Rebuild Diary

Let me start with the part I didn't have to rebuild: the frame and body. The frame was built from 3/4 inch aluminum square tubing from the Home Depot. The frame pieces were cut on my bandsaw and then "welded" together with an aluminum welding rod from Northern Tool and an oxy-acetylene torch. The body consisted of fender skirts made from 1/8 inch aluminum sheet, which I had lying around.I don't have any decent pictures of the original bot, so we'll have to pick up where I finally dragged out the camera. The first thing to do was gut the bot, removing the original gas system and relays. The chassis was still intact and had just enough room for the planned enhancements. The original foam tires, which had been severely torn and squashed were removed. The only thing remaining at this point was the Aluminum frame and the CO2 tank. I further stripped it down to just the base frame, to which the motors mounted. Each drive motor is the motor and planetary gearbox from a Black & Decker Versapak screwdriver.

The motors and gearboxes were torn down, cleaned, inspected, lubricated, and reassembled. The original output shafts were replaced with 1/4 inch drill rod. The wheels were replaced with casters from Northern Tool .

The wheels were removed from the casters and the bearings removed. I had to create hubs to mount the wheels, with a 3/4 inch I.D, to the 1/4 inch axles. I couldn't find any 3/4 inch solid rod, so I took 3/4 inch steel tube and put in 5/8 inch steel rod to create the hubs.

Since I don't have a lathe with which to center drill the hubs, I had to make a jig to hold them straight on the drill press. I took some wood, cut it in half, put spacers between the halves, and drilled a 3/4 inch hole across the cut. After removing the spacers, the hole is just slightly under 3/4 inch to grip the hubs tightly. Since the hole in the wood was drilled on the drill press, the hole itself is straight in relation to the spindle. It worked fairly well. There was a bit of inaccuracy, but overall tolerance was satisfactory.

The wheel pieces were drilled to accept 1/4x20 screws to hold the tires on. The hubs, wheels and axles were then brazed together. The axles were then pressed into the gearboxes.

Next, I tackled the electrical system. The original relays had been thrown out. I opted to replace them with 12V 30A automotive relays from Radio Shack.com. At this point I still didn't have a radio control rig, so tethered full-on / full-off was my plan. I know that the 30 Amp relays are still underrated for the stall current of over 75 Amps per side, but at about 35 pounds total weight, there was just no chance of ever stalling the motors. I opted for 8 of the relays in a dual H-bridge configuration. These were mounted to the bottom of the main frame, between the front and center motors. I also installed a terminal strip to simplify some of the wiring.

The next step was the wiring. On the original bot, I had used 16 and 18 gauge wiring for the high power section. Big mistake. I actually melted the insulation off of a significant portion of it. This time, I decided to go heavy. I was shopping for wire at Radio Shack.com and was pleasantly surprised to find 10 gauge speaker wire pair for $.59 per foot. Yeah, I know, speaker wire, but at this size I didn't see any problem. Besides, it was extremely flexible, which was a boon in the tight space I had to work with.

Several of the wiring connections at the terminal blocks had to be jumpered. I didn't really have space for heavy wire jumpers, so I pulled out some steel and a hacksaw and made some.

The next step was to add the low power wiring to drive the relays. I used some 18 gauge wire for this as it didn't require any real current capacity and I had some already. Next, I mounted the motr / gearbox / wheel assemblies to the frame. The mounting was accomplished with 6x32 screws run through the gearboxes and into the frame. The outside end of the shaft was mounted to the bottom of the frame with small teflon bushings from Lowe's secured with small conduit straps. Since it was on the bottom of the frame, and I didn't expect any serious horizontal shock, this was an adequate mounting scheme. Were this for a larger arena, where some serios shock could be expected, I would have had to redesign the mounts. I went ahead and mounted the rest of the frame as well as the CO2 bottle. I installed the batteries and it was ready for a test drive. That big red key is a 12V 300A disconnect switch from Northern Tool. The batteries are 12V 4.5 Ah gel cells that I picked up for free. I have about 12 of these lying around.

*Note* Due to time constraints before competition, the following pictures were taken after the competition.

Now for the fun part, the weapon. The original weapon was a camp hatchet that I had rigged with a screen door closer piston. As previously discussed, the original design left a bit to be desired. However, it did have potential. I opted to stay with the same basic design, but reengineered it to greatly increase the diameter of the gas passageways and to add a reservoir of regulated gas. I searched around for a suitable cylinder, but was unable to find one with at least a 6 inch stroke, 1 1/2 inch bore, and a spring return. I needed the spring return to simplify the home built valve system. I settled on using the old faithful screen door closer piston for the main driving piston. In the arrangement I designed, this was to be the slave cylinder. I cut open the piston to reinforce the mounting on the end and to weld my valves directly to the piston. This arrangement would reduce the distance from the valves to the piston, thereby reducing efficiency robbing pressure loss and turbulence in the system. The valves were two 1/4 inch bore gas ball valves. I removed the handles and mounted the two valves with the valve stems pointing towards each other.

I affixed the two valve stems into a short piece of 5/8 rod to connect them together. Into the joiner, I threaded a handle which would operate both valves simultaneously. When the handle was moved 90° towards the front of the piston, the input valve wuld close, and the exhaust valve would open. When the handle was pushed 90° to the rear of the piston, the input valve would open and the exhaust valve would close. Next, I created a buffer tank from a 2 1/2 inch by 8 inch pipe nipple and two caps. I drilled two holes in one cap for inlet and exhaust. I brazed 3/8 inch nipples into the holes and threaded it into the input valve. I had to come up with a master cylinder to drive the valves for the slave cylinder. I found a gas spring, the kind to hold open car hoods, at an auto parts store. I took home the gas spring and proceeded to modify it. I cut the end off of the cylinder, slowly to let the gas charge escape, and pulled out the piston. I found that the plate on the piston had holes which allowed the gas to flow through from on side of the plate to the other. I filled the holes with epoxy and let it sit overnight. Then, I drilled a hole in the side of the piston, at the bottom, to insert a pipe fitting. I reinstalled the piston rod and plate, and brazed the end piece back on as well as a male quick disconnect fitting from Georgia Paintball. I brazed a long carriage bolt to the end of the buffer tank to mount the master cylinder. I mounted the piston to the bolt, and pinned the rod / clevis end to the handle of the valve joiner. I added a spring return and was ready to test. I tested on shop air from 60PSI to about 125 PSI. The higher the air pressure, the better the performance.

I never saw any signs of potential "explosive deconstruction", so I was ready to proceed. I mounted the regulator, an old soda fountain CO2 job, to the frame with zip ties.

I connected the output of the CO2 tank to the regulator with a hose, a quick disconnect, and a "gas check" valve from Georgia Paintball. Coming from the outlet of the regulator, I put a tee with lines running to the master cylinder and to the buffer tank. The hoses from the paintball shop were a bit expensive, so I hunted around for a cheaper option. What I found was grease gun extension hoses at Northern Tool. These hoses are 12 and 18 inches long, rated at 3000 PSI, and under $5 a piece. The inside diameter of the hoses was pretty small, but large enough for my purposes. The entire piston / buffer tank assembly was mounted to the frame with one large bolt at the back end as a pivot.

The axe now needed a bit of work. The axe was an old metal handle camp hatchet to which I had bolted and brazed a clevis made of steel flat bar stock. The axe was OK as it was, but not quite good enough. The wide head tended to disperse the force over too great an area. I pulled out the cutting torch and whittled the edge down to a fine point. I then took it to the grinder and sharpened it.

I installed the whole assembly in the bot with a bolt at the rear pivot and a bolt at the axe pivot. The gas storage tank was a 9 0z. paintball tank and the regulator was an old soda fountain first stage regulator set to 150 PSI. This The hose coming from the tank was connected to a gas-check valve and quick disconnect fitting. The valve would relieve pressure on the hose and seal it off so that when the quick disconnect was disconnected, the contents of the tank did not come spewing out. This ws for safety and to allow filling of the tank. The tank had to be filled in place as the design did not allow its removal without full disassembly. After the quick disconnect was the regulator. This was followed by a tee which fed the buffer tank and the solenoid valve. The solenoid valve fed the small master cylinder.

The next task was to install a control system. The radio control was handled by a Futaba 9VAP PCM that I bought at a swap meet. It's a sweet rig, but a bit old. No one I've ever talked to has ever heard of that model. Since it's a computer radio, I was able to do elevon mixing at the radio, eliminating the need for a mixer on the bot and allowing single stick navigational control. I used a Parallax Basic Stamp II to handle conversion of the radio receiver signal to PWM for the motors, as well as fire control.

Not exactly a speedy processor, but it seemed to work. I homebuilt an H-Bridge from 75 Amp MOSFETs. This is trickier than one would think. Eventually my testing released all the magic smoke from (blew up) one more FET than I had to spare. Due to time restrictions, I had to go to full-on/full-off control with relays. I used the remaining FETs to drive the 12V 30A automotive relays and reprogrammed the stamp to handle single speed control. I added a 5V voltage regulator so that I could run the control system off of the 12V main power and not worry about charging a receiver battery. I housed the whole thing in a plastic electrical switch box and mounted it to the bot. I also added a serial cable and connector for reprogramming and calibration of the stamp.

The last task was to reinstall the armor and build a cover. The original armor consisted of four aluminum skirts which were attached to piano hinges riveted to the robot. The idea was that lifting the skirts would not lift the bot off the ground, making me immune to lifter bots. The problem with this design was that due to the steep angle of the skirts and the fact that they were unsupported was that when Lizzie was hit, it pushed the skirts in, lifting me off the ground. This time I used the same skirts, but screwed the hinges to the frame to allow easier disassembly. I also supported the skirts from underneath with 8 long carriage bolts screwed into the frame. These were adjusted for length to allow the skirts to ride on the floor, but not to be pushed in any further. The next step was to construct a top. This was the day before the start of the convention and I had not even come up with an idea. It needed to be strong, spacious enough to house the weapon assembly, and not interfere with the radio signal. As I looked around the garage for an idea, I stumbled across (literally) an empty 5 gallon bucket. I pulled out the sawzall and cut it in half lengthwise. I added a mounting plate and hinge to the bottom, and 2 angle brackets to latch it. I duct taped the antenna into the lid and mounted it. Perfect fit! It did leave a bit of the front unprotected, but I was counting on no one getting that close with the axe. The entire bot was taken out to the driveway and paintd blue, mainly to make the white bucket match the aluminum body.

Testing of the axe yielded some impressive results. It was able to pierce the cover and 200 pages of a hard back book. It was also ablt to create a .4 inch long hole in 12 ga. sheet steel.