Red Menace - the 2nd Battlebot (Robots)

We learned a lot of lessons from our first attempt at building a BattleBot, and high on the list was, “Robert needs to know how to build hardware as well as software”. Thus it came to be that I spent the summer helping Team TSR Robotics rebuild one of their robots, picking up a basic education in machining along the way.

However, this caused a problem; I didn't have time to work on a new robot of my own for Battlebots in November, and with my kids threatening “Daddy, either you build us a Battlebot, or we'll get a new daddy who will!”, it was with a sense of relief that I noticed an eBay auction for a lawnmower-style spinner. I was fortunate enough to win the auction, and thus it was that at the beginning of September, a new robot entered the family. We call it…

The Red Menace

Formerly owned by a very nice guy who simply didn't have the money to compete any more, this was more an adoption than a purchase. The plan is simple: starting with the original robot, we'll make changes and improvements in an incremental way, and at all times, we'll have a working, ready-to-compete robot. With the help of the TSR guys, I could probably have whipped together a robot in time, but this way, there's no risk of missing the tournament.

The first week was consumed with unpacking the robot, checking it out, and getting it running with the IFI robot controller left over from Evil Twins.

The weapons motors on Red Menace are controlled by a solenoid switch, which itself is controlled by a Team Delta RC switch. At first, it wouldn't work with the IFI robot controller, but eventually it dawned upon me that the Team Delta switch requires power supplied through its RC connector, and the IFI doesn't provide this by default. While there is a connector on the IFI you can use to provide +5V for distribution to any connected RC devices, I found an easier way; the robot came with a 5V DC/DC converter with a RC connector (originally used to power the radio), and so I just plugged it into one of the unused PWM connectors on the IFI. Since it only connects to the +5V and GND lines, it's electrically the same.

I also ran into the same RF coupling interference problem we had with Evil Twins. Wrapping everything but the antenna of the IFI in aluminium foil provided a quick fix; I'll probably build an aluminium box for the final incarnation.

Since I hope the kids will be able to drive this robot in combat (assuming I get through the prelims; they'll only be attending for the TV rounds), it is essential that they get in lots of driving practice. So I yanked the weapon blade and added a caster so they could mess around with it on the driveway.

I'm also working on better software for the IFI to help them out. For example, since I'm using a CH FlightStick, which has a twist input, I'm using that as a “fine turn” controller. I also hope to interface a heading-lock gyro.

Other planned improvements: upgrading to BattlePacks and souping up the wiring; changing the weapon drive to vbelts (which can slip for shock absorption) and the weapon to a solid bar (currently, it's a flail); and perhaps even rebuilding the frame.

09/12/2002 - Gyroscopic Stability

In an attempt to improve the directional stability of the robot, I purchased a helicopter heading lock gyro. It basically changes the left/right commanding of the robot from “turn left/right” to “turn left/right and maintain heading”.

This means that even if a wheel slips, the robot will maintain its heading when being driven. It also makes the turns crisper. A complicating factor is that now the command path is IFI → Gyro → Mixer → Motors; since the gyro only deals with a left/right signal, I have to mix that with the forward/reverse signal outside of the IFI. I'm currently using a mixer, which is “smart” but it has dead zones that are a bit too big, so I'm thinking of playing with dumber mixers. Even so, with the gyro on, the bot is a lot more controllable.

The parts for the v-belt drive have arrived, and I'm going to be installing them over the weekend. I've also come up with a cute slip-clutch design for the blade; attached to the weapon drive shaft will be a metal plate (basically a big washer), a disc of UHMW plastic, the blade, another UHMW disc, and a bottom plate. A bolt going through the bottom plate into the shaft (up the middle) will compress the sandwich and couple the blade to the shaft by friction, yet give in a major impact.

No idea if it'll work, of course.

09/21/2002 - More gyro work

Played around with getting the gyro to talk directly to the IFI, or, more precisely, to talk to a basic stamp that then talks to the IFI; initial results promising, and have been written up here.

12/04/02 - It's Robot Fightin' Time!

Well, a lot of work in the last few months, but all scattered about. Got all the parts I needed, and installed the new batteries and wiring. Machined the weapons blade and mounts; it now uses two plates and two pins, with a small bolt and washer that locks the assembly together. On the head of the bolt a delrin furniture slide is expoxied to make a bearing surface that can slip around the floor of the arena. Also changed to a v-belt drive, as this acts as a slip clutch; this was a major pain in the butt as it required boring out pulleys to metric dimensions.

Here are some shots of the robot and its first opponent, “Sacrificial Lamb.” Needless to say, the lamb got slaughtered! We estimate the blade is wizzing around at well over 1000 rpm, and we can probably get it going faster if we do a little tuning of the blade shape. The only problem we ran into was that the blade rotation direction is such that if the bolt isn't locktite'd, it eventually loosens.

This weekend we head off to the North Carolina Robot Street Fight to test it out. No sacrificial lambs there, alas, as you can see by reading our report on how well (ahem!) we did.

12/22/02 - Revisions and Revisions

You can learn a lot from getting defeated, and we certainly learned a lot at the Street Fight. So as soon as we got back to Wilmington, we got hard at work on revisions to Red Menace.

The electrical system was ripped out and the control electronics now have their own dedicated 12V/5V power supply (non-rechargable lithium-ion batteries to save weight). Internal components were shock-mounted and the radio was relocated. The battery mounts were reconfigured to save weight and absorb some impact, using styrofoam offsets. The weapon blade was bevelled using a CAD/CAM mill; each long edge of the 4041 steel blade took 6 passes with the cutter (and about an hour) to complete. A new weapon shaft (and a spare) was made, and ball casters were added to top and bottom to make the robot fully invertible. The shaft is actually a little longer than it should be, but we didn't have time to shorten it.

All in all, we are ready for Robocide in a couple of weeks!

01/12/02 - Robo-Suicide

Back from Robocide, where again I went 0-2. Had a lot of fun, but every time the robot took a hard hit, it would die. After much debugging, I found out that when the radio battery (the new one I added!) got hit just right, it went intermittent. Both fights were lost due to that. I also had some steering problems.

Because of the above, I've decided to KISS the problem, and change from the IFI to a simple R/C control system. That ought to reduce the complexity and increase the reliability.

I also suffered a couple more bent weapon shafts. To deal with this, I'm going to try and get tougher shaft material, and I'm going to mount the bottom bearing on the outside of the robot so that it, and it's shaft collar, can nestle right up against the blade mounting plate for extra reinforcement (that's where it was bending). I'm also changing from the pins to basic bolts to hold on the blade, as the current system is too fragile; the ball casters get broken too easily. Back to carriage bolts for me!

04/04/02 - Retired

Went 0-2 again at Battle at the Beach, and have decided to retire the robot. The big problem was the weapon shaft, which no matter what I did would bend. In one fight I also bent the weapon blade and had to use a fork lift to unbend it. In the final fight the lexan cover got shattered and the batteries got rearranged, but apart from minor tire damage, the internal components came through OK, so I'll be able to use them again.

My other robot, Anorexia, did better.