Last week I gathered information from the web, scratched my head, and tried to figure out how to build the basics of a robot. Several decisions were made, as a group, such as electing to put the PWM function into Altera chips, in order to shift some of the load away from the HC11. We decided on near IR sensors, electing to use the Sharp sensors provided. The date is: Thu Feb 4 14:02:36 MST 1999 This week I continued gathering information, colaborated on a decision to move the Pulse Width Modulation function into Altera chips (in order to reduce the load on the HC11), and began a collaborative design effort with Jack Landes to build the Altera PWM function. In addition, several group functionality issues were settled. The date is: Thu Feb 11 18:40:58 MST 1999 This week, Jack Landes and I completed the implementation of a motor control system within Altera. Unfortunately, the design requires 178 logic cells. Since the design only requires 21 pins (input and output both), we are faced with the problem of finding an Altera chip into which our design will fit (since cramming with a shoehorn won't work). I designed a system to implement a Pulse Width Modulated signal. The system consists of a 7.8kHz input signal, as well as a desired speed. The 7.8 kHz signal runs a counter, which counts up to the desired speed (which, after all, is only a number between 0 and 256), placing a logic 1 on the output line until the counter is equal to the desired speed. This has the effect of generating a PWM signal that varies between 0 and 5 Volts (an op-amp can easily push that range to 0 to 12 Volts) at 600 hHz. I also designed a controller of sorts, which, through a simple series of IF-THEN statements, compares actual motor speed to desired speed. If the speed is two low, the controller pushes the speed up a bit, and vice versa if it's too slow. Since the speed is being checked at 600 kHz, the system has plenty of time to operate. Jack and I put his encoder decoder together with the above systems, and make it compile within Altera. The problem of too few logic cells is one we are still considering. The date is: Fri Feb 19 13:53:35 MST 1999 I spent the entire week preparing for the midterm presentations. Designing slides and co-ordinating group activity took up most of my time. I also worked with Jack Landes on fitting the PWM generation, Encoder decoding, and rudimentary motor control into two Altera chips: an EPM7128ALC84-12, and a EPM7064-44. The date is: Thu Feb 25 15:41:37 MST 1999 This week Jack Landes and I ordered parts for the motor control system. Specifically, we ordered the: EPM7128ALC84-12 The date is: Thu Feb 25 15:46:14 MST 1999 This week Jack Landes and I ordered parts for the motor control system. Specifically, we ordered the: EPM7128ALC84-12 EPM7064LC44-12 and an 84 pin PLCC socket - the 44 pin socket we are going to cannibalize from Mike S.'s dead HC11EVBU board from www.wyle.com In addition, we attached a caster wheel (donated by the Instrument Room) to the body of the robot, allowing our robot to stand in place without outside support. The entire group elected not to linearize the sensor data, as it was good enough for our purposes in raw form. The date is: Fri Mar 5 16:04:42 MST 1999 This week I: - obtained a small water pump from an auto junkyard. $5. - obtained a small length of hose from ACE Hardware. - obtained an adapter to fit the hose to the water pump. - obtained a small piece of perf-board to mount the Altera chips (for motor control) on. - assassinated the ruler of a small, third world country to obtain misc parts (just kidding). - implemented various clandestine operations. The date is: Thu Mar 11 18:29:51 MST 1999 This week, Jack Landes and I obtained the parts for, and put together, the fire suppression subsystem. This system is virtually complete - all that remains is to mount it on the robot, and implement the interface to the HC11. The fire suppression subsystem consists of a windshield washer pump from a broken down Dodge Colt (a car), two lengths of hosing, a sprinker head, a makeshift water tank. The water tank was improvised using two plastic salt shakers, and (of course) epoxy. A small adapter was used to attach the hosing to the water tank ($0.75 at ACE Hardware). A similar adapter was used to mate the hosing to the windshield washer pump (the hose was too small). From there, the hosing goes (by means of a third adapter) to the sprinkler head. The sprinker head, also obtained from ACE Hardware for $0.89, sprays the water in a wide, flat arc (the better to whack the candle with). The entire fire suppression system came to a total cost of less than $20, all of which was donated by Jack Landes and myself. The pump draws far too much power (12 V and 1 Amp) to leave attached to the robot's primary battery. This problem has been solved by obtaining 2 secondary batteries ($4.86 each at good old ACE Hardware, made for answering machines, and included in the cost figures above). The batteries are capable of supplying 550 mA hours, and 500 mA hours, at 6V, each, and thus solve the problem. Both Altera chips (EPM7128ALC84-12 and EPM7032LC44-12) came in last week, are now programmed, and ready to place in their sockets when the sockets are wirewrapped (there are far too many connections in far too small a space to etch a board without extreme quality etching equipment). Jack Landes will wire-wrap the sockets over spring break (having earlier agreed to do so). I will use some time during spring break to prepare transparency slides for our formal midterm presentation. The date is: Thu Apr 1 11:22:24 MST 1999 This week, Jack and I found and purchased all-thread and masonite for the robot chassis... We also tested the motor control system and found the H-Bridge Boards to be defective. We are currently rebuilding and re-testing the motor control system. The date is: Fri Apr 9 14:44:46 MDT 1999 This week, all of Group 7 redesigned motor control 4 times. These multiple re-designs were neccessary because of a flaw in the conceptual view of motor control (number of pulses from the encoder was mapped to an 8 bit number based on revolutions, instead of PWM cycles.) This would have resulted in total and complete failure if not corrected. These multiple re-design took almost all of our week, but a few other things were done. I, for example, helped Mike begin writing the control code for the HC11. The sensor algorithms, as well as basic right wall following algorithms, are now in place. The code is not complete, though we hope to spend next week completing the code, testing the motor control subsystem, and starting to debug the code.... The date is: Fri Apr 23 12:49:22 MDT 1999 This week we had to completely re-program the Altera chips we used for motor control. This was made necessary by several programming errors, and the fact that once the errors were corrected, the programs had to be re-arranged to fit. In addition, we had fried (at some point in the past) an EPM7064LC44-12 chip, and not noticed. This chip was replaced through the generosity of another group in Junior Design... The date is: Fri Apr 30 16:11:09 MDT 1999 This week I helped finish the sensor systems. In addition to this, I also helped re-design motor one more time!! I programmed the PWM program into an EPM7032LC44-15T and tested it - it worked perfectly, and was capable of varying duty cycles from 0 to 100%. That question answered, the motor control was moved outside of Altera, and the EPM7128ALC84-12 was used to hold Encoder Signal Processing (the latest design) and the PWM program. It fit with a minimum of bitching and whining. The H-Bridges were tested and proven to work. The date is: Thu May 6 15:11:05 MDT 1999 This week in Junior Design we: - worked on the report - worked on the report - worked on the report - mounted most things on the robot I was assigned the PWM and Fire Suppression Subsytem parts of the formal report. I wrote the Fire Suppression (it has yet to include diagrams) section. The PWM section is mostly done (somewhat similiar to mostly dead) because I needed it done for another class (English 341). Except for some minor editing for context, and making the separate sections of the report fit together as one, the report process is well on its way.