TiltPanPlatform
From TheRobotGroupWiki
Contents |
Tilt and Pan Platform
GlennCurrie and EricLundquist of the Robot Group are working with Austin Robot Technology to provide a large tilt/pan platform. This platform will be mounted on top of the DARPA Grand Challenge vehicle and will carry one pair of high resolution video cameras and a laser scanner.
Frame
The entire frame is constructed from one inch square steel. This makes the frame very rigid, very strong and very durable. Here you can see the base, pan (with drive gear), and the tilt section all laid out.
Pan
The ability to pan the cameras is required in order to be able to navigate around tight or hairpin turns. The pan mechanism is built from a go-kart hub that came off of a robot drive train owned by GlennCurrie. This provides very sturdy and very smooth ball bearing motion. The chain will be driven from a Everest and Jennings wheelchair motor also scavanged from the robot drive train.
Tilt
The tilt mechanism will allow flexibility is choosing the optimum viewing angle with respect to the ground. Tilt is driven by a heavy duty linear actuator Electrak that was provided by BillCraig.
History
July 9, 2005. Initially, we tried using the Motor Mind C controller from Solutions Cubed. The nice thing about the controller is that it supports both rs-232 commands and radio/control servo style commands. This is useful for testing. Of course, because we already had some in stock was another big incentive. Unfortunately, the main problem is that the linear actuators use 4 Amps and the Motor Mind C is only good for 2 Amps per channel. Even with no load, the controller would quickly go into thermal shutdown mode.
Closed Flat Position. These are a couple of photos of the tilt portion of the tilt/pan unit. The scissor mechanism was created from standard hinges from the hardware store. The pin used on the hinge was ground off at one end and then hammered out. The pin was replaced with a 1/4 hex head bolt that was long enough to go through the hinge and the push rod on the linear actuator.
Open or Max Tilt Position. Note that we will still need some sort of pivot hinge on the rear end of the actuator so that it can swing up/down.
July 30, 2005. Our friends at Tecel have agreed to provide high current H-Bridges in the form of their D200 motor controller. In addition, they are providing a P500 8052 based microcontroller to provide the command interface, for position data acquisition and to drive the motor controllers. We had to go with the P500 because we needed the built-in A/D conversion capability, as well as, the RS-232 communication capability.
July 31, 2005. We got the main structure for the pan motor mount done today. The tilt mechanism still needs some work.
August 3, 2005 Worked on getting a smaller footprint prototype to work. Here it is in the "closed" position.
Prototype tilt mechanism on a test stand in the open (raised) position.
August 13, 2005 The hinge idea just wasn't working out for the tilt mechanism. Version 2 of the tilt mechanism was built with "Heavy Duty" strapping located in the section where they have various metal joiners for 2x4 construction like roofing joists, etc. These metal bars are rather thick and sturdy with pre-drilled holes to be used to attach 2 2x4's together. Metal stock is grossly overpriced at Lowes/Home Depot (over $3 per foot!) but since these were joinst materials, they were $1.43 for each 1 foot section. A little work with the drill and as you can see, we now have a scissor mechanism. With 2 sections, it is about 5 inches in height.
Here you can see the scissor mechanism in the extended position. With only 2 inches of throw on the linear actuator, the mechanism has extended it to approximately 8 inches of throw!
August 15, 2005 Just received the new P500 8052 based microcontroller and D200 
high current H-bridge motor controllers from our friends at TECEL 
August 16, 2005 Our second site visit is tomorrow. Obviously, we won't have the tilt/pan platform ready but we are committed to continue working on it. This evening, we soldered a 7805 +5V voltage regulator onto the P500 controller as per the instructions. I put a +6V battery pack on the input and measured around 4.3 to 4.6V. The data sheet says that we need a minimum of 7V and a max of 14V to get a stable +5 out. I have several small +12v lead acid batteries that I will rig a switch and cabling harness for tomorrow. I will also make a trip to Frys or Alltex to get a female DB9 serial cable that I can hack up to connect it to the 3 pin .1" space header that the controller uses for the rs232 connector.
Started looking at the code requirements for the project. Tecel provides an unsupported version of Small-C for the board. It is unfortunate that the Small-C compiler is not being actively developed. I have fond memories from the mid 80's when I ported the Small-C compiler to a Centurion Computer 7000 by bootstrapping it via a Xenix system! It was a very cool metaphysical experience to eventually get the compiler to compile itself on the target machine from source! Those were the days! However, I understand how much work is involved in continuing to develop and maintain a compiler. Perhaps they will eventually release the source code to their port into the public domain or perhaps under the GPL. On the plus side, there is included sample code for both rs-232 communications and a/d conversion. It looks pretty straightforward so we may be able to use the code with very little effort. If we run into a brick wall, we will probably go to the SDCC Small Device C Compiler and see if we can make it work with the P500. This may be a good project anyway just on general principles. Looks like it is a reasonably complete almost-ANSI C compiler. The cool thing is that they support several different processors such as PIC and Z80 but their "native" target is the 8051!
August 17, 2005 Went by Alltex Electronics and picked up some cable making supplies. Got the RS-232 cable built. Also rebuilt the power harness to use a +12V battery and added an on/off switch. Got it all plugged up and after a little bit of work with the Troubleshooting section, I was able to compile and download a serial i/o test program.
I heard that the site visit went pretty well, so we are still in the running!
