How to Build a Combat Robot
by Robert Woodhead (Metal Munching Maniacs Producer)
So you've thought about building a robot, but think it's too difficult? I've got news for you.
Anyone with some basic mechanical skills can build a competitive combat robot, and if you don't
have those skills, picking them up is much easier than you think.
Heck, I'm a computer geek. I'd never built anything in my life before I got into this sport,
and if my kids hadn't come to me and told me, "Daddy, either you build us a robot, or we're
getting a new Daddy who will!", I still wouldn't know the difference between a tap and a die.
But they did, and I learned, and after a year or so (and several failures), I'd built myself
a reasonably competitive featherweight wedge named Anorexia (because it was so skinny). You can see it on the
Robot Club & Grill MAYHEM DVD. Then Anorexia got
bounced off the ceiling by a spinner, and the rebuilding process mutated it into a lifter, which
was also fairly successful, and would be more so if I was a decent driver.
(By the way, you can click on any of the small images to see a larger version)
In any case, to make a long story short, this year my kids came to me and said "Daddy, we want
a bot of our very own, and it's got to have a mean weapon." So it was back to the drawing board.
My design constraints were that I wanted to make something that would be easy to construct,
would reuse a lot of parts I already had (like batteries, speed controllers, etc), and would
use standard parts for the tricky bits to save time and make the build easier (my time was limited
and I only had a few months to be ready for the
Southwest Divisional Championships).
This is one of the best things about the evolution of our sport; there are now a ton of
well-tested standard parts that make it much easier for novice builders to build decent robots.
I mean, heck, I've got some decent basic machining skills, but there's no way I could build a
huge gearbox for a spinning weapon. Fortunately, I don't have to, and neither do you.
What I came up with was a design that's "totally offensive", and that's what we ended up
calling it. Everything is sacrificed
in favor of the weapons system, a Magmotor-powered
lawnmower blade. I tried to pare off every
ounce I could so that the blade could be as hefty as possible, and it ended up being a 6.9lb 22"
hunk of tool steel.
By using a lot of standard parts, I figured that if the design wasn't successful, I'd have lots
of bits left over to reconfigure into a different layout. As it happened, this turned out to be
a wise decision. This article is about version 4 of the robot.
The first iteration was a simple squat "T"
of aluminum C-channel, aggressively hole-sawed to reduce the weight. The drive wheels were set so
that they were tangent to the arc of the weapons blade.
This meant that the bot could easily turn left and right without moving the blade at all, at the
cost of having to scrub like nuts to drive forward or backwards. A pushybot this wasn't, but
given the lack of armor on the bot, being able to always put the "nasty" part of the bot towards
the enemy seemed to me to be a big plus.
Two weeks before its first tournament, however, I started having some serious second thoughts.
The problem was that because the hub of the gearbox scraped on the ground, the bot wasn't very manuverable.
The askew orientation of the wheels let it turn on a dime, but it often had trouble moving
forwards and backwards.
So I went back to the drawing board and came up
with a tilted spinner design, somewhat reminiscient of the lightweight 2EZ. Worst case,
I figured that I'd get some extra practice welding.
At the tournament, we found out that the robot should have been named Totally Suicidal,
because
every time it hit something, it would flip upside down and lie there helplessly. After one memorable hit
it danced its way on its side around the arena.
So it was back to the drawing board -- again!
This time I came up with a simplified version of
the original undercutter design. The drive wheels were exactly in line with the spinner hub,
so I could rotate around that axis while still having full forward and backward drive without
scrubbing the wheels. The frame was made out of aluminum T-angle with some minor lightening holes,
and a UHMW plastic bonnet protected the chewy parts.
At its debut tournament it went 5-0 with 4 KO's and not only came in first, but also won a coveted
invitation to the 2004 Nationals. Wow! We've got a winning design, right? Nope. At its second tournament,
we found out the fatal flaw. In almost every fight, it would deliver massive hits, pop in the air,
and slam down again... and because the blade and gearbox were suspended between the drive motors, that
long strut of aluminum would get bent, and the wheel drive motors would break from the shock.
I spent the
whole tournament doing "precision" engineering with a sledgehammer on the frame and replacing motors, and
in our final fight, we lost both drive motors and the frame got so bent the blade was touching the
floor of the box. This turned out to be a good thing because my son James, the driver, found that by
turning the weapon on and off he could move and turn the robot! We won that fight and came in 3rd.
So we're back to the drawing board again, trying to come up with a frame that is sturdy enough
to take the damage yet light enough to permit us to keep that nasty 6.9lb blade. And along the way,
I hope to show you how easy it is to build a robot.
Before we begin, a quick word about safety. It's not
enough to be safe when you're fighting your robot, you've got to be safe when you're building it.
And that means using protective gear like safety glasses. Wear safety glasses ALL THE TIME, even
when you're not using power tools. You'd be amazed how easy it is to get something in your eyes when
doing stuff as simple as hammering a nail. Don't believe me? Do a Google image search for
"nail in eye". Eyeeew... If that won't convince you to be safe, nothing will. Wear
proper clothing, and foot and hand protection where appropriate. If you don't know how to use a
tool properly, don't use it until you get some instruction on how to do it right.
Do yourself a favor, hang your goggles on your workshop door and don't even go in there
without putting them on. Shop safety is something we at Metal Munching Maniacs really care about, because
if you're blind, you can't watch our DVDs, and if you accidentally chop off your hands, you can't
easily use the remote. The jobs of our employees depend on you not hurting yourself. Needless to say,
building and fighting robots can be dangerous, particularly if you are a careless idiot, so do it at
your own risk, and if you accidentally drill a hole through your thumb, it's your fault, not ours!
Next, I want to introduce you to your new best friend, calipers. If you don't have a set,
pop on over to Harbor
Freight and pick up a basic 6" caliper. Buying inexpensive tools from Harbor Freight is sort of a
rite of passage in the sport, and by the end of this article you'll probably have several items you'll
want to buy there.
Calipers let you measure stuff very precisely and scribe
marks on your work. This lets you do a lot more with hand tools than you'd expect. While I've got a
small Sherline mill and lathe in my workshop, I
decided to go all macho and build the new frame entirely with hand tools. Such are the sacrifices
made by someone who wants to suck you into the sport and get you to buy all our DVDs.
When designing a robot, some builders go all crazy with CAD (Computer Aided Design). I
prefer to use CAD (Card Aided Design). I cut out the shapes of the various components out of
index cards and move them around to see how I can get things to fit. Then when I've got something
I like, I draw it out in a simple drawing program on my Macintosh.
The new design for Totally Offensive goes back to the original design idea, with some
improvements based on experience. The weapon will be at the front, the drive motors at the back,
and the weapon shaft will be touching the ground (thus, no big suspended weight, and hopefully,
less bending stress on the frame). The frame will be two big chunks of 3" aluminum channel connected
in the shape of a "T". I happened to have some channel handy, but if I hadn't, I would have gone to
my local scrapyard to get some,
or popped over to Online Metals for
what I need (they'll even cut it for you). The nice thing about channel is that it's strong and
you can tuck stuff inside it; in the design, I'll be putting the batteries on top of the
front spar (with the weapon motor underneath), and the speed controllers on the back spar, covering
everything with a UHMW bonnet (Ultra High Molecular Weight plastic, great stuff. I get mine at
McMaster-Carr, which, btw, is a great source for parts
and high quality tools. Not cheap, but great service and a huge selection!)
Day 1: Mounting the weapon gearbox and motor
OK, let's start building. The first step is to chop-saw
(if you're messing with metal, you'll want one) the spar that
will support the weapon and battery. In this case, it's 15" of channel. In the picture, I've put
the
Whyachi TWM3RS gearbox, with Magmotor and
backplate attached under the channel, so you can see how it fits.
The gearbox will fasten to the channel using 4 bolts. Actually, they're button-head
cap screws. It turns out that everything you and I think of as a bolt is actually one of a huge
variety of cap screws. What Home Depot claims is a bolt is actually a hex-head
cap screw. However, you can magically change a cap screw into a bolt
by threading a nut onto it. It's a yin-yang thing, without the nut, there is no bolt. Believe it
or not, the US Department of Homeland Security has a
21 page document devoted
to the complex problem of distinguishing a bolt from a screw! The things you learn in this sport,
it never cease to amaze me. And after telling you all this, I'm just going to keep on referring to
them as bolts, because it makes everything much easier to read. I just wanted you all to know that
I wasn't some ignorant doofus that didn't know the difference between a bolt and a screw.
Anyway, we have to drill holes in the
channel for the bolts. Fortunately, the engineering diagrams that come with the gearbox tell us
exactly where the holes have to be; our challenge is simply to put them there. If I used my mill,
I could position them to within .001 of an inch, but this kind of precision isn't required. If I
can get to within a few hundredths, it'll be OK. This is where calipers become
indispensible.
First, I painted on some marking fluid, which is basically a thin dye that quickly dries, to
stain the general area where the holes are going to be. Then, setting my calipers appropriately,
I scribed lines that intersect at the drill points. A good practice is to use one corner as your
reference point, and measure from there, so in my case, I measured and scribed all the horizontal
lines from the bottom edge, and all the vertical ones from the right edge.
With the locations marked, it's time to use a
center
drill to make starter holes at each drill point. Center drills are these stubby little drills that
have the great virtue that they don't bend and wander (much) when you drill with them. You start out
by placing the center drill over the right location, and drilling out a tiny little dimple, then you
take a look at it to see if it is in the right spot. If it isn't, you continue drilling, using a little
pressure in the right direction to walk the drill to the center of the crosshairs. With a little practice
you can use your eyeball, a hand drill and a center drill bit to make nice dimples that will guide your
normal drill bits to a precision of a few hundreths of an inch.
Once the center drill dimples were made, I popped in a regular drill bit (use the calipers
to measure a bolt, then go a size larger; it helps if you've got a nice set of drills in 1/32" increments)
and gently drilled out the bolt holes. But there was a problem. The two holes that mount to the front of the
gearbox are so close to the edge that they intersect with the vertical rails of the channel. So
that part of the channel has got to go. I could spend a lot of time and sweat hacksawing it, but it's much
easier to pop a cutting disc into a hand grinder and chop it away, then swap in a grinding disc and
smooth things out. Grinders are great fun, but remember to wear goggles, gloves and a mask; the
metal dust gets everywhere, and your lungs isn't a great place for it. A test fit showed me that I
also had to grind a little cutout
into the vertical rails near the rear bolt holes to allow the bolts to lie flat without touching the
rails, and that when bolted down, the rear motor mount plate wasn't flush with the side of the channel,
which meant that my bolt holes weren't quite in the right places.
To fix that problem, I simply picked one of the 4 holes that I felt was closest to the right
position, put in a bolt, then
picked another hole, noted how it had to move in order to line things up right, and
used a round file to elongate the hole in that direction. A few minutes of filing and I had a perfect fit.
My next task was to add 3 bolt holes to allow me to secure the rear motor mount plate. This
was easy. While the gearbox and motor were still bolted to the channel, I scribed a line along the back of
the rear plate, removed the motor and gearbox, used my calipers to scribe location marks, then
center drilled, drilled the actual holes, and it fit perfectly without any adjustments.
Finally, I used a hole saw to cut out a big hole at the front of the channel to match where
the top bearing is on the gearbox. The metal there doesn't really add to the strength of the robot,
removing it saves me a little weight,
and having access to the shaft mount bolt might come in handy if for some reason I wanted to remove
the shaft without removing the gearbox. Plus I got to use my hole saw, which is fun, if messy.
I also drilled 3 bolt holes at the other end of the channel that I'll use to fasten it to the other
bit of channel (that forms the top of the "T"). Not bad for two hours work, but that's enough for day 1.
Day 2: The top of the T
Before I started constructing the other spar of the robot, there was a leftover job
from day 1. I needed to drill and tap some holes in the rails so I could attach the UHMW bonnet I will
later construct.
Tapping holes is one of
the basic skills all robot builders learn. Tapping is the process
of forming threads on the inside of a hole so you can screw in a bolt without using a nut. It's
actually pretty easy.
Step 1 is to drill a hole of the right size. The correct size depends on the size and
thread of the screw, which you look up on a
tap size chart. So, for the
10-32 (size 10, 32 threads per inch) screws I'll be using, in aluminum, I needed to drill a hole
.1590 inches in diameter, which is a #21 drill bit (in addition to the fractional size drill bits
most people have, there are a whole series of numbered and lettered sizes).
After that, I chose a tap. It was a 10-32 tap, of course, but there are a zillion
different styles of taps for special circumstances. Fortunately, 99.44% of the time you'll either
use a tapered tap or a bottoming tap. A tapered tap is tapered at the top so it's easy to get
the tap started, but this means you can't put threads at the bottom of a blind hole (a hole
that has a bottom, that doesn't go all the way through the material). So what you do in this
instance is tap as far as you can with the tapered tap, then switch to a blunter bottoming
tap and finish out the last few threads.
In this case, since I was tapping through holes, I only needed the tapered tap. I dribbled
a little tapping fluid into the hole (it lubricates things, making the tapping easier, and also
helps remove the metal filings),
inserted the tap, and gently started turning it, a half-turn at a time.
It's important to be very gentle with taps. They are incredibly hard, so they cut
easily, but in consequence, they are incredibly brittle. Dropping one on a hard floor can
cause it to shatter like glass. If you accidentally put some bending force on a tap, or
keep on turning after it's bottomed out in the hole, you can easily snap it. Since extracting
a broken tap is a serious headache (sometimes you have to corrode them with nitric acid),
err on the cautious side; if you get any unexpected resistance, back off.
My standard procedure is to make 4 half-turns, then back off 1 half turn to let the
shavings the tap has removed accumulate in the flutes of the tap. If I'm tapping a blind
hole, then after 6-8 full turns, I'll remove the tap, clean it, blow out the hole will canned
air, add some tapping fluid, and then continue. For some expert advice on tapping, check out
the Ultimate Guide to Tapping Holes on the
BattleBots Forum.
Tapping can be very zenlike, and the
history and technology of
screws is fascinating. And my ability to resist a bad pun is a sad sign of
maturity on my part.
On
to the second spar: after chop-sawing a 16" section of channel, I marked, drilled
and tapped some holes so that I could bolt it to the main spar. Some of you may be saying
to yourself, "Those bolts can't possibly hold up to the stress of combat, they'll break or
the threads will strip!" You're absolutely correct, but that is not their purpose. At this
point in the construction process, I was mostly concerned about how things fitt together. Later,
after everything is nice and snug, I can weld the frame into a single chunk of metal.
With the spars connected, I next had to mount the two drive assemblies. I'm reusing my
Whyachi T-Box gearboxes and wheels
for this purpose.
From the top of the weapon gearbox to the bottom of the weapon shaft is a distance of 4.625",
so the distance from the top of the motor spar to the bottom of the wheels must be the same. Since (in
the orientation I'm using) the Whyachi gearbox is 4.034", I had to displace the gearbox
0.591" down. Actually, I decided to shift them down 0.650, so that with brand-new tires, the blade
tip will be very slightly tilted down in the front, and as they wear, it will gradually level out and tilt up.
At this point, I was only be able to use 2 out of the three mounting holes, but later, once I get everything in
place, I can build some mount plate extensions. On an earlier version of Totally Offensive, I
also tapped the motor mount screw access holes on the side of the gearbox, and I plan to
weld on some endplates that will bolt to the motor.
For the initial test fits, I just bolted the motors onto the front of the spar, then once
everything was just right, I made cutouts in the bottom spar rail with my grinder and a cutting disc,
and mounted them in their proper
place. Hey, it's looking like a real robot! Let's put on the blade and see if that fits...
Total project time so far: 5 hours.
Day 3: Mounting Motors
One of the design flaws in the
previous version of Totally Offensive was the mounting of the motors. The motors mount to the
T-box gearbox via a couple of small screws, and the stress inflicted by a big hit, transmitted entirely
through those screws, was actually bending the faceplates of the motors, loosening the endbells,
and even deforming the motor can itself. To reduce this stress, I need to clamp the motor firmly
to the frame as well as to the T-box. I could have machined a custom motor mount that clamps onto
the endbell of the motor, but since I'm building this robot with only hand tools, I simply went off
to Home Depot and got a couple of small hose clamps, dremelled some slots into the frame, threaded
the clamps through, and used them to secure the motors.
The velcro strips attached to the frame serve two purposes. First, since the motors are
very slightly smaller in diameter than the gearbox, they act as spacers and let me clamp the motors
tightly without putting bending stress on them. Secondly, they act as electrical insulation and will
help prevent a battle-damage-induced short through the frame (which is why I wrapped the endbells
with electrical tape). I'll probably add some tape under the clamp itself.
Total project time so far: 6.5 hours.
Day 4: Installing speed controllers
Speed controllers are an essential
part of every combat robot. They read a signal from the RC radio receiver and convert it to a
proportional output voltage that drives the motors forwards or backwards at any speed you want.
The way they work is pretty cool; basically, several hundred or thousand times a second, they
turn on and off the flow of electricity to the motors, based on their control input. So if you
command 100% forward, the juice will be on all of the time, whereas 50% forward will result in
the speed controller pulsing the electricty half on and half off. Because they do this pulsing
so quickly, as far as the motor is concerned, it's getting a constant flow of electricity whose
voltage varies, and so it spins faster or slower. When you tell the speed controller to reverse,
it swaps + for - and runs the motor in reverse.
People in the sport often talk about the "magic smoke" which makes all electronics
work (when the smoke leaves the robot, the robot stops working!). The #1 source of magic smoke
is a blown speed controller. It's pretty but can be expensive!
The speed controllers I'm using for my drive motors are
IFI Robotics Victor 883's.
These are actually serious overkill for the motors I'm using (Monster Max RC Truck motors), but I had some
extra lying around, so I figured I'd use them. In normal use, they won't even get warm, so I didn't bother
mounting the cooling fans.
Mounting them was pretty simple. I markeded the position of the mounting holes, and drilled them out
(I also drilled a cable hole to let me pass wires from the back to the underneath of the main spar),
then attached some velcro both to the spar and to the bottom of the speed controllers. This will not only help
them stay where I want them, but act as a bit of a shock absorber. Then I mounted them using, of all things,
a couple of wire ties. Think of one wire tie as the bolt and the head of another as a nut. It's cheap,
simple, and won't come loose unless you clip if off, not to mention the fact that since it's not
conductive, it doesn't provide a path for a short-circuit to the frame.
Next, I clipped the motor leads to the correct length, crimped on some ring terminals, and screwed
them onto the output sides of the Victors. Those of you with good eyes will note that because I've connected
the motors identically, and because one motor/gearbox is the mirror image of the other, when they get the
same input, one will drive the wrong way. But since I never can remember which way I need to wire them,
I'll just wait until I test it to figure it out, and swap the motor wires on one of the controllers.
You may also notice that I've rather sloppily hot-glue'd the PWM cables into the speed controllers.
This is to prevent a nasty impact from popping them out.
Wiring up the input side of the speed controllers requires a bit more care, as hooking them up the
wrong way is usually an expensive mistake. Just remember
RED = POSITIVE, BLACK = NEGATIVE/GROUND. At this point, I haven't decided
how I'm going to run the power from the batteries to the motors, but for testing I'll just leave the
cable hanging free. Oh, and by the way, whenever possible, twist power cables together, it helps reduce
interference.
One important tip: get yourself some good wire crimpers and strippers. You'll save yourself a lot of
aggravation, not to mention heartache when you find out that you lost a fight because a bad crimp came loose
(I speak from experience). A good crimper costs $20-25, but is well worth it.
Total project time so far: 8.5 hours.
Day 5: Wiring and testing
At
this point, all that's left to do is attach a couple more components, wire everything up, and I'll have
a working robot. First, I drilled and tapped some mounting holes for the
Whyachi C1 Contactor that I use to
control the weapon motor. It's basically a big relay. The relay is controlled by a
Team Delta D-Switch, which
uses a R/C radio channel to switch a circuit. I was too lazy to remove the motor and gearbox while
drilling and tapping, so I covered them with a plastic bag to keep out the metal shavings.
I mounted the contactor under the front spar behind the motor mount plate. It's actually fairly well
protected, since anything that wants to hit it has to get past the blade, but I'll add some extra armor
(aluminum or UHMW, depending on how much weight I have available) to give it some extra protection. The
major hazard will probably be something getting swept into it by the blade.
Because this version of the robot is electrically identical to the previous version, all the
wiring work was already done. It was just a matter of plugging everything together, piling it on
top of the robot, and testing it.
The two speed controllers plugged into channels 1 and 2 of the radio, and the D-Switch plugged
into channel 6, because that's the way I have my radio set up. One of the nice things about having
a really good radio, like the
Futaba 9-CAP, is
that you can really tune the performance of your robot by making adjustments in the radio.
One basic thing that most robots need to do is "mixing". When using a standard airplane-style RC
transmitter (which most people use), it's common to want to do "1 stick driving". This means that the
right stick on the transmitter does all the driving and steering; you push forward to go forward, pull
back to go backward, and push left or right to turn the robot. So you've got a forward/back input, and a
left/right input.
The problem is that what you really want is a left motor input and a right motor input. If you've
got a simple radio, you have to do the mixing on the robot using a
mixer module that does the
conversion for you. But if you've got a radio like the 9CAP, you can
do the mixing in the radio.
Once I had everything wired up, but before I powered up the robot, I made sure to check for shorts
to the frame from all of the powered points on the robot, in particular the motors and speed controllers.
It only takes a minute and can save you from an expensive and smelly mistake. I also put some blocks
under the gearboxes so that the drive wheels were spinning in the air. Finally, I made sure that the
weapon drive motor was not connected; I know it works and I have no need to spin it up inside the
house!
Then I cautiously applied radio power, main power -- no smoke, great -- and gently tested the
drive motors. As expected, one of them (the left one) needed to be reversed. With that done, I
was ready to give it a real test drive.
The battery I was using is the right size (24 volts and 3.6 Amp-Hours of capacity), but the wrong shape.
Now that I've got my basic assembly and testing done, I'm going to send it back to
Robotic Power Solutions and get it split and
remade into two 12V packs. So the final battery system will be lower and longer that the one I
temporarily installed. Also on my upgrade todo list: the current battery uses NiCad cells, and if I
change over to NiMH cells, I can go from 24V to 36V without increasing the weight, and substantially
increase the speed of my weapon. But that's down the road a bit.
Off to the garage for a test drive! You can see a short quicktime of testing
here. In the background you can see the carcass of
one of my earlier, less successful robots, and my knees, which my wife assures me are incredibly
sexy.
And hey, what do you know, it drives fairly well
on concrete, which is very slippery compared to the traction-painted steel floor of the typical
arena. During this test, I was running the motors at a maximum of 20% of full speed, and even so,
was able to spin the wheels. This is not unexpected due to the fact that a lot of the weight of the
robot is over the front weapon shaft, but isn't a real problem (I hope) because this bot isn't
designed to race around, it stalks the opponent and disembowels them.
Day 6: Merrily we weld away
After
spending a few hours making some stiffening triangles, it was time to weld up the frame. Since I'm only a novice welder (and don't
have a welding machine of my own), I took
everything down to ProFab of Wilmington (conveniently located right next door to the AnimEigo offices) to get the work done. My
son James (the main driver of Totally Offensive) came along to learn a bit about welding.
In this particular case, we used "Tungsten Inert Gas" or "TIG" welding. The area to be welded is bathed in a puddle of
inert argon gas, and an electrical arc flows from a tungsten electrode into the metal to be welded. Extra filler metal is
added to beef up the weld. TIG is slower than the more common MIG or stick welding processes, but when done by someone who
knows how to do it properly, usually gives better results.
In addition to just welding the parts together, we also welded on some stiffening triangles to transfer impact stresses
around the joints, plus a couple of tabs that prevent the weapon gearbox from moving side to side (reducing shear stress on
the mounting bolts; they are weakest in shear). Finally we added some motor mount end plates that both let us mount the motor
gearboxes more securely, and also act as fenders.
Once I got the frame back home, I popped it in the oven (350F for 8 hours, then air-cool) to artificially age it. The heat of welding
causes the metal around the welds to weaken, but over time some of this strength comes back as the metal ages. By heating it up for a while,
you can speed up the aging process. There's a more radical procedure called solution heat-treatment that can make the metal even stronger,
but it requires much higher temperatures and can cause deformation in the structure if not done properly, so I'll settle for what I can
do at home. Hopefully this robot won't be a turkey, even though it was cooked like one.
Day 7: In the US, it's a hood; in the UK, a bonnet...
Spent a couple of hours installing two of the three UHMW bonnets that will protect the various components. The
curved UHMW panels will protect against shrapnel and act as shock absorbers. UHMW is flexible but extremely tough, which
makes it a great material for this purpose. After cutting the UHMW with some shears, I used a punch
to mark the locations where I needed to drill holes so I could bolt it to the frame; UHMW is soft so
it's easy to make nice dents that will guide the drill.
After the holes were drilled, I tested the fit by attaching the bonnets to the frame with
some bolts. A couple of the holes weren't quite in the right place, but there's a cute trick for
adjusting them. With bolts loosely screwed in through the correctly positioned holes, I inserted
the correct tap through the incorrectly positioned ones and threaded it into the tapped hole in
the frame. This caused the tap to chew into the UHMW on one side of the hole and ream it out just
enough to allow a proper fit by the screw. A silly trick, but it works well.
Day 8: She's all done!
When the rebuilt Battlepacks
came back, it was time for final assembly. I strapped down the packs to the main frame using
pipe clamps, and tucked in little UHMW mounting plates to mount the radio and radio battery
on top of the main batteries. These plates will do double-duty as shock absorbers. The radio
and its battery are attached to the plates with velcro and wire-ties. Then I quickly made
a UHMW hood to cover everything; the big slot allows me to connect and disconnect main power.
On the other side of the hood are is the main power light and a small hole for connecting
and disconnecting the receiver battery. The radio antenna attaches to the top of the hood with,
you guessed it, more velcro.
The initial driving tests are all satisfactory. All that remains is to put some threadlocker
on the structural bolts, and head to the competition!
Total project time: 20 hours.
Robot Assault 2004: Getting the bugs out
It's always a good idea to shake down a new design so, I
decided to give it the acid test at Robot Assault over the Labor Day Weekend.
That way, I could see what was going wrong and have a month to fix it before Nationals. Unfortunately, my kids couldn't attend
because of school conflicts, so I had to fight the robot by myself.
As you can see, the bot dished out
some nasty hits in the process of working its way to a 3rd place finish (pictured is
Intrusive Interloper,
built by Team Radicus). TO's two losses happened in two different ways: the first one
was a classic "turtle" knockout; a really tough wedge named
Xhilarating ImpaX (mounting a new tool steel wedge and fenders in an
obvious "Totally Anti-Offensive" countermeasure) popped the bot up into the air, it came down on its back, and no amount of tweaking
of the weapon could get it to quite bounce onto its feet again. Oh well, you have to hand it to
Team Rolling Thunder, they had a game
plan and they executed it! Then in the loser's bracket finals, TO delivered a massive hit to veteran robot
Eat Hitch and Die that totally mangled one corner of the bot, but the shock was so massive that several of the solder
joints in one of TO's battery packs failed, knocking it out. Eat Hitch wasn't quite so dead, so it won.
Plus, in the fights that TO won, self-inflicted damage included a twisted front
spar, some broken welds, and another battery pack failure (fortunately, the final hit in that fight also KO'd the opponent).
So there was lots of room for improvement. On the bright side, the electronics and motors performed flawlessly; those pipe
clamp motor mounts really did the trick and kept the drive motors from getting damaged.
One of the great things about tournaments is that it gives you a chance to really chat with the other builders and
swap ideas. So it was that Paul Ventimiglia and I were discussing our robots; his spinner, Green Wave, was hitting
so hard that it literally vaporized its own bearings, and I was wondering what I could do about the frame twisting on Totally
Offensive.
The problem was that when TO hit another bot, the shock caused the bottom of the gearbox (near the blade) to want to move
laterally, and this caused the main spar channel (to which the top of the gearbox was bolted) to twist. Unfortunately, while channel is very resistant to bending, it's
weak when it comes to twisting. So clearly, the channel had to be reinforced, but unfortunately, I didn't have much weight
to play with; TO was right at the 30lb limit already.
I knew that I was going to have to get some new battery
packs, and that if I changed from NiCd to NiMH batteries, I could
save a couple of pounds. So I had a little weight to work with. Finally, Paul came up with a great suggestion -- run some
straps or cables from the bottom of the gearbox to the back spars, and put them into
tension. That way, when the gearbox wanted to move due to recoil from a hit, a lot of the force would be transmitted down
the strap into the back spar, and before the gearbox could twist the main spar, it would have to bend the back spar! Or
at least, that's the theory...
I thought about this evil plan all the way home (a long drive), and emailed
Terry at Team Whyachi to see what he could come up with. A few days
and $40.00 later, I had some custom-made Titanium straps, with tabs already bent into them. After a little trimming to fit,
I cranked in the tension on the bolts and (I hope) had one problem solved.
A little extra welding fixed the cracked welds, and I was ready to tackle the battery problem. I asked around on the
BattleBots Forum for suggestions, which lead to a most generous offer from
Dick Stuplich (Driver of New Cruelty, Goosfraba and many other great
bots) to build me some custom NiMH packs. Dick evolved what looks to be an really sturdy pack design; the cells are assembled
as long sticks of 4 cells, end to end, then 5 of these sticks were assembled in an X configuration and fiberglassed together
into a rock-solid brick that fit exactly into the channel on top of the robot! If one of these packs breaks, then I'm not
concerned because the rest of the robot will be a twisted pile of wreckage. The only possible downside I see is that the
packs may get very hot and won't be able to cool down quickly because of all the insulating fiberglass. Oh well, we'll
see how they work out; worst case, I'll cool them in liquid nitrogen and
make ice-cream with the leftovers.
Dick, your humble guinea pig abases himself before you.
Finally, with the 1.3 lbs I had left over, I fabricated some new covers for the robot, and added a big hoop as
a passive self-righting mechanism. The combination of the hoop and the fact that so much of the weight of the robot is
located low and forward means that it has no stable positions that don't automatically roll it back onto its wheels!
Actually, that's not quite correct -- it is possible, though extremely unlikely, for the robot to end up resting on
its side. However, when that happens, one of the drive wheels is in contact with the floor, and a quick pulse of
forward throttle pops the robot upright again!
So, $280.00 of repairs and improvements later, Totally Offensive is ready for Nationals. Wish us luck, my kids
will need it, as the field in the 30lb class contains some seriously nasty robots.
Total project time: 30 hours.
2004 Nationals: Totally Offensive is Totally Triumphant!
We went to San Francisco hoping to make a good showing. What happened exceeded our
wildest dreams!
James and Alex drove their robot to a perfect 5-0 record with 4 knockouts, and Totally Offensive
is the 2004 Robot Fighting League National Champion in the Featherweight class! Dishing out the
damage to all comers, the bot made it to the final match largely undamaged, but there, a bit of nervousness
and overconfidence resulted in a bruising fight with Bear Tracks, a similar bar spinner, that basically
totalled T.O's frame. Despite the damage, James was still able to manuever the robot and delivered a double
wheelectomy that left Bear Tracks helpless! Victory was ours!
While the titanium stiffening straps helped a bit, they didn't completely solve the frame twisting
problem; that will have to be addressed as we design a new version of the robot for the 2005 season. But
Dick's batteries performed flawlessly!
So now you've seen how it's done. Why aren't you busy building a robot that can trash ours and
take the title away from us? Good luck! If you're wondering what you'll be up against, check
out the Totally Offensive 2005 Build Report.
Final Build Cost Tally
The following table lists the component costs and sources for all the parts used in the construction of Totally Offensive, including
spare parts. My actual cost was much lower since many of the parts were scavenged from other robots. A simple starter robot, such
as a wedge or brick, would be much less, and once you've built your first robot, your second one is much cheaper, because a lot of
the components can be shared (like radios, spare batteries and motors, etc.)
| Item |
Source |
Quantity |
Price |
| Aluminum Channel |
Online Metals |
4 feet | $ 19.50 |
| 0.125 UHMW Sheet (Part #8752K411) |
McMaster-Carr |
24" x 24" | $ 12.64 |
| TWM3RS Weapon Gearbox |
Team Whyachi |
1 | $385.00 |
| Magmotor Short Support Kit |
Team Whyachi |
1 | $ 42.00 |
| S28-150 Magmotor Weapon Motor |
RobotCombat.com |
2 (1 spare) | $600.00 |
| 22" 6.9lb Hardened Tool Steel Blade |
Team Whyachi |
1 | $180.00 |
| C1 Contactor (Weapon Controller) |
Team Whyachi |
2 (1 spare) | $ 82.00 |
| TBox Drive Gearbox |
Team Whyachi |
2 | $360.00 |
| 40SH10 TBox Shaft |
Team Whyachi |
3 (1 spare) | $ 86.00 |
| W238-100-44H Wheel |
Team Whyachi |
3 (1 spare) | $ 93.00 |
| Titanium Tension Straps |
Team Whyachi |
5 (1 spare) | $ 40.00 |
| Trinity Monster Maxx Mild Motors |
Tower Hobbies |
4 (2 spare) | $170.00 |
| Robinson Racing 32P 12T Pinion Gears |
Tower Hobbies |
4 (2 spare) | $ 15.00 |
| Victor 883 Speed Controllers |
RobotCombat.com |
3 (1 spare) | $450.00 |
| Victor PWM Signal Driver Cables |
RobotCombat.com |
3 (1 spare) | $ 45.00 |
| Team Delta RCE200(B) Switch |
Team Delta |
1 | $ 20.00 |
| Dick's Bricks 24v NiMH power packs |
Dick Stuplich |
2 (1 spare) | $200.00 |
| Futaba 9CAP Transmitter & Receiver |
RobotCombat.com |
1 | $370.00 |
| Misc Bolts, Washers & Screws |
BWS |
Various | $ 10.00 |
| Hookup wire, connectors, pipe clamps, etc. |
Lowes |
Various | $ 20.00 |
| TIG Welding |
ProFab |
5 hours | $240.00 |
| Total (Hope my wife doesn't read this): |
$3440.14 |
| My kids winning the championship: |
PRICELESS! |
| |
