The date is: Fri Jan 29 12:42:22 MST 1999 01/29/1999 12:45 PM Anthonius Sulaiman (Group 2) This week we discussed how we are going approach the first problems (i.e. motors and HC11) and how it is going to look like (i.e. the body, the wheels, and the sensors mounted). Other than that, we also divided our group to two sub- groups of two persons. Each will work on a subject and I am assigned to be the main HC11 programmer. And since we are going to use HC11 as the brain of the robot, we also have to re-open our books in Microcontrollers. Personally, I have done researches on the web and tried to re-learn HC11 programming during the last week. Our plan for next week is to order parts and tools required. Our next scheduled group meeting is on Saturday, January 30. We will start the motor programming on that day. One member will start doing the sensor programming. My idea is to start on the last program of HC11 (EE308 Lab 13) and modify it for the robot. The date is: Fri Feb 5 11:23:51 MST 1999 This week I started working on the HC11 programming. Basically, I snagged my HC11 program for speed control from EE308 Labs and redesigned it to work with the H-bridges. Our plan is to have the HC11 control the H-bridges via 2 lines of PW modulation (for speed control). My work plan for next week is to continue on the HC11 programming and I plan to test it out on Saturday. To control the direction, I am going to try using the same PW modulation (by changing the duty cycles). If that does not work, I will have to use one line to control the direction bit on the H-bridges. If the programming goes well, we are going to start on the sensors. The date is: Fri Feb 12 11:47:38 MST 1999 This week I finished most of the HC11 programming for motor control. I still need to test it out with the actual motor. My HC11 program is to have two lines to control the PWM on OC2 and OC3 outputs, and one extra line (undecided on which output) to control the direction of the motor. For now without the sensor, I only have the program to run the motor at a specific duty cycle. We plan to get together on Saturday to prepare for the design review presentation. We will gather all the information and design that we did and put them into PowerPoint or slides. Another thing is that for next week I plan to test my program on the motor and the H-brisges and start debugging and integrate the sensor program afterwards. The date is: Tue Feb 23 17:08:34 MST 1999 Basically during the previous week we tried to gather information from each other and got together to discuss the presentation (i.e. how we were going to present it and each person's presentation). As far as the robot goes for the previous week, I already started on the basic sensors programming with some knowledge and information that I got from Saeed. I have thought of using Altera for some parts of the programming that can take most of HC11 resources (like the PWM for example). As for the next week or two, I plan to work more on the sensors with Saeed (in the programming), get together with Najib and Sean for doing some tests on them on the weekend. We plan to have the program for left wall following. On the other hand, I already got one big cookie tin for the body of the robot. From my personal point of view, what our group needs right now is to have all the external parts (motors, sensors, and body of the robot) together and start combining them physically and in the program. The date is: Fri Feb 26 07:58:40 MST 1999 I have run a test on my HC11 program. The program for the HC11 does not work too well. I will have to do more debugging and probably also change the way the program works. Right now the program takes quite a long subroutine on the interrupt (which probably is the main problem). My solution to this one is to clean it up and to take away some parts and turn it to Altera program. We also ordered some more parts which hopefully will arrive within a week or so. My plan for this weekend and next week is to have some of the controls to Altera (e.g. the PWM and the duty cycle). This way the HC11 only has to control how much duty cycle required and send the signal to Altera and let Altera run the PWM accordingly. As for the sensors, I plan to have some GP2D12 sensor subroutines integrated to the HC11 or Altera next week while waiting for the ordered parts to arrive. The date is: Sun Mar 7 12:10:18 MST 1999 Since the Altera cannot be implemented for now (because we do not have the socket to plug it into the circuits), I continued to program the HC11 this week. With all of the group members, we tested the sensors for a few different configurations this week. Some failed, but some worked. One of the configurations that worked can even detect the candle from more than two feet away. For next week, I personally plan to finish up the program and start combining the program for the sensors and then test it out. For the spring break, we plan to finish all the parts. The date is: Fri Mar 12 13:15:01 MST 1999 This week, we got the flame sensors and the line detector circuits working. I prepared a few subroutines in the program for these sensors. They are not fully integrated yet, because I have been working on the HC11 programming for the movements. The main HC11 program for the movements that I have been working on theoretically is finished. I still have to debug it and run more tests on it. We plan to meet again on Saturday to work on the hardware and the software together (that way everybody will know what each person's doing). Our plan for next week (during the Spring Break) is to finish everything that's needed for the MidTerm evaluation. My part is to have the motor control ready, if possible with the sensors connected and integrated to the program. The date is: Thu Apr 1 16:28:12 MST 1999 Since the report for last week is not posted, I will summarize it. Up until the mid-term review, I kept on programming the HC11. My first program did not work properly, but my second version worked as I demonstrated on the mid-term review day. After the mid-term review, we were working on the fan. We got it running by last Thursday. The fan can blow up a candle up to the range of a little bit more than seven inches, which is about enough for the maze (the radius of the candle circle is six inches). Meanwhile, I started on the full program for the sensors and the robot. For the sensors, Saeed will be in charge, while I will help on integrating it to the HC11. Up until now, I have the preliminary flow chart for the wall following and the flame/white-lining combo. I plan to work on the codes this weekend and I plan to have parts of the codes ready for testing in a week. The date is: Thu Apr 8 16:04:58 MDT 1999 This week I have so far tested the movements of the robot without the sensors on. This week I also got the full algorithm for wall-following and white-line combo from Saeed, so I'm going to start writing the main code for the movements around the rooms (excluding fire-extinguisher). The algorithm in summary is: * Do a left-wall following (left has the priority). * Every time it needs to turn, it moves about 8-9 inches forward and rotate. * Every time it scans a white line, it stops to scan the room (we exclude the fire extinguisher for now). * If no fire is detected, move backwards again about 8-9 inches and rotate to the right (otherwise, logically we want it to put out the fire and go home). * As for the fourth room, we make assumption that after the three rooms without fire, the fire must be here so enter using a right-wall following, put it out and go back home using a left-wall following. Obviously, my plan for next week is to write this algorithm into a C program. The date is: Fri Apr 23 11:40:54 MDT 1999 Last weekend when we tried the program, it works pretty good on the straight move. But after a few hours of testing, it slows down a lot and it couldn't move as it should before. We figured out that it is caused by the heat in the voltage regulator. It seems that the voltage regulator takes a lot of currents and gets hot really fast. So we put a heat sink for every possible heat source. After that, we tested it again and it worked better. Also I tested on the frequency to voltage measurement, it turned out that the measurement really affects the speed of the motor like it should. As I slow down one side, the other also slows down to almost a matching speed. So in summary, the program now is working with the motors and the frequency to voltage converter. The next step would be adding the wall sensors. We currently are working on the wall following program which is under some serious testing and debugging right now. So my plan is to get some more testing and debugging and also preparing for the presentation. The date is: Fri Apr 30 11:10:22 MDT 1999 Last Sunday I found out that the frequency-to-voltage converters were not giving the correct value when we hooked them up from the motor encoders. It turned out that the frequency-to-voltage converters that we built only take values from at least >>negative<< unipolar waveforms, but the motor encoders are giving 5V amplitude >>positive<< unipolar waveforms. One fix that we tried this week was to put the motor encoder waves into Schmitt Triggers and hooked them in series with capacitors. We tested them in the oscilloscope. The waves now are >>bipolar<< square waves at about 2.5V amplitude. Another problem rises. The capacitors take some time to charge up. If the motors turn very suddenly, the circuits won't work. Luckily, I put a ramp function in the PWMs so that the HC11 gives time to speed up. This is working OK, but now the problem is that the frequency-to-voltage converters sometimes still give the incorrect values. We haven't figured out what actually caused this, but my guess is that we have some grounding problems (though the grounds are OK when I checked them). For the last two or three days, we've also been concentrating on the presentation. We have prepared almost 30 slides ranging from the circuitry, the robot's skeleton, and the programming. For the next week, we plan to continue working on the robot as much as we can while also preparing the final report. The date is: Fri May 7 07:39:27 MDT 1999 This week I continued on working on my program. At least I got the simple wall following program finished for the final review. We sort of by-passed the frquency to voltage converter for now. As I said in the review, given another two weeks we could have finished it (provided no other errors in the wiring, programs, etcetera). We got the program almost at the very last minute, so we didn't have the time to tweak the shaky moves when 2B|~2b (Bleep, our robot) is following the wall. The turn is still also wide. The things that we have implemented in that are combo of left and front sensors to detect the wall and to stop at the wall in front of it. There will be no plan after this report other than finishing up the report and go to the finals.