Archive for August, 2008

 

Nuclear Kitten 5 and Pop Quiz 2: Final Update

Aug 31, 2008 in Bots, Nuclear Kitten 5, Pop Quiz 2, Project Build Reports

I got a new camera.

It won the features-and-price contest for having a full manual control mode, like a micro-DSLR, which I felt was worth the extra pile of change. However, the vast majority of the time I’ll be using it on dumb point-and-shoot mode anyway. Regardless, this means REAL PICS!

Robot Microbattles is TODAY! Nuclear Kitten and Pop Quiz are ready. They’ve been ready, but I have been making random little changes to make their operation more reliable. Since build reports stop at the competition, this will be the final report for both bots.

Nuclear Kitten 5

I’ve been hitting random objects with NK to test its durability and driveability. So far, it has passed on all accounts. NK has historically been my most successful bot, so I hope I’ve passed the reliability onto this build to let it continue.

I made a change to the drivetrain in the form of getting rid of the timing belts. While having synchronous belt drive made the bot handle that much smoother, they also transmitted shock through the drivetrain much better. On a good impact, the recoil of the weapon would throw the bot backwards. It used to be that the O-ring belt drive of NK4 would absorb this impact, but the timing belt translated this into a violent torque impulse at the drive motor.

The result was that I blew out the left side drive gearbox. After frantically ordering a replacement set and having it 3-day delivered, I decided to bring back the O-ring drive, just running on the timing pulleys instead.

These are square-profile BUNA o-rings from McMaster. I selected a loop that was a full inch shorter than the circumference of the timing belts in order to allow the O-rings to stretch – without this stretch, they will not grip the pulleys.

This works extremely well. The bot still handles like a 4WD, but no further impacts have affected the drive motors. The motors are even more isolated from shock than NK4, since the weapon pod is floating.

And so with this revision, here is the final finished pic of NK5.

Pop Quiz 2

If I learned anything from this build, it’s that the new 2.4Ghz RC systems rock. Both NK and Uberclocker run them, and I have no issues with reception, signal loss, or range, at all. Pop Quiz is the complete opposite of that.

It uses a 75mhz GWS micro-receiver, designed for use in tiny indoor and park R/C airplanes. This means it really doesn’t have much in terms of noise rejection or good filtering, nor that much gain. Throw it from the environment of a balsa wood airplane with at most a few servos and one prop motor and into a tiny confined space, antenna trimmed, surrounded by both sides by conductive ground planes, with 4 high-current open-frame DC motors and 3 high-speed switching ESCs, and there are problems.

The bot is completely undriveable with the weapon on (Oh, did I mention the ring of spinning permanent magnets and a big metal twirly thing?). I actually can’t predict which direction it will take off in when I move the stick. I only have the ability to declare that movement is needed.

So PQ is semi-autonomous in the drivetrain. Hopefully I can emulate controlled movement enough to not get counted out. The weapon works great, and should bounce it around enough to not need controlled movement.

I lost one drive motor in testing to a thrown winding (Not a gearbox failure like I suspected any failure would be). I didn’t bring spares, and these are not motors you can run into RadioShack and get. Fail.

So I had to cut it out of the bot. This front wheel will probably be taped over with cellophane tape to give it more slip. Fortunately PQ2 is 4WD and I have three more motors. It actually handled better with this motor gone.

When all was said and done, I squished everything back in and closed the top. Here’s Pop Quiz 2.0.

Yes, the blade hits at the 3/4 mark of a quarter (about 18 and a half cents). It’s actually raised about 1/16″ higher than the design by fiber clutch washers which I put in to avoid metal-on-metal grinding. I will leave it this way, since .6875 is good enough, and it gives more top clearance to prevent disaster.

I leave for Microbattles in 5 hours.

Ãœberclocker Update 18: I think I’ve fooled the Robot Gods… for now Edition

Aug 23, 2008 in Bots, Nuclear Kitten 5, Pop Quiz 2, Überclocker

Alright, so I’m in Atlanta.

Sunday and Monday evening were mostly spent tuning the signal module. I found out that the Victors do indeed have optocouplers larger than the asses of most Hollywood women, and not even the Atmel chips could drive them. In the end, I still have to splice some small signal transistors (ran open-emitter style for non-inverting buffering goodness) into the PWM cables so the Victors would notice my desperate attempts to communicate.

The signal module is an Arduino embedded Atmel controller, and the code was all written by me. It’s my most complicated recent microcontroller adventure. I excavated a little of what I learned about discrete-time control systems from 6.01 to implement the control loop. Essentially, the AVR waits for an incoming R/C signal and reads each pulse width to determine my throttle stick position (and thus the fork position). It then has 18 milliseconds to determine what to do with it. 18 millseconds at 16MHZ is alot of time.

The AVR chips have a 10-bit analog-to-digital converter, so it reads the arm potentiometer and returns a value between 0 and 1023. Conveniently, R/C PWM signals have pulses between 1000uS and 2000uS, which is close enough to 1024 and 2048, 2^n values. So it was rather easy to make a proportional-only controller by scaling the armpot values. I added an integral term to the output equation also to get rid of the steady-state positioning error that resulted from the Victor controller’s deadband around 1500uS. This got the arm responsive enough for me to call it a night and tune it later.

Then stuff broke.

Of COURSE I had to go back and make a few solder joints more permanent, and otherwise touch up the PCB I mounted everything to. I didn’t notice that I had made a very tiny solder bridge between +22 volts (the robot’s main battery) and the output side of the 5v regulator chip.

When I plugged it back in, the AVR controller quite literally exploded in my face. It’s in several pieces.

So then I really did have to call it a night…and started packing the bot up. I ordered a new Arduino board (since the roasted one is very securely soldered to the PCB and not worth the effort to extract) and should have it next week. For anyone interested in learning from my adventures in making a giant servo, I’ll post my code when it’s cleaned up and tuned.

Uberclocker

I stuffed two suitcases full of robot equipment with clothing and personal supplies interspersed between the goods. Here’s Uberclocker locked into its handy bot carriage, with spare parts and hardware. As there was plenty of space in the suitcase, I zip tied NK and Pop Quiz in, to keep the bots in one place. The other suitcase contained the charger, transmitters, and tools.

Somehow this all got past Homeland Hilarity, but the botcase suffered a bent caster wheel. In order to make the thing self-supporting, I borrowed a hammer from one of the baggage offices to beat it back into shape. Wait, I just got handed a massive claw hammer by an airport desk clerk? Huh?

Anyways, in Atlanta, I got to work on the new clamp arm actuator. Here’s a rendering of the new design.

The gist of it: Solid aluminum structure, indirect gear drive. The leadscrew is fixed to the fr0k and does not rotate – instead, this whole assembly climbs up and down it. This takes all the loads off the motor shaft and instead transfers it to some big chunks of metal. The only way to destroy the clamp arm now is to rip the leadscrew clean out, or applying so much force that the nut strips out.

I’m not saying this can’t happen, but it’s unlikely to happen at D*C.

Here it is implemented. Yes, I’ve been COMPLETELY spoiled by easy access to machine tools – while this would have taken me a few minutes walking to get done at MITERS, I had to drive no less than 50 miles round-trip to make this sucker, and of course make Dale put up with my really bad habits that I’ve accumulated from working without supervision.

Suburban Atlanta needs a few you-build-it freelance engineering places, so I think I’ll make it a mission later in life to start them. Although I suspect the culture to support it is missing…

This actuator is a throwback to something I made in late 2005 for Science Olympiad. My robot entry for that year had a linear actuator attached to a movable jaw that could clamp down on objects. I custom-made the actuator using some gears hacked out of an R/C car, a block of UHMW, and patience on the drill press. The output gear is threaded on the inside to ride up and down a piece of threaded rod.

Pop Quiz 2

In the same machining session, I made a new blade collar for Pop Quiz. This retains the blade on its drive motor, so it’s a pretty critical part. Stock steel shaft collars didn’t come in the width I needed, and would have been a tad too heavy. So I made a quick aluminum clamp-style shaft collar, 1/4″ wide, 1.25″ OD.

I’m still facing severe failsafing issues with Pop Quiz. It’s been a reported problem with the tiny GWS receivers, and there seem to be no workarounds besides the usual anti-RF stuff, but the thing is still pretty jittery. It goes nuts if I turn the radio off when anything is moving.

Pretty bad for something that can remove outlying body parts with ease. I might switch back to the other micro-style receiver and mod it for height.

Here’s Pop Quiz in its final state.

In other bot news, Nuclear Kitten’s replacement battery pack arrived and will be swapped in shortly.

Dragon Con is A WEEK AWAY!

Ãœberclocker Update 17: This is not funny Edition

Aug 16, 2008 in Bots, Project Build Reports, Überclocker

Actually it is. From the past several builds, I have come to a conclusion that beginning exactly one week before I am to leave for an event, things will go horribly wrong that will leave whatever I am building crippled and disabled and unable to function up to its full potential. We saw this with TB4.5MCE for Battlebots IQ, where I lost an arm ESC and couldn’t obtain a main battery worth its weight in beans. TB4.5SP1 for Motorama had a whole third of the bot go out two days before I left, and only some last minute hacks allowed it to run at all.

There are plenty of rational explanations, of course, since I usually get to rigorously testing my engineered systems only in that last week or so, and untested systems like to break if they’re not properly designed and built. But the fun explanation is that the robot gods are just out to get me.

Oddly enough, nothing wrong wrong with NK except losing a LiPo cell, which is not a terminal disaster.

Over the past few days, I have been slowly wiring up Uberclocker. Working in an environment where everyone else is also working on something interesting means I get distracted far too easily, so it’s really taken _that long_ to wire the bot up.

Here’s most of the wiring. For a bot with 5 motors, 4 controllers, and 2 batteries, the wire layout is surprisngly clean. In this picture, the master power switch had not been mounted and no cell balancer wiring is installed.

This was the condition of the wiring for the bot’s first remote-controlled powerlift – the previous one was by touching wires to battery leads, which isn’t exactly a legit in-arena tactic. So what weighs more than or equal to a 30lber?

Well, at MITERS, plenty of things, but the only thing compact enough was this…

…45 pound rackmount uninterruptible power supply battery pack.

I approve of this bot. It can raise the 45 pounds pretty briskly and can suspend it in the air with no motor power applied (with the controller in Brake mode), and can bob it up and down. I don’t have an amperage number yet.

As expected, the back end of the bot lurched upwards on beginning the lift, but fell back down due to the support from the rollers in the front. I didn’t practice balancing the bot with a load on the rollers, but that should probably happen some time.

Then stuff broke.

First, while running the clamp arm with no load, the Banebots 12-45 controller popped its reverse direction. It was a very audible (and frightening) pop that made me think one of the FETs on the Victor 883 controlling the fork motors fried.This was strange, since I was using a motor far, far smaller than what the BB controller should be capable of handling, and it was not clamping down on something or hitting a travel limit. Just small reverse throttle.

So I have started the replacement-under-warranty process with Banebots. We’ll see how THAT one goes – chances are they’ll be considerate and ship me a new one in Atlanta.

After the mild scare, I decided to continue testing by dropping in the ANT150 controller from Nuclear Kitten. But again, that didn’t last long, because…

… the actuator motor shaft snapped. The B62 has a reported history of fracturing output shafts, and it looks like the bug hit me. However, it looks like the shaft actually broke in flexure, which tells me that the radial support at the base of the leadscrew is poor.

I’m not surprised. When the clamp arm hits something on its way down, the load on the actuator is essentially pure compression. Any buckling of the part connections at all could lead to catastrophic failure, which seems like what happend in this case.

This is bad. Very, very bad. Considering the whole leadscrew assembly was green Loctite’d together, it’s going to be nearly impossible to extract the broken shaft stub. A replacement motor is a hefty $30, and I’d have to buy and machine a new section of leadscrew.

Bad indeed. Without the upper clamp arm, Uberclocker is a giant spatula of questionable sexiness.

I’m now in the process of specing out parts and designing a new leadscrew machanism. This will be an indirect drive type device that will take the stress off the motor shaft itself. The motor, furthermore, will be mobile along with the upper clamp instead of being pivoted in a fixed location on the forks. Instead, the leadscrew will be fixed with a pin type joint at the base.

Using an indirect drive leadscrew mechanism, similar to the “beak” of my 2005 Science Olympiad robot, should make the whole thing more reliable anyway.

This will have to be fabbed in Atlanta, since I won’t get any parts I order now by Monday and actually get anything done. So with the clamp out of commission, I tied it to the fr0k to secure it and continued finishing and finalizing the wiring. I installed the battery balancer plug, another Convenient DB9 Connector of LiPo Balancingâ„¢ , physically mounted the power switch, and cut holes in the top plate to accomodate them.

Besides the DB9 connector, I also put in a direct connection to the battery using a Deans female side connector mounted to the left Ebay. For this bot, the charging current needed will exceed the limitations of the 24-22 gauge wire that fits in the DB9 connector, so I will use that only for the charger’s balancing function, and the big connector for actual charging current.

After tidying up the interior some, it was time for a beauty shot. Cheesy cell phone camera style.

The bot’s now only missing stickers and blinkenlichten.

And also the signal module. The signal module is my name for the receiver + fork controller combo that will occupy the left side Ebay. The fork controller will take input from the Spektrum Rx and do two things – one is buffer the signals for the Victors, and two is correlate my transmitter’s Throttle stick position with the position of the fork. It will take input from the potentiometer mounted to the fork and drive the ESC accordingly.

I decided to let it process al l the receiver signals (and passing drivetrain commands straight through) instead of making a separate buffer board for the drive only – it keeps the wiring cleaner and allows for future expansion of more advanced control schemes.

There is minimal hardware left to implement, since the microcontroller already comes with its neato support board.

There are exactly three days left. Can I get everything working?!?

Nuclear Kitten Update 4: Uh oh, shit’s starting to break Edition

Aug 13, 2008 in Bots, Nuclear Kitten 5

It *ALMOST* worked.

I wired through the night to get NK up and moving. I found out that the frame has enough space in it to stuff in TB4.5′s Spektrum receiver, which means one less bot I have to worry about receiving its own noise and interpreting it as a signal. And it worked – the bot is hellishly zippy, and the disc is absolutely frightening.

However, the drive motors draw more current than the Barello ANT controller can handle – it enters thermal failsafe after a few seconds of hard driving.

This is due in part to the massively increased current capacity of the batteries, being LiPo cells, the higher voltage I’m running the bot at (11.1v versus 9.6 for the previous frame) and the fact that all four wheels have traction, as opposed to NK4 which really drove on its two front wheels.

Secondly, while finalizing the wiring, I resoldered the main power connector in backwards (the consequence of working so early in the morning?). While normally this would just result in a big spark and me pitching the bot halfway across the room, it actually blew out one of the tabs on the battery pack. Thin nickel tabs on a pack that can dump 30 amps, absolutely brilliant.

The blown tab is too short to link back to the rest of the pack. This means I will have to get a new battery pack, or risk running 7.4 volts. It’s too close to departure date to have a pack shipped to me here in Cambridge, from Hong Kong, and so I will order a battery and have it shipped to Atlanta.

Meanwhile, NK5.0 IS DONE AND MOVING! One down, two to go! What else is going to break?

Ãœberclocker Update 16, Nuclear Kitten update 3: A picture is worth 9000 words Edition

Aug 10, 2008 in Bots, Nuclear Kitten 5, Pop Quiz 2, Project Build Reports, Überclocker

So I can’t find the sub-micron sized grain of dust or metal shaving that is caught in the lens actuator of my camera. Unfortunately, this means that I will have to start haunting Ebay and local consumer electronics outlets. Until then, I suppose everyone who reads this site (all -0.000001259 of you) will be in suspense, save the occasional grainy cell phone camera picture.

The basic rundown is that Ãœberclocker is ready to be wired and NK is also approaching mechanical completion.

Ãœberclocker

I drew up the “EBay” assemblies and prepared flat patterns to make them out of sheet metal. Unlike TB4.5SP1 which had all its electronics in a bunch, Ãœberclocker features “distributed electronics”. There’s not really an advantage to either method, just that I couldn’t find space in this bot to slam all the electronics in one place.

Each is a work of 1/16″ aluminum origami (read: smashing in a vise and banging on it while wielding a torch) that bolts to a close frame member. There is nothing secured to the baseplate, unlike TB (and all my previous bots). Furthermore, all of the Ebays are dismountable from the outside.

This is the left side Ebay, which houses the main switch and Convenient DB9 Connector of LiPo Balancingâ„¢. The big switch came off a 1980s era PC power supply which, despite being a 7 inch cube, could manage maybe 200 watts. It is mounted such that the top bezel is flush with the top cover plate of the bot. A little rectangle will be cut out of the top plate to pass it.

Since Ãœberclocker isn’t a severe duty bot, I decided to forego making a normally closed switch like I did for TB. The detent is strong enough for me to not worry about it randomly clicking off.

The rear Ebay clamps the batteries between itself and the rear of the bot. It also carries the Victor 883 controllers for the drive motors. The right side Ebay houses a third Victor and a small Banebots ESC to control the clamp motor.

fr0kp0t mounted. This was an afterthought, and I really didn’t want to take the entire thing apart again, so it was just clamped and hand drilled. The potentiometer is rather exposed for something which will be telling the fr0k everything about where it is in life, so it might get a larger metal cage. The pot is coupled to the fr0k through the shaft set screw.

There is currently no bottom hard mechanical stop for the fr0k, and so it can swing all the way around the bottom of the bot (no doubt ripping off the entire bottom plate in the process). To prevent this, I’ll add some things that stick out of the fr0k towers. While I could easily limit the travel in software, a hardware backup is good for preventing self-eating disasters.

So Ãœberclocker is ready to be wired up and programmed. The fr0k, with the chain tightened (by removing one pitch with an offset link), successfully powerlifted 30 pounds, so I know torque isn’t a problem. Maybe there will be a drive test soon…

Nuclear Kitten 5

NK went from 0 to about 50% done in a day due to the “snap-together” chassis. The only thing I needed to build for that were some nutstrips ( 1/4″ aluminum squares with regularly spaced tapped holes), which was tedious but trivial. Past that, I had to machine up the weapon motor and drive wheels.

I specifically bought a 4-40 spiral-tipped tap so I could powertap all the holes, 7 each in 12 nutstrips. Sadly enough, I’m short about 4 inches worth of 1/4″ square aluminum stock to make the last of the nutstrips. This is within range of just biting the proverbial bullet and machining down some bigger stock.

Blurry ass-picture showing some frame bits and the nutstrips. Ignore the obvious non-trusses that populate the inner rails. My excuse is that it was 5AM Joltgineering – I’m not sure why I didn’t just link the corners with triangles instead of making an ugly V shape.

I like this “Chinese Puzzle” frame – I think I’ll keep the tricks in mind for future projects.

While the waterjet pump was still primed, I tossed on a plate of 1/4″ 4140 and cut out the blades.

Because the blades are now steel, they are thinner in profile. This should be more than enough for 3lber duty, especially after heat-treating (famous last words).

After discovering that the grungy MITERS horizontal bandsaw cut much faster when I dropped a weight on the end of it, I started on the disc motor. The motor itself is almost identical in structure to my wheelmotor and much simpler in construction. I also didn’t have a stiff and consistent tool holding device when I built the wheelmotor. Overall, this resulted in a very fast build of the disc motor structure.

I made two side plates out of 3″ aluminum round and the center axle out of 1″. A steel pipe was turned into the magnet ring. This steel pipe appeared fine on inspection, but upon contact with a cutting tool, turned to powder on the inside. Fortunately, the rust wasn’t deep enough to affect the final ring dimensions.

I also made all four drive wheels. Each wheel consists of a stock SDP (redundancy?) pulley, an aluminum “rim”, and an O-ring “tire. The process was fast, since each rim only had radial features that could all be made by selective parting tool use. The drive motors themselves are in transit and should arrive Monday. Previously, I ordered a few Speed 300 size RC motors off Ebay to retain the absurd speed and maneuverability of NK4, which featured those motors spliced into the aforementioned gearboxen.

Upgrading the wheel size the morning before I ordered all my parts (more 5AM  Joltgineering), I failed to account for the fact that one of the frame connector pieces now interfered with the front wheels. Something will probably be sanded down.

After taking a Taco Bell break, I put the disc motor together for a test fit. This motor has a very tight airgap to maximize torque, and I was worried about the tolerances adding up and causing the stator to grind against the magnets. To my surprise, it ran concentric and true – with no rubbing. Cheers for the robot gods. I then went back and preliminarily installed all 28 magnets (in groups of 2) with ultra-thin CA glue. The empty spaces will be filled in with epoxy.

So, here’s a picture of NK in that “well, it LOOKS done” pose.

Old NK is undergoing the scrapping process to extract the goodies (there’s not much left).

Only a few spacers here and there remain to be done on NK.

So the summary of things to do is:

Ãœberclocker

  1. Wire up!
  2. ???
  3. Profit!
    1. Make bot carrier

NK5

  1. Weapon pivot axle and spacers
  2. Wind motor
  3. Tune drivetrain for belt tension
  4. Wire up

PQ2

  1. Wait for giant plate of titanium (est. Tuesday)
  2. Cut blades out of Ti
  3. Machine a blade retaining collar
  4. Integrate receiver and mixer into a single module
  5. Think of how to get reception

A week remains…