1/31/2007 Today, Mike and I decided to go to a 2 processor system on the Hummer. We got a good deal of the re-wiring today. We put two different Bx-24 chips onto some breadboards that we got from Jameco. One will handle all of the sensors and movement commands, and it will send commands to the other processor, which will only manage the motors. We left the "old" wiring system in tact, so we can revert to it if we need to. I intend to make a new schematic of this system as soon as we are finished. 2/1/2007 Today, I worked more on implenting the new system with two BX-24 processors. We pulled a switch from one of the hexapod kits and installed it on the hummer. I soldered some wires to the switch, and installed some power and ground wires on the breadboards which hold the 2 processors. 2/7/2007 Today, Nalaka and I worked on the Hummer projecet. Nalaka helped finish wiring up the power to the control boards. I wrote some very simple code to be sure the the Bx-24 that will be controlling the motors is wired up and working correctly. The code uses 4 of the chips inputs. Unfortunately, pin 16 appears not to work anymore, so I am not using it. I am using pins 13, 14, 15, and 17 as inputs. These are signals which will tell the motor driving bx-24 to go forward, go backward, turn left, and turn right. I manually tested the program and chip using our new power supply, and everything behaved as expected (with the exception of the discovery of dead pin #16 as already mentioned). Tomorrow, I will document the test code used as well as the schematic for this system that we have built. I also plan on getting out the eATV and working with Nalaka to thinnk about what parts we will need. If I have time, I will look into wiring the 2 processors together and having the "sensor proccesor" command the "motor processor". 2/8/2007 Today I did find time to program some commands into the sensor processor, and they were succesfully executed in the motor processor. I documented a pin-diagram, and two sets of code in a subfolder of my log folder called "hummer-dualproc". The motorproctest.zip file contains the code for the motor processor, and the sensorproctest.zip file contains the code for the sensor processor. The next step will be to attach sensors to the sensor processor and write code to drive the motors differenty depending on what happens with the sensors. I will wait for Mike to get back to do this, as he has more experience with the sensors and probably already has written this code. The code in the motor processor also needs to be updated to pulse the forward and reverse pins instead of simply turning them on and off as I am doing now (the hummer moves too fast this way). I also helped Nalaka some in looking at the eATV to brainstorm some ideas for the mechanics it will need to become autonomous. 2/14/2007 Today I worked on the ATV. I took the seat/cover off. It ended up being a pretty big job. In the process, I seem to have broke the ignition switch. In order for the ATV to operate now, the ignition switch needs to be bypassed by connecting the two wires together that go into the switch. Also, there is a button inside of the brake assembly. The brake handle needs to be in its uncompressed position in order for the hummer to move. There is a white box and a silver box on top of the frame of the hummer. The white box is mostly basically a pass through, with all the wires interconnecting through plugs. The silver box must be where the circuitry is, but I have not opened that one yet. Important Warning: The brake system is curretnly unconnected. I will not be in tomorrow, so I will not be in until next wednesday. One more thing, I bought some tools. I small crescent wrench and two alan wrench kits - one for american and one for metric. All of them are in our big black toolbox. 2/22/2007 I worked today and yesterday with the eATV. I have been thinking about a solution for steering the hummer. I drafted up a sketch drawing of one possible solution, which would have the servo motor mount directily in line with the steering shaft. I have uploaded a scanned image of my draft files in my log folder, in the eATV folder. The file is called steering.jpg 2/28/2007 Steering Research: Interesting way to do it: Used a belt right on the steering shaft http://www.its.umn.edu/research/reports/CTS-04-12/I/ch5.html#steer We will pursue the idea that I had drawn out. I sent an e-mail to a representative at PacMation, a motion control company, telling him our situation and asking for some advice on motor selection. 3/01/2007 I talked with some people at PacMation. I told them it should be able to turn from full left to full right under 15 Nm of torque in about 2 seconds. They are currently looking into what will work under our space requirements. A question they have for us is what communications method we will use to link to the motion controller... serial? USB? Ethernet? I wasn't sure. Clinton came in and we come to the consensus that serial would be the best idea. I spent the rest of my day figuring out how to electronically control the speed of the hummer. I had originally taken data points with the sensor in place now, but later found out those data points will not matter so much, because an extra resistor is needed in the circuitry once the throttle lever is removed. There are three wires that come out of the electronics box that go to the throttle. Red, Black, and Blue. They currently go to a plug with a sticker on it that says "hall". The red and black are 5V and gnd, while the blue takes an analog voltage input to set the speed you want. When I used the power supply to apply a voltage to the blue wire, nothing happened at first. Clinton came in and figured out that there needs to be current going through the red and black wires. We connected a 2.2k resistor (red red red) from the red to the black wire, and then used the power supply to drive the blue wire. This time it worked. The hummer's wheels started moving at 1.4v, and seamed to max out (based on the sound of the motor) at 4v. The good news is the blue wire does not draw any power, so any anolog device will be able to control it without having to amplify it. 3/07/2007 Today I went over the quote that PacMation sent. A screenshot of this quote is in my eATV folder. I also put in there some specs on the gear motor and servo motor that they recommended. I made my own sketch and saved it as a jpeg in the eATV folder, called motor and reducer.jpg. Overall, the combination will have a "long and skinny" look. The diameter will only be 2.5 inches, but the length will be around 9 inches. It is a good fit for the space that is available in that section of the eATV, however. I found some links to some web sites that offer solutions for mounting a bracket to the large shaft on the ATV. I have not had a chance to look into these with more detail to see if they would work for us, but here are the links. (I believe the top 3 links refer to the same manufacturer) http://www.staffordmfg.com/pages/newprod.htm ... More Specifically, http://www.staffordmfg.com/pdfs/mtng_collars.pdf http://news.thomasnet.com/fullstory/483682 Shaft supports Misumi USA http://www.misumiusa.com/Catalog.aspx?CategoryID=-1,59912,60018 found on global spec more specifically, http://www.misumiusa.com/PDFViewer.aspx?Metric=true&Page=151 or https://www.misumiusa.com/PDFViewer.aspx?Metric=false&Page=130 I have now looked at these in a little more detail. I believe they will work, and we have several options to choose from. Because the outer diameter of our mounting surface is not a standard shaft size, we made need to use a bigger size shaft clamp and machine some "filler" to go in between. At the end of the day, I sent an e-mail to PacMation asking for some advice on how to couple the gear reducer to the 7 mm shaft on the eATV. I have yet to do analysis to make sure the shaft can hold the torque which will be applied to it. One thing we really need to know: 1. Is 18 N-m torque going to be enough? I believe an additional test would be appropriate. Another thing we need to know before ordering from PacMation... Will we be using a braking system? They could help us there, too. 3/14/2007 I spoke with PacMation today. They are going to send a revised quote with the following changes: A slightly larger motor and reducer to be sure we have enough torque AND speed. A two axis controller and amplifier setup. The second amp will be for the braking system. Bob at PacMation is going to look around for a linear actuator that we can use. Meanwhile, we should get a better idea of what kind of forces we will want to be able to apply to the braking system. I printed out some shaft clamps that I can look at tomorrow. I would like to design out a good shaft-mounted motor bracket next. 3/28/2007 Today Mike, Nalaka and I did some force testing on the brake system. We estimate that a linear actuator that can provide 20 lbs or more should be sufficient. We decided on a motor mounting clamp to use and planned how to design the brackets. We plan on buying 4 clamps, each costing about $50. We will eventually buy the metal for the clamps and machine it ourselves. While Mike may be going back to working on the hummer soon, I may start working on specific Solid Models of a completed mounting and shaft-modification design. 3/29/2007 I talked with Nalaka about some over-all project ideas for the eATV. We talked about different sensors we can use and a little about interfacing. I got a finalized quote from PacMation. They send it in a form that is difficult to document, but I have it in my e-mail and a P.O. should be made shortly. I put the brochure for the linear actautor we are buying in my eATV folder: Actuator Flyer Rev X15.pdf 4/11/2007 I went ahead and docmented the final quote from PacMation. It is listed in the eATV folder as PacMation Quote 3-29-2007A.jpg and PacMation Quote 3-20-2007B.jpg. Between the two files, you can see the whole quote. In case anyone was wondering, the motor modification refers to the need to create a flat surface on the shaft of the motor so it could match the mating style of the linear actuator. Terms to remember & research - Xbow MicroNav - Penn State Mini grand challenge http://www.linuxjournal.com/article/2202 - article giving ideas for robot development http://cede.psu.edu/~avanzato/robots/contests/outdoor/index.htm - Mini darpa.... looks fun and challenging Today I found the Crossbow Sensor package that we have and gave it to Nalaka to look at. I also looked into the rules for the Penn State Mini Grand Challenge. It looks challenging as you have to stay on an asphalt path. The path often merges right into grass or loose gravel. How to tell when your between? There is also some GPS requirements involved. I would summarize the competition as path following, obstacle avoidance, and GPS following. 4/12/2007 We have decided to use the rules for the Penn State Mini Darpa challenge as a guideline for the eATV project. http://cede.psu.edu/~avanzato/robots/contests/outdoor/ I have made an initial block diagram sketch of all the components that I think the eATV will need., it is called "system sketch 4-12-07.jpg" The next step is to figure out if the BX-24's that we have are sufficient to run the sensors and motor control that we have, or if I need to look into getting a different uPC. I believe that the major algorithm and decision making will be done on the ToughBook laptop. I need to figure out how many uPCs we will have, and how they will communicate with the toughbook. USB to serial, with a USB hub? Maybe one uPC that communicates with the Tougbook serially, and others that communicate with it using I/O pins. The Penn State course mini DARPA requires that you stay on an asphalt path through a certain number of GPS waypoints. An idea I have to stay in the center of the path is to use two camearas looking out the side of the ATV. It will be looking for the line of the end of the asphalt where it transitions into something different (grass, leaves, gravel, wood, whatever). I need to find out how, through software, that line can be found and then analyzed. You could try to keep the line from both images at the same height in order to keep the ATV in the center of the path. Another sensor module it would have would be color sensors using photoresisotrs, as we once did in the robotics class. We could have a skirt around it to block out any light, and have them constantly detecting the ground to make sure it is on asphalt. There would be two of these modules, one on either side of the ATV. Maybe in the front corners. Instead of something completely home-made from scratch, maybe there are commercially available "color dectectors" we could look at first. 4/18/2007 Today I worked on the ATV mechanically. I took the brake and throttle off of the handlebars, so the handlebars can now be completely detached form the atv. I also looked closely at the forward/reverse switch and the kill switch. The fowrad/reverse is a Single Pole Single Throw switch. We can buy a relay switch that a TTL output can control for that. I am fairly certain that the kill switch will be the same way. We also need a mecahnical kill switch in series for competition requirements. Unfortunately, in doing some testing it is very likely I blew a fuse in our MultiMeter. I will need to be on the lookout for a new one. In the process of debegging that, I found out that another one of our yellow multimeteres doesn't work. I tossed it. I worked a little more on the system diagram for the eATV. 4/19/2007 I found that none the microprocessors we have or the Galil Moition Controller don't provide an acceptable analog output to drive the speed wire with, so we will need to get a DAC. We should be able to use some general purpose output pins on the Galil to use as inputs to the DAC. I also planned on using 2 of these outputs from the Galil to operate 2 SPST relay switches. I started looking into the commercial color sensors, for example: http://mysick.com/partnerPortal/eCat.aspx?go=FinderSearch&Cat=Row&At=Fa&Cult=English&FamilyID=239&List=1&Category=Produktfinder&Selections=14493 Model Name CS81-P3612 Part Number 1028225 I finally found a price listed for the part above and it was $700. That is way too much, considering we can make our own that may work all right. Here is another one to look at http://www.emxinc.com/colormax.html http://www.thomasadams.org/robots/experiment.htm - info on homemade color sesnors 5/21/2007 Today I came in to finalize one last order, and get it turned in. I made a "to do" list for the eATV project that lists several different things that can be done, including some things that should be ordered. I put the "Summer to do.txt" file in my eATV folder. I bought some replacement fuses for the one I burnt out buy I forgot to bring them in. Someday... 12/22/2007 I deleted that to do list, because I did it all. I replaced that fuse that I burened on the large plug-in MM, but I think I burnt another one in the small yellow one. The status and documentation for the eATV, as I left it, can be found in the eATV folder in Autonomous ATV report.doc