spent the following hour altering and debugging the code to no avail. We then went back to the circuitry and found out that our voltage divider was giving out 5 volts and that the soldering on the voltage divider leg had been partly been broken. After soldering it back together it still wasn’t working. As a group we came to the conclusion that the voltage divider chip was exposed to too high of a heat, causing damage from the solder. After applying the new voltage divider to the circuit the class was over. It was quite discouraging as the class felt like one step forward and 2 steps back. But with that beginning said there was a lot to be understood while problem solving especially with the coding. My soldering skills have definitely improved too. Hopefully in the next class there will be a huge leap as we’ve resolved a lot of issues