LOLrioKart takes up a good portion of floor space in the back half of the room and is occasionally used to store all my crap. It’s also a pain to move around because of its mass, and a pain to work on the electricals because they are all very low to the ground. If I want to test the drivetrain, I had to lift the kart and balance it on a set of automotive jacks. Don’t even mention that time I had to swap the battery packs…

So for a while, I had wanted a lift or crane to suspend the kart from. I didn’t take the idea seriously until Spring term ended, when I started looking for options. I became partial to a ceiling-mounted hoist because of the ability to send the kart all the way to the top for storage and extra floor space.

The kart only weighs about 200 pounds empty, which is a essentially trivial load in the world of winches and cranes. MIT building N52 used to be a factory, and factories in the early 20th century were built to last forever. The ceilings are all solid concrete, more than a foot thick. Essentially, almost anything would have worked, and I briefly considered just gearing down a beefy DC motor instead of buying a specifically designed winch or hoist motor.

But as luck would have it, Craigslist produced a pristine example of an ATV winch for sale locally, so I quickly jumped on it.

This Master Lock (I thought they just sold locks, but I was wrong) unit seems to be a pretty standard offer in the world of cheap generic utility winches. It made some substellar sounds when loaded and the drum finish was pretty rough, but I’m going to assume that it won’t kill me. Too badly, anyway.

The mounting holes in the winch frame were located in a place where I couldn’t access them with a powered screw driving tool. They were designed to be mounted to brackets first, which are in turn mounted to your choice of stationary reference frame.

…so I had to devise my own. This 3/8″ aluminum plate was just hanging out in the cave of materials. I could probably have waterjetted any number of small robot parts from it, but hey.

The two large middle holes are countersunk on the other side to fit 5/16″-18 socket head cap screws. The six surrounding holes are countersunk to fit some 1/4″ flathead concrete bolts. MITERS had a large stock of “tapcon” style concrete screws (which do not use anchors), probably from back when we bolted stuff to the ceiling all the time.

Because bolting things to the ceiling isn’t exactly a precision machining activity, I used spraypaint and sprayed a pattern into the ceiling, using the mounting plate as a template.

The winch itself is mounted using two 5/16″-18 socket head cap screws and grade 8 nuts and washers. This should not be the first failure point.

A little while later…

I borrowed a hammerdrill and a 5/32″ concrete bit and went to town on the ceiling. Reportedly, the hammering noise was ungodly loud, even on the third floor. I guess that’s what happens when you bang on a solid concrete building.

I learned that a hammerdrill is best used not under intense drilling pressure, but rather under modest pressure. If you push too hard, you dampen the chiseling action and the effect is diminished. This was well reflected in me taking almost half a minute to drill the first hole – but the last only took 5 seconds.

Unfortunately, attempt #1 to mount the winch didn’t go well. I made the mistake of reading a different box of screws (which specified a 5/32″ drill). 1/4″ tapcon screws need a 3/16″ drill. The difference meant that I managed to shear off the screw halfway into the final depth.

Epic fail.

I quickly went back with a 3/16″ bit to expand the other holes, and the rest went well. Friends and cohorts spotted the high-altitude work and helped hold the 25 pound winch up while I drove in the screws.

But after much sweat, concrete dust, and loud construction noises…

FLYING LOLRIOKART

Well, at least I know my five-bolted rig can hold up 1 LOLrioKart. The only power supply strong enough to supply the current demand of the winch was a big Optima lead-acid battery.

Because this is a shady winch being held up by shady screws into shady century-old concrete, we started piling heavy things into the kart to see if we could find the maximum load. Helmets and face shields were aplenty during this exercise, because even if the kart was only 1″ off the ground during it, the falling 20-odd pounds of steel were a concern.

Things piled into the kart include two truck disc brakes (60 pounds), the lead acid battery (50 pounds), the Defibrillator (a.k.a the kart charger, 25 pounds), one brushless Etek (25 pounds), a huge linear power supply (25 pounds), and a milling vise (40 pounds). Powered lifts and interrupted drops were attempted to put force on the mounting, but it was solid. I think it’s proved itself.

Assuming I did install the screws correctly and that the concrete is not crumbly, each 1/4″ concrete screw is rated to a maximum of 1,100 pounds assuming a 1.5″ depth. There are five holding the plate to the ceiling for a cool ultimate tensile strength of 5,000 pounds. Even accounting for imperfections, the winch should stall long before the screws pull out. It’s still unlikely that I’ll ever allow anything more than the kart by itself to be hoisted, or items of similar weight, because it can get overhead and that can end badly.

Yup, it’s a hoverkart. Any questions?

…so it’s actually just a very small cross section of all the testing that’s been going on. I’ve been thrashing the new controller mercilessly. Never before could I throw the kart around like that without the risk of something setting on fire – but it’s been bulletproof.

That J-turn took about 15 tries to get right, by the way, and it still isn’t quite up to spec. I’d need much more space to do it correctly…

I am satisfied. Next: 6 quadrants.

As of about right now, LOLrioKart features a full bidirectional and regenerative motor controller. This might not be true soon since I still have to perform stress tests, but nobody can point at me and say PICS OR IT DIDN’T HAPPEN either because there are pics and it did happen!

Continuing from Day 1, here’s most of the high current bus wiring. I was able to salvage much of the weird grounding strap/braid from the previous controller to use on this one, since it had much shorter wiring spans. The braid is much more flexible than the equivalent amount of ~6 gauge wire, and passes right through the copper screw lugs.

Also in the picture is the large 10 ohm power resistor that shunts the contactor. It’s the precharge resistor for the bus capacitors. I included a check in this iteration of the software to prevent the motor driver from outputting any power if the contactor itself is not latched. When not latched, the logic can still draw power through the resistor.

Check out the double bus capacitors. Some 8 gauge noodle wire links the caps in parallel and also to the  gigaFETs next to them.

It turns out that 3900uF of bus capacitance isn’t really enough, so I might have to double-deck the caps later on.

Now comes the enormous 375A Anderson powerpole. Since the whole controller is upside-down this time, I decided to put the Etek on connectors. I received a set of these for my birthday as a gag gift (they were spraypainted red and black to prevent me from plugging them in backwards!).

The pigtails are 2 gauge welding cables. The kart seems to feature  a very diverse conductor menagerie, ranging from 10 to 2 gauge. Overall, the longest runs get the biggest gauge, so I think it all balances out.

It’s time for mounting! I found some abnormally long wood screws, so just decided to use them to mount the controller instead of trying to use machine screws and tapping plywood like last time.

I was too lazy to actually draw out the mounting hole pattern on the wood first, so I just set the controller up on the machined standoffs and dimpled the wood in the center of each hole with a punch. Then the wood screws were too large to thread themselves into the wood without splitting, so I pilot drilled the dimples. All the precarious controller balancing and punching probably came out to more effort than just drawing the mounting pattern, but hey…

The main battery connector and power switch are located across from eachother, joined with the shortest possible wiring runs. Overall, I cut a few inches out of the amount of wire needed to link all the parts together.

Here’s a look under the upside-down controller. I probably could have dropped the standoff height another half inch so the wires keep themselves in place.

Alright, now we get to my favorite part of this controller design – the entire reason why I chose the Epic Heatsink of Epicness and the upside down design: the case mod.

Two 120mm LED fans park right on top of the Epic Heatsink and tap off the 12 volt line. They are probably the worst possible choice for real controller fans, since they flow as much air as I do by sleeping, but look at how awesome they are.

I dropped the assembly in for a quick test spin of the drivetrain. To my utter surprise, nothing exploded on application of battery power. This is a new development in the line of Kartrollers – generally I have to rebuild it every time I plug it in. The precharge circuit works much faster than the old one now, and without danger of being set on fire by the motor trying to pull current through the bypass resistor.

Downside: That’s probably the most awkward possible place for the battery switch.

Well that didn’t last long. The first wheels-up test resulted in a sudden and mysterious total system shutdown after entering a particularly stiff period of motor braking. No fire, just a sudden crowbar-like effect. After half an hour of diagnostics (the gate driver chips were both toasted and replaced), the controller worked again… barely.  It managed to spin a small test motor at about…. 10 RPM or so.

The above high-side scope trace led me to discover that, horrifyingly enough, I was using bootstrap diodes at almost twice their rated blocking voltage.

The bootstrap diodes are pointing towards the brown capacitors in the picture.

I randomly picked some diodes out of our “power diode” bin which turned out to be 1n5819s on closer inspection – rated to 40 volts absolute maximum. First off, the kart battery was 60+ volts charged, and on regenerative braking, the voltage spiked beyond that. Hitting the brake hard most likely resulted in a mortal blow for both diodes, momentarily firing( x >> 12.0) volts into the 12v auxiliary bus.. This caused the DC/DC converter to freak out and shut down, powering everything off.

Better than setting everything on fire.

The diode carcasses still conducted a little, just barely enough to make the high side FETs also  conduct just a little to run the test motor.

The solution was to replace them with some more random diodes from the bin. This time, they were 1n4007s, good to $MORE_VOLTAGE_THAN_THE_KART_SHOULD_EVER_SEE (namely, 1000 volts).  The test motor almost flew off the table, a good sign.

The negative current is a sign of the regen brake working! Negative current in this case means it’s going back into the battery bank.  This was captured by revving the kart drivetrain up to full speed (a scary ordeal with it balancing delicately on an automotive jack, by the way… if it fell off, it would probably have broke into TMRC) and then letting off the throttle. The synchronous H-bridge took care of the power conversion.

So what always happens after I quickly check that the vehicle works in midair?

Yup. With the reverse switch and brake pot still dangling precariously off the terminal strip, I hurriedly threw the throttle back on and took the kart out for a hallway joyride.

…but not before rigging myself an easily accessible emergency stop. The little chunk of wire is normally wedged into the contactor switch terminals, turning it on. Making sure I was tied to it at all times meant that I could quickly disengage the contactor if… stuff happened.

Fortunately, it wasn’t necessary. There’s not any test video since I decided to save face in case the whole thing detonated in the middle of the hallway, but rest assured video is forthcoming. After doing the drooping-wire test run, I decided to clean everything up and actually consolidate the controls into one easily-accessible location.

…yeah, um, so my user interface design skills need some work.

But the whole thing is conveniently located and fulfills its requirements. The Big Red Switch is a legitimate emergency stop button now! It opens up a normally closed circuit when I pound on it, turning off the contactor. As some people may remember, the previous kartswitch was actually a normally open button, so you had to push to engage and pull to disengage.

The direction selector is located next to the Big Red Switch , and above it is a potentiometer to adjust the drag braking force. This pot should really be a logarithmic pot, since the filter time constant calculation behaves exponentially near zero – very fine adjustment of the pot yields very large coast/brake differences at the “coast” end of travel.

The current-voltage-micrometer makes a return.

And here’s the whole nest from the other side.  I was lucky to find an “INCREASE OUTPUT” knob, which really should have been reserved for cruise control.

The nest enters into the terminal block in a neat and orderly fashion.

A new day and an overnight trickle charge on the batteries means plenty of testing and fire to come! Stay tuned for more pics and video of Syncrec LOLrioKart!

LOLrioKart in a forest of scooter handlebars

After being totally whored out for CPW by giving rides to prefrosh and generally being hooned around campus, I remounted the kart sound system to use as a portable music device and all-around pimpmobile for Mini-Maker Faire the weekend after. But thirty minutes before departing for the Faire, I powered the kart on.

Nothing happened.

It wasn’t a spectacular explosion. It wasn’t even a smoke column or a quiet sizzling sound. Through a half hour of diagnostics, I only managed to deduce that the gate driver stopped driving the gate. The power FETs appeared to be fine. Whatever – we were late already, so I just pushed the kart a mile or so to the Cambridge Public Library where the event was being held, blasting bad eurodance and girl pop along the way.

Afterwards, I decided to just remake the controller, yet again. The latest iteration, version 5, did work, but was haphazardly put together (i.e. pretty par for the course). I wanted something that had more structure than a bundle of wires. I also wanted to keep advancing on the number of features and simplifying the schematic. This resulted in the great gate driver hunt of 2010, out of which the IR2184s came.

As described, the existing controller core is kind of a kludge. It was just an Arduino with a protoboard shoved onto it which had the gate drive componentry. Alongside was a 12v DC/DC converter. Everything was just packed inside a little box with foam bits, and not otherwise secured.

Running the controls in open air for half an hour to diagnose the latest failure resulted in me discovering that the DC/DC converter got hot. Way hot – hot enough for it to shut itself off, which explained alot of kart flakiness.

So I did what any shady college engineer would do given a totally rigged and quasi-functioning project – rip everything the hell apart.

I dismounted everything and took the mounting board out of the kart.

The big black slick in the righthand corner is from grease and oil being thrown upon it by the drive chain.

Let’s begin again. We start with an epic heatsink:

I found this in the materials cave at MITERS. There are a few reasons why I wanted an Epic Heatsink, the foremost of which is epicness.

The other reason is modularity. At 300 x 200mm, this heat sink had enough area on its back to mount practically all the core componentry. I wanted to move away from point-to-point hardwiring onto something a little more modular, like a commercial controller.

Mechanical issue: Only one side of the heatsink had space for mounting holes or other provisions. The other seems to have been trimmed off by someone else.

It just so happened that removing the bent fin would open up a width on that side of the heat sink equal to the width of the flange on the other.

…so that’s what I did. Remembering a trick someone showed me a while ago, I slammed the heatsink on the mill and cut down the bent fin.

The 3/4″ rod fits into the 5/8″ T-slot on a Bridgeport mill table and provides a straight edge to bump a long part against, for axial trueness without resorting to hammer bashing and a dial indicator. Normally the rod should be precisoin ground or otherwise some length standard, but in this case, “precision” meant “closest one to me that qualified for diameter”.

With the edge cut down but no holes drilled yet, I started laying out the components. There were a few iterations – this is the first. The FETs would be centrally located so one fan can blow on the back of the heatsink to cool them. The Arduino would sit right next to the gate pins.

But that was before I found these cool double capacitor holders, which allowed me to pack on more bus capacitance than the kart currently had (2400uF to 3900uF) and keep it on the heatsink mount. Of course, caps like this don’t need heatsinking themselves, but the mounting surface was there.

Through more cramming, I decided to also move the contactor onboard, such that the unit was essentially self-contained. Other parts of the power system, like the battery cutoff switch, terminal blocks, and power connector, would remain offboard.

I took a quick break and laid out the parts. The heatsink itself already had quite a few holes drilled in it from its previous, more legitimate application, but sadly none of them lined up. Thus, it was back to the mill to drill the mounting holes, which was a rather boring half hour of trying to match a digital readout with a hole chart.

Some screws and washers later, and the pieces are mounted. The contactor is case-grounded, and my intention was to keep the heatsink itself isolated. So, it’s secured by large nylon screws and insulative washers.  The Arduino and capacitor block are mounted with nylon standoffs.

The DC/DC converter has an isolated case, so it was bolted straight to the heatsink. It will probably be alot happier this way. It, and the two Ixys gigaFETs, got a liberal coating of Arctic Silver thermal compound on the bottom.

Here’s the new driver board. It’s another Arduino-perfboard assembly that houses two IR21844 dual gate drivers. Unlike Kartroller 5, the high sides are bootstrapped, not supplied via isolated regulator. I decided this was a good simplicity tradeoff.  Additionally, it’s much more failsafe – the boostrapped high side can’t maintain 100% duty cycle, so it can never get stuck on like the low side drive can.

The wiring nest on the underside, in typical me-fashion. I’m using some cool 30 gauge solid tinned wire that has very unique insulation properties. It doesn’t smoke or burn when it contacts a soldering tip, but rather just melts out of the way. It’s very thin and easily damaged.

But those same properties let me just mash the wire into the solder joint with the tip and it will automagically fuse with the solder ball that’s already present. This has enabled me to just go pin-to-pin and link up an entire net very quickly.

I think it’s wire-wrapping wire, but I’m not sure. What is it and where can I  get more?! MITERS is almost out!

Check out the integrated signal terminal block. I even labeled it so I don’t forget which wire goes where!

Most low power wiring joined. Notice the small 5mm power plug coming out of the DC/DC converter and into the Arduino. I figured it was better than soldering directly to the input pins…

Here’s the gate wiring from the Arduino shield to the FETs. Gate resistors are integrated into the heatshrunk portion of each cable end. I really should have local pulldown resistors too, but elected to keep those on the shield.

The rightmost gate wire got red shrink because I ran out of black.

I began to add other signal interfacing hardware to the shield at this point. First to be completed is the contactor detector. It’s a wire (the red one) connected to the contactor (through a resistor divider) which enables the software to see whether or not the contactor is latched. One of the problems with Kartroller 5 is that it just let me (or someone) gun the throttle even with the contactor was in the precharge state.

This just caused the controller side voltage to fall under 36 volts, resetting the DC/DC converter, thus shutting the logic off. In turn, the FETs turned off, causing the voltage to rise again, which… well, it led to an unhappy cycle of kart-twitching until I reached for the battery switch.

The next few lengths of wire connect the Arduino to the important I/O – throttle, brake, and reverse.

With all the signal side wiring done, I could start on the software. Here’s a makeshift user interface with a throttle pot, a brake pot, and a reverse switch.

Wait, what do you mean “brake pot”?

Does that mean I’m removing the disc brakes?!

Nope. But for the longest time, I’ve been telling people that LOLrioKart has regenerative braking. Now, this is pretty much a total lie, because Kartrollers 3 through 5 (the ones with a proper half-bridge layout) couldn’t actually switch the high side transistor to build up the motor current which allows regen to occur.

Now that I have both the confidence and a driver chip that allows this to happen easily, I’ve run into a smaller but not insurmountable (and actually convenient) problem. The IR21844 is a synchronous rectifier chip that takes one input and internally splits it into two complementary outputs. That means as long as the chip is enabled, the high side and low side are always switching in complement with one another. This is the way it’s supposed to work in a switching power supply.

It works that way in a DC motor too, but the experience is much different. In most vehicles, if you stomp on the accelerator and let the vehicle build speed, it will just coast until friction brings everything to a stop. In an EV with no regen capability or that enabled by another switch, the same is true.

However, in an EV driven by a synchronous regulator  that is always on, you’d faceplant into the steering wheel on throttle lift-off because the motor will try its best to brake when the rectifier is in its circulation state.

To put it another way, in an always-active synchronous regulator, the motor is braking some percentage of the time and accelerating the rest of the time, and you just control the percentage. Full stop is just like 100% braking. To get the throttle-coast effect, I’d have to alternately switch the IN and #SD pins on the 21844, instead of just driving IN low. This takes more processing and pin-banging than I care to think about. Additionally, when I do that, I bypass the internal built-in shoot-through protection of the 21844, ruining everything I have ever loved.

Simulated Inertia

So what to do in this case? Now, one method to achieve this coast state is to use current control (instead of voltage control, which is much simpler and what most motor controllers perform) and command a zero current. Zero current necessitates the regulator switching at just the right ratio such that the motor’s back EMF equals the output, and following this zero current state all the way down to a stop.

I didn’t design this system with a current sensor in mind, nor do I feel like making another feedback disaster yet. So I opted to take advantage of the fact that an average speeding shopping cart and the average coasting motor are both readily approximated by a first order system.

Figure 1: Example of first order decay with varying time constants

Refer to the above pretty picture . The pictured time axis is arbitrary. This is a simplified model of “I floor the accelerator on LOLrioKart for a short amount of time, then lift totally off”.

In an ideal, friction-free world, any object accelerated to a certain speed will just stay at that speed forever, represented by the purple line. That doesn’t work in real life without a net power input. Friction, a force dependent on speed, will eventually bring everything to a halt. In an electromechanical system, any opposing force dependent on velocity can be lumped into a “friction” term. Varying this causes the rate of decay to change, resulting in the red, yellow, and green curves. The red curve represents very sharp decay, or high damping. The yellow is an intermediate value, and the green a low value – the system coasts for a long time before stopping.

Now, in an EV, two major forces can dominate the “friction” component. Real mechanical brakes and tire scrubbing with the road, or the electric motor itself. An electric motor can provide a resistive force proportional to the speed of rotation by acting as a generator, extracting work from the system’s kinetic energy. Even better is that this rate of work extraction can be modulated – that’s the whole point of regen braking.

Conjecture 1: By programming in an artificial throttle decay curve, I can simulate the effect of neutral coasting.

This involves interpreting the throttle in the following manner:  Follow it exactly if I am depressing the pedal (value increasing), but send it through a digital low pass filter if the direction changes, such as if I back off. This software LPF will slowly ramp the motor down over a long period of time.

In a synchronous rectifier arrangement like I have the gate drivers set to be, regenerative braking occurs automatically if the kart is trying to drive the motor faster than it wants to go given a certain duty cycle of the output. Essentially, the motor acts as the dominant source of “friction” in this case, but instead of dumping the energy as heat, it dumps it back into the battery bank!

Conversely, if the kart wants to coast down slower than the set decay curve, the motor will spend a little energy prolonging it. If I set the decay to zero, it will keep driving the kart at a constant speed forever, at least until I run into the nearest dumptruck. As such, I’ll have to make sure the filter maintains a “minimum decay rate” so it eventually WILL just stop.

Conjecture 2: By making the decay time constant adjustable, I can adjust how much apparent mass the kart has, or modulate the regeneration strength

The job of the “brake pot” is to make true conjecture #2. For now, it will just be on an instrument panel. By adjusting the pot, I can set the coastdown time constant of the kart. Alternatively, if I start with “infinite” coastdown, then increase it, I can rapidly stop the kart through regenerative braking only. This latter effect would be more useful on a brake pedal.

The phenomenon of applying a tiny amount of braking at all times is called “drag braking” in the EV world. The same effect occurs in an automatic transmission car, except it’s not adjustable… on purpose, anyway.

So let’s do it!

Say hello to the ghetto-ass screenshot of my laptop reading an Arduino program through the serial port. I’ve implemented the aforementioned one-way digital LPF in a test program that just makes throttle values.

You can see the variables besides the printout window.

• lastGoodThrVal is just the last output value that the loop calculated
• curThrVal and curBrkVal ought to be self-explanatory – it’s the magnitude of each input.
• brakeFraction and coastFraction are the two components of the digital LPF. coastFraction is the “hold previous value” component, while brakeFraction is the “input component”. That is, coast is the second term and brake is the first term in this equation, which I implement in the program.

This video of it working will save me (30 pictures/s * 1000 words/picture * 30 s) words! Notice:

• with “hard braking”, or curBrkVal high, the output value closely tracking the (scaled) throttle value on the way up as well as down.
• with “slow braking”, or curBrkVal low, the output takes a long time to decay back to zero.
• that if i suddenly hammer the brake while the value is coasting, it will fall very quickly.

Alright, so what’s next after making sure the code works?

HOOK IT UP

I tested from a current and voltage clamped power supply just in case I Did It Wrong™.  You can see the result for yourself.

First, I perform a few start-stop cycles with hard “braking” – or high virtual friction, whichever.  Next, I toggle the reverse switch while the motor is moving slowly. I do this slowly because if I do so too fast, the voltage spike causes the power supply to crowbar, shutting everything off. Not exciting at all.

Finally, I do a few spindowns with the “virtual inertia”(/slow braking/little friction) effect visible.

The shaking motor is due to my very slow update rate of 100 milliseconds. This is well above the motor’s mechanical time constant, so it actively responds to the stepped changes in the controller’s output. LOLrioKart will have a much higher mechanical time constant, but it might still help for me to increase the refresh frequency.

Here’s a short clip of the regenerative brake effect when I suddenly increase the brake percentage while the motor is coasting. The power supply pegging above 50 volts is the best indicator. If I brake too hard, it goes over 55 volts and the PSU gets mad.

4 quadrants of love.

]]> http://www.etotheipiplusone.net/?feed=rss2&p=651 2 http://www.etotheipiplusone.net/?p=642 http://www.etotheipiplusone.net/?p=642#comments Fri, 14 May 2010 08:47:36 +0000 the chuxxor http://www.etotheipiplusone.net/?p=642 As previously illustrated on this site, LOLrioKart’s motor controller is one big lesson in power electronics that I’m learning as I build it (…and blow it up…and rebuild it). While I must admit that it’s frustrating to have a controller that never really works, and which just sits on the metastable edge between functional and explosive, it’s a valuable learning experience to figure out exactly how to steer this edge towards the functional and reliable side.

You know, like finding out that your oscilloscope inputs are grounded so you absolutely shouldn’t probe a floating ground with it because it’ll dead short your power supply through itself and everything else in the path. Many times.

I found out that’s reason #1 why Segfault’s dual custom H-bridges all exploded. Live and learn, I suppose.

Cross-conduction

Another dominant cause of power amplifier failure is shoot-through, or cross conduction. In your average half-bridge circuit, the worst thing that can happen is your top and bottom side switches on the same leg being turned on at once. That usually results in very high pulsed current draw from the power supply, then…

Alright, so it’s not that bad. But shoot-through will generally blow fuses, melt wires, or destroy an output component (in the worst case causing it to fail short) if you‘re not a pussy don’t use fuses like me.  3 phase block commutation generally avoids this shoot-through phenomenon because it ensures the top and bottom sides of one phase are never switched consecutively.

However, DC motor drivers do not have that luxury, because there is essentially only one phase to switch. Therefore, whether in software or through hardware, provisions are generally made for shoot-through protection. It usually takes the form of a section of delay-circuitry, combinatory logic, a delay state in software, or if you’re Shane, driving optocouplers in inverse-parallel because they form a convenient delay circuit already.

During the times I have experimented with full-bridge or half bridge DC motor drivers, I’ve generally used a variant of the delay circuit presented in 6.131, the same one that is documented on Segfault’s progress report post. It takes a while to make, is nonintuitive, and more complex than it needs to be, taking up 5 of 6 inverters in a 7414 hex inverter.

For a little while I’ve been searching for a FET driver chip that has this function built in. After some digging on Digikey and places like FindChips, I found the IR21844.

Like the Ixys IX6R11P7 chips I favor now, it has a standard low side driver and an isolated, bootstrappable high side. The difference between the 2184(4) and the 6R11 is the input. The 6R11 takes two individual inputs, one high and once low. They do not affect eachother, and you could build in a self-destruct mode for your controller if you wanted because it can command both switches to turn on at once.

The IR21844, however, takes a single input and splits it internally into an inverted and noninverted output. So that means if your input is logical low, then the low side is on and high side is off – and vice versa. Not only that, but it also has internal shoot-through protection (IR calls it “deadtime”). Even better is that the 21844 variant allows you to tune the deadtime to the taste of your favorite mutant semiconductors.

It basically tells me that I wasted a year fumbling with this crap when someone already made something I can just buy – like everything else in life. The 21844 comes in 8 and 14 pin DIP through-hole components (+1) as well as SOIC (ewww… but handy if I ever come out of the course VI closet). And so, I gave a few quarters to the Internet semiconductor gumball machine for some units to experiment with.

Here’s the test circuit I rigged together using the schematic found in IR’s datasheet. I’m reading the two outputs using two channels of the MITERscope. The input is a 50% duty cycle square wave, 5Vpp and 10Khz. The single potentiometer is a 100Kohm unit that I’m using for deadtime adjustment.

See? I wasn’t lying. One input to the 21844 is split into two complementary outputs.

This scope screenshot is actually a little deceiving. The vertical scale is 5 volts per division and the zero levels are located on the X axis (channel 1) and 2 major divisions down (channel 2). What looks like the two square waves not meeting in the center is actually them overlapping. I should have used a wider division setting.

What’s really interesting – and what sold me on these – is the built-in deadtime.

Check this zoomed shot of the above square waves. What you see is an approx. 1us delay between when the low side turns off (channel 1 high to low) and the high side turns on (channel 2 low to high).

This is good. This break-before-make behavior ensures that your +V and GND rails never see eachother in the middle.

Even better is the fact that it does the exact opposite thing on the other side. On a low-to-high transition of the output, the high side turns off before the low side turns on again. Again, break before make. This proves that one output isn’t just a time delayed version of the other, it’s legitimately being trimmed to fit neatly into the other.

One implication of this is that the high side will completely shut off at a point before 0% input duty cycle. This seems to be inherent in the nature of delay stages and I conjecture is part of the reason why control deadbands exist. Complementary to this, the low side will completely shut off at a point before 100%, which must be considered if your drive circuitry is bootstrapped on the high side i.e. does not have its own isolated power supply at all times.

The deadtime adjustment is internally grounded on the 2184 but brought out to a pin on the 21844 variant. You connect it to ground through a variable resistance of your choice – datasheet page 10 gives a rundown on deadtime vs. resistance.

As advertised, the deadtime is adjustable from about 400ns (0 ohms)…

…To roughly 2.5us when my 100Kohm pot was pegged. The 21844 appears to support deadtimes of up to 5 microseconds.

Indeed, when I just straight yanked the pot out of the circuit, all the output shut off. I suppose an infinite deadtime is just like being permanently off.

With this in mind, I went ahead and made a little buck converter just like I did in 6.131. I used two IRF2807 FETs that I had leftover, running in bootstrapped high side mode. It was still running on the 50/50 square wave input. Will it work? Will everything explode?!

Well, what happened first was that I shorted the high side through the scope probe, killing half of the 21844.

Sad.

That’s why I bought a rail of 30.

But then it worked! On the (real) first try, even! That is, well enough to totally smoke a power resistor of too small value while I wasn’t watching.

hey guys i made a power electronic widget that worked on the first try do i win a prize do i do i do i

The next stage was to replace the resistor with something more meaningful. Something that actually does some useful work besides making a small smoke column… I mean, it can do that too, but it should be designed to do otherwise.

I have an idea – let’s try a DC motor!

HO REGEN BRAKING YAY

I probably played with this thing for half an hour – just gunning the motor, then quickly dropping the duty cycle and watching the power supply’s voltage spike up. This half bridge circuit brings me one step closer to a 4 quadrant synchronous regenerative drive for LOLrioKart, and who knows what else. To add the other two quadrants, I just need to make another half-bridge and connect the motor up between them.

Speaking of LOLrioKart, can these little driver chips actually handle semiconductors the size of those I have an excessive amount of? One way to find out is to try driving the gate. This is an IXYS VMM65001F, good to 100 volts (same as a Swapfet) and 680 amps, which is at least three Swapfets. This is a bigger transistor than LOLrioKart can ever hope to push. Just because of that, it’s a good example to test the 2184s on.

Also because it is a convenient half-bridge module, and two of them will easily make an oversized H-bridge for said LOLrioKart.

I drove the gigaFET using the bootstrapped high side.

It took a full 1.7us to turn on, where I defined turn-on as the gate voltage reaching approx. 67% of the drive voltage. There are many ways of defining “turn-on”, this is just one of them.You can even see a little bit of the “Miller Plateau“.

This is not bad given that the on period is very long compared to the switching period. During this switch is when the FET acts like a resistor, dissipating power. LOLrioKart’s PWM runs at 4Khz, which is a half-period of .125 milliseconds, still about 70 times longer than a switching cycle. Not the best, but most likely better than anything else I have on the kart at this moment.

And the turnoff occurs a little faster, since the gate voltage only has to reach 4 volts before it falls under the threshold voltage. The turnoff itself is mostly clean.

So where does that leave me? I now have a method of switching a half-bridge in an orderly manner, without external glue circuitry. I also happen to have enough large semiconductors to make a 600 amp peak capacity H-bridge. I have this dumb shopping cart with a motor on it.

Will there be a kartroller version 6?!

Überclocker Remix

Let’s start with some pictures of epic motor ownage. I cracked open the toasted HTI gumball machine motors out of curiosity after removing them from the bot. What awaited me inside was a scene of utter devastation.

That doesn’t look very healthy. It appears the commutator decided to just melt off the backing material. This motor actually still ran, just throwing blazing white sparks everywhere. The discoloration of the copper next to the crater attests to the extreme heating that occured.

The brush cap from the left side, which simply failed open circuit at the event. Well, now the reason is clear why it failed open. Half of the brush conductor spring just sort of flew off and melted itself into the other side of the plastic brush holder.

The carbon brush itself was bouncing around inside the motor.

Another view. That bit of spring must have been pretty hot to instantly melt itself into the plastic.

And another view of the copper droplet that is the commutator. Oddly enough, the windings themselves seemed for the most part to be just fine.

Here’s Überclocker looking decrepit on a table. Since a robot with no motors is akin to a dog with no legs, or a fish with no fins, I began the quest to search for a… well, more legitimate motor. That’s when I remember that I found these, from Way Back.

DeWalt drills are classic musclebot motors. Sadly enough, these were of different voltages (!?), which not only surprised me as to how on earth their previous user expected their creation to move in a straight line, but saddened me because I… well, want mine to.

It was enough to perform a fit test and draw up plans to modify the gearbox to accept these motors while UPS channels their Brownian Motion to get a matched set of motors to me. They are the “new” DeWalt motors, where “new” is relative to 2003 or so. These drills have 3 speeds and are infinitely more of a bitch to mount. So I will only be using the motors in my custom frakenb0xen.

Fit test. The good news is that these motors are roughly the same length as the 700-size HTI motors, but a little fatter. No issue, considering the gearboxes have plenty of wiggle room.

The gearbox modified to accept a DeWalt motor, with its Alien Technology Motor Pinion of neither metric nor Imperial tooth pitch. Needless to say, this will be removed and the HF motor pinion crammed on in place.

So am I over-motoring the HF drill gearbox parts by putting a real motor on them? Perhaps. However, I think it’s a legitimate move in a 30 pound robot, because the laws of physics dictate that I can only put so much power to the ground. I’m mostly after the “real motor” bit, not so much increased drivetrain power, because the robot doesn’t have the traction to use it.

With motors now on the way, this conversion ought to go quickly since I’ve already drilled the new mounting holes to accommodate them. Überclocker should then be able to attend more events.

The (Possibly?) Final Chapter in the Tragedy of the LOLrioKart

So by now all ya’ll have probably heard of this.

While the details surrounding the citation were totally illegitimate and imply a degree of recklessness that was not present at all, the bottom line is that the kart is not going on any more open road adventures until it’s legit. And by legit, I mean registered and insured and fully street legal.

Whatever measure this takes, it will happen. It will simultaneously the most confusing and most glorious thing on the planet.

But the good news is that through two weeks of intense demos and driving during Orientation, the kart didn’t explode. The motor controller, version 6, is more or less stable. That’s huge. That’s like, me doing something right in electronics for once.

Of course, if I actually run the numbers on the electrical characteristics of the power converter, it would probably make real EEs run away to vomit. But the kart has survived more than twenty power cycles without misbehaving, save for the flakey DC-DC converter that caused the initial failure of version 6. A replacement module with better-designed (read: existent) filtering solved the problem.

So I’m satisfied. There will be little active work on LOLrioKart this term, with most of the fleeting effort concentrating on the battery system. After said weeks of operation, two cells in the battery pack are now just resistors. I regularly saw the voltage dipping under 45 volts on acceleration, which is concerning to say the least. Battery management solutions are condensing around me, so I may make the jump to lithium iron phosphate cells.

Now let’s move onto the new shiat.

This is a Xootr Street push scooter.

Gee, that looks kind of like every other push scooter on the planet. You know, like a Razor scooter. I thought you already had one of those? With like… a motor on it, right? That you built? I heard you built a motor. Can you show me how that works? Can you build me one?

… </average_miters_visitor>

Oh, that’s the difference.

As much as I love RazEr when it works, it’s time for me to realize that it’s too freakin’ small. I’ve managed to hit the tiny-but-functional goal, and at the same time the maximum recommended rider weight a few times. With some more scrupulous design, I could probably fit more batteries in there, but otherwise all the useful space is essentially occupied. And while 5 inch wheels are great for shoving your average copier motor core into, they are not great for shoving into your average pothole.

I need something bigger. More legit™. So thanks to MITERS for coming up with an engineering sample of the Xootr Street. I won’t actually be making any mods to this one, since it’s … not mine, and stuff.

This thing measures a bit over 3 feet long when fully deployed. The wheels are 7 inches in diameter and cast aluminum. It’s the smoothest thing ever on the ground because of the large wheel-to-bearing diameter, which minimized rolling friction. And the deck is absolutely enormous….

… and HOLY MAGNESIUM JESUS ON A STICK. It is in fact CNC machined from billet aluminum. These guys are just like me, except with infinitely more style. A scooter? Made from real metal?

And the 10 center-side pockets are just big enough to comfortable seat two A123 26650 cells apiece! How about that. 20 cells yields almost 150 watthours of battery pack energy.

Ground clearance check. The deck height, oddly enough, is almost the same height of the Razor A3 frame. The wheel line is just an inch and a half or so higher to fit the 7″ wheels. Overall, as can be seen, there is about 1.25″ of “fiddle space” from the bottom of the deck (not including the pockets) to the top of the 1.5″ parallel.

This is good, because hiding all the goodies under the vehicle frame contributes to vehicle aesthetics and the illusion that something which is not supposed to be motorized is moving under the directive of an unknown force.

There are no motor drawings or plans for this yet, but the profile of the wheels and their spacious internal diameter make them amenable to stuffing axial flux coreless motors inside, maybe even one per wheel. I’ve been itching to build a real surface-wound (no iron core) pancake motor for a while, but have been put off by their complexity in manufacturing.

As more details condense from the bot-aether, I’ll give this project its own page, category, and possibly a snappy and witty name. This is not a high priority project, as I don’t even have the vehicle yet, and it might spill over into Spring term.

Spring is a better time to blaze around anyway.

Now Announcing Project SEGFAULT

Segways.

A bad pun on the word segue. A fundamentally unstable faceplant-waiting-to-happen of an inverted pendulum. A cool exercise and great demonstration of basic control theory.

DIY balancing personal transporters have been attempted and perfected many times before. It’s almost passè. There’s even instructions on how to do it and code for your choice of microprocessor. All you need is two fat motors, a rate gyroscope, an accelerometer, and determination.

The whole thing about “microprocessors” is what has been putting me off. I like to think I’m familiar with mechanical engineering principles. I’m shake on electronics and EE. But I’m the last person you want to ask about anything software related. I hate software. With a passion. Even though I use it ALL day, I shudder to see what goes on under the glitzy Web 2.0 interface, or under the ultrasonically-welded sealed cap of an Atmel chip.

…so that’s why I want to do it all in ANALOG ELECTRONICS.

That’s correct. Op amps, comparators, linear components, passives… it’s a 6.002 (or 6.101) paradise. I stochastically arrived upon this idea near the beginning of the term, but it took a few weeks before I took it upon myself to do some research on gyros and accelerometers, and sketch out a rudimentary control network composed primarily of rail-to-rail op amps.

Then I remembered that Dale had built an analog balancing robot, so naturally I read the site and discovered I was doing it totally wrong.

I have a sneaking suspicion that a relatively non-chaotic differential equation like the one that governs inverted pendulums can be pretty easily translated to a continuous time control system (analog, as opposed to a discrete time digital control system). The idea as a whole is to have a purely analog, continuous-time front end controlling a Class D amplifier, also known as a switching amplifier or if the output is bidirectional a locked antiphase amplifier. Basically this just means your transducer wiggles back and forth really quickly… but some times more in one direction than another, so the summation of the movements is a velocity.

But Charles, isn’t a switching amplifier a digital thing?!

Yeah, if I implemented a real linear motor driver, I would have a battery life of 30 seconds and require heat sinks the size of Hannah Montana. Sssshh…. don’t tell anybody.

With the plan now more grounded (HURRRRRRRRRR PUN) than before, I’m moving forward with the mechanical details, as I always tend to do first. Once I have a rolling frame, I could conceivably roll analog or digital, or mixed-signal. As always, this entails a trip to MITERS and a few hours of mining for parts.

…

Yeah, so it’s nothing much yet. I grunged these 9″ pneumatic tires for the project as they were the only two matching wheels in MITERS that weren’t already on something.

9 inches? Isn’t that a bit small (&thats_what_she_said;) ? It is, but there’s nothing fundamentally wrong with having smaller wheels on such a machine. It just makes obstacle negoatiation tougher. If anything, I can get away with having less torquey motors because of the increased mechanical advantage.

The design work continues! After I get my control theory a bit more in line, I’ll sketch up a schematic of what I think should work. There are endless supplies of linear circuit components at MITERS and kicking around the EE labs for experimentation. I have accelerometers and gyros on the way from Sparkfun for experimentation.

…Oh, that’s the other cheat here. Real, modern MEMS sensors. I’m going for the analogginess here, not period-realism.

SEGFAULT will get its own page and category as it develops. This is my number one goal for the end of the term, and I’m actually going to try to get the controller graded. Here goes… something!

There are three reasons why I work on the kart more than any sane person world. The first is if I’m not doing anything else at the moment and need a distraction from the tribulations of life. The second is if I’m preparing for an event or situation where it would be publicly seen… after all, a working model is better than a nonfunctional sculpture.

And the third is if I have a neat idea or cool part and it HAS to be implemented NAO.

Like some instrumentation. Because operation of the kart is always a game of power electronic dice, I decided that some kind of readout of system conditions was necessary. It just so happened that MITERS had some old skool panel meters hidden deep within its bowels.  There were a few interesting options, such as a leak rate meter… what on earth does that measure?

I decided to start with a simple battery voltage monitor, since I had no convenient Hall Effect sensor,  shunt, or other low-value resistor (besides the SwapFETs’ incredibly low 2 milliohms) or a real constant current supply to calibrate a current meter.

A semi-known fact is that most ammeters are in fact sensitive voltmeters. While it’s easy to make a loop of wire and a magnet respond to 10,000 volts, it’s not nearly as easy to do with 10,000 amps. So a resistor game is played to turn the 10,000 amps into a very small voltage, like 100 millivolts or something. Enough to tick a needle on a voltmeter that has “10,000 amps” written on it.

You can easily convert an “ammeter” to a voltmeter if the “full scale deflection” voltage and current draw are known.

I settled for this meter for the voltage monitor, since the other one actually says amps on it. This one measures “Current-volts-microns”.

I have no clue what kind of SI unit that is, but it was the winner because of its simple 1-millivolt-per-tick scale.

So let’s convert this 100-mV meter into a 100 volt meter.  To not explode the meter, it should span a voltage of no more than 100 millivolts (0.1V). In a 100 volt system, that means 99.9 volts must be dropped across a resistor in series with it before it’s connected to the circuit under scrutiny.

The meter drew approximately .5 milliamps (0.0005 amps) at full scale deflection. So the resistor in question must drop 99.9 volts while passing 0.0005 amps. Now just pimpslap Georg Ohm and you have the resistance value needed – (99.9 / 0.0005) = about 200,000 ohms. Actally 199800, but I didn’t have one of those, and the kart isn’t going into space or something.

…yet.

Zip tied to the kart.

Through this meter, I found out that the batteries drooped in voltage under a good hard launch from 61 volts (freshly charged) to about 56. So they’re not too dead.

Or they are, but even being completely fucked are still awesome just by virtue of being cacknormous.

In a continuation of Reason #3, I found a road blinkie. You know, those things on top of orange construction barrels. It contains a few amber LEDs, runs off D-cells, and automatically switches on and off via photocell.

Well that was easy enough. A half inch bolt threaded through a spacer and into the mounting point of the light and I had improved the road safety of the kart hundredfold.

Let’s move onto more imporant things. For the past while, the kart has been randomly cutting out. The 12 volt DC/DC converter has been resetting for apparently little reason – not just under acceleration, but even sitting still. I wasn’t sure what was causing it, but suspected some sort of transient effect scaring the DC/DC unit.

In tearing down the electrical system, I decided that it was a good time to build a more legitimate motor driver.

It was time to get away from the cobbled-together hardware PWM generator. Producing signals in software makes for a much more versatile controller that can be reconfigured easily. I happened to have some Arduini kicking around, and a Protoshield kit leftover from last year’s Überclocker build.

Rounding out the components is an IXYS 6 amp dual gate driver with isolated high side. Using a halfbridge driver like this lets the controller perform regenerative braking. The high side required an isolated power supply, so I yanked out this 12v-12v DC/DC converter-converter from another motor driver board. I wanted some more electrical isolation between the fragile microcontroller and the harsh environment of my non-EE projects, so I salvaged some optocouplers from some weird board that had to have been made in the 80s.

Finished a few hours later.

I was able to use the Arduino language’s built-in PWM command, so the software was extremely simple. Normally it operates at 500Hz – far too slow. But changing the timer/counter initial counts causes the PWMs to run substantially quicker. I ended up going with the 4kHz option.

The two gate outputs are on the bottom side of the board. The left is the high side, and the right is the low side. For now, to keep backwards compatibility, I left the high side unconnected in the kart.

Scoping the gate driver outputs. This was using some test code where I had independent control over each channel. The waveforms look good, except for a bit of twanging in the high side, which I suspect is just a ghetto scope probe.

Making the driver board fully modular meant that the system wiring could be cleaned up substantially. Before, I had a mess of signal wiring and power wiring all meeting at the terminal strip. However, I could now devote the entire terminal strip to power connections. The system DC/DC converter (a 12 volt, 3 amp unit) fans out into 5 outputs now, so I don’t have to try putting two wires into one terminal. Overall, everything became more organized.

…

So did it work?

…

No, of course it all blew up. The problem obviously does not lie within the gate driver system, because everything worked fine for about 20 minutes. Then the aforementioned DC/DC converter began repetitively cutting out.When it dies, the entire kart shuts off because the contactor opens up.

I had gotten into the bad habit of curing these brownouts by hard-cycling the battery switch to reset the converter. It worked a few times.

Then when I hit the switch again, the drive FET made a muffled popping noise and the kart jumped for a split second. Then all was quiet.

Okay, so this explosion wasn’t as spectacularly fire-filled as the other 5 or 6, but I still have to remove the whole electrical system to replace the brick. Amazingly enough, the gate driver assembly survived the whole ordeal.

Explanations that my EE friends (who still refuse to just build me a working controller, eh guys?) offered up include transients on the 12v rail resulting from inductive spikes coming from the contactor, or the lack of a local bypass capacitor  on the input side of the converter causing very short voltage dips to shut the converter off.

Either way, the DC/DC unit is now the problem child. The ghetto moves to another part of the city.

At least I got this cool Volvo dashboard gauge cluster  for free at Swapfest. Sort of defeats the purpose of me adding my own voltmeter.

But first, I am proud to announce that the kart is now able to stop.

That is, in under half a block’s distance.

Here’s the reason why. I bought a set of 140mm disc brakes and cable-actuated brake calipers from electricscooterparts.com, which incidentally sells all kinds of electric scooter parts. Now that I know that things like this exist, I wonder why the hell I didn’t spec them out for the kart originally.

Oh, right, because I didn’t know they existed. The quality of components seems to be about par for Orient-imported small vehicle parts; by which I mean the brake disc vent holes had burrs around the edges, the bolt circles were not quite concentric, and the left and right brake calipers were, while sharing the same mounting dimensions, different parts physically beyond being simple mirror images.

So it was out with the old and in with the new. I dismantled the front wheel assembly on each side and cleaned everything off.

The inside of the wheel rims were thoroughly caked in small brake band particles. The fact that there was little in terms of brake left over on the bands themselves probably contributed to the kart’s dismal stopping ability (read: none).

So extremely bald front tires.

The tread was not exactly deep on them to begin with, but all the rough handling, skidding, and serendipitous toe angle of the kart has essentially vaporized the tread off the tires. The rubber thickness is still adequate, but I just shouldn’t be driving in the rain.

Then again, there are bigger problems to expect if I try to drive in the rain.

Integrated wheel-o-brake hub, to be made from hugeluminum round stock. It carries the bolt pattern for the cheap wagon wheels on one side and the brake disc on the other, and is bored for a .5″ bore R8 type ball bearing. I tried my best to perform an interpolation of the intended bolt circle diameter using the three bolt holes on the brake disc.

Here’s the hugeluminum billet in question, set up in a position ripe for disaster. Real machinists and South Bend lathe lovers avert thine eyes.

I needed to turn this billet into two smaller billets, but our horizontal bandsaw was broken, the N51 auto shop’s was optimized for steel cutting, and I was not going to wrestle this through a conventional bandsaw. The last option was chucking it in the lathe and parting down the middle, which filled my imagination with vivid images of tooling setup explosions and broken back gear teeth.

However, with judicious use of centers and power crossfeed, disaster was averted, and I had two equally sized not-hugeluminum billets.

A little while later, a hub emerges. The bolt circle was drilled using my handy dandy indexing fixture.

Test mounting everything. Surprisingly, taking the average of the three bolt radii resulted in a disc that was centered with minimal wobble. It almost makes me think they did it on purpose or something…

Nah.

It’s time for Pretend-O-Brake. Here is a setup testing prospective brake caliper mount positions. The brake caliper was designed by real engineers, so there’s not a single straight line on it to reference dimensions from. Mounting it would be a nontrivial matter, so I decided to resort to some cheating in the form of the abrasive waterjet.

I designed a caliper mount based off existing part dimensions and alot of caliper-balling (eyeballing the dimension measured from an imaginary line projected off the end of your caliper tips, directed towards the feature in question).

Here’s the designed part. Just for kicks, I threw it into Inventor’s built-in ANSYS stress analysis add-on to see theoretically what might happen if I brake too hard. The verdict is that I could make this part out of jello and still have it be able to lock up the front wheels and skid.

Alright, so not jello, but at least birch plywood.

The mounting points for the calipers are slotted such that I have an ability to adjust them a small amount if I found that caliper-balling wasn’t enough.

A few hours later, parts cut out of some leftover half-inch aluminum. Abrasive waterjets are beautiful things.

You know that extra hole next to one of the caliper mounting slots? That was originally for a design-on-the-fly widget to connect the caliper mount to the steering pivot block. However, when I performed a test fit, I realized that I could just cut out a step in the caliper mount and have it slide over the block in question. Square objects cannot rotate over eachother by nature, and the Nut of Wheel-Retaining will hold everything in place.

Well then. Cutting down a portion of thickness is certainly easier than making two more whole parts.

The end result is a quasi-floating brake caliper. Better than a fixed one, IMO, in taking up for the lack of alignment inherent in stuff I build.

All mounted up. The spacer length between the wheel-o-brake and the steering pivot block required a bit of trial and error to get right, but once everything was cranked down, the assembly was solid.

Another view of the assembly, all cabled up.

And now duplicate for the other side, accounting for chirality.

It turns out that real brakes stop moving objects substantially better. I was able to lock up the front wheels and skid during runs in the hallway – sort of the opposite extreme of not being to stop at all, but at least I have the choice of locking up or not. The tires have substantially more traction outside on concrete, and I was not able to lock up (without stomping excessively hard), but that’s a good thing. Stopping distance from top speed was reduced to “under the length of the N52 parking lot”, scientific tests be damned.

Overall, I consider adding brakes to the kart a great success.

Guess what? It’s SWAPFEST time! LOLrioKart has been the unofficial promotional vehicle for Swapfest since May. This time, I could show up and have a fighting chance at not extensively damaging property or causing wanton personal injury.

Then I

Well, I sure as hell didn’t, but something did. I put the kart in a Prominent Advertising Position™, then went around to gather cruft. When I returned to start putzing it around, this happened.

As soon as the battery switch was engaged, the precharge resistor set on fire. This tells me that the ESC is stuck wide open, and so the Etek tries to drink all few hundred amps of its stall current through a 50 ohm, 1 watt straw. It, in turn, does not last long.

Because the ESC was damaged in the ON position, I elected to not hit the contactor button. At that exact moment, it would have taken off and landed in a pile of server parts. Servers are fundamentally more expensive than anything on this vehicle.

And thus I kept the kart off until I grabbed a friend and rolled it back into MITERS. I have not yet opened up the electrics to see what went wrong (besides it existing in the first place), but my suspicion is on the gate driver again.

Anyone have a real, 60 volt commercial DC motor controller of over 300 amps capacity they want to donate to the cause?

Swapfest finds

Swapfest is always an interesting adventure because of the variety of people it brings. By variety, I mean old ham radio enthusiasts. However, the distribution of cruft and oddities is quite Gaussian in nature. There’s tons of the usual – electronics supplies, small discrete components, computer parts. A steady amount of the esoteric but not out of the ordinary, such as vacuum tubes, antique radio equipment, and random shit from someone’s attic/basement/garage/hole-in-the-ground. But every once in a while, you stumble upon something that is so weird or awesome that “Holy Iridium Jesus” is the only proper response.

This is one of those finds. I was told that it is a military aircraft radio component of some sort, an early form of spread spectrum radio called a data translator. I prefer to call it AWESOMSESAUCE.

I didn’t have a real camera available, so the multi-kilopixel cell phone camera has to do. But I think the astonishing engineering detail is visible even from here. The thing is packed solid with conductors, tubes, motorized digital-to-analog converters, and crazy components that I don’t even know the function of. Seriously. A motorized DAC. It even has NUVISTORS. I don’t even know what the hell NUVISTOR is, but it sounds badass.

This is from an era when mechanical engineering and electrical engineering were truly intertwined and engineers had to have a deep understanding of eachothers’ practice. This is not like “mechanical engineers design a structure and the electrical guys slide a board in”. This is “your product is so incredibly part-dense that your fucking components are structural members.

I stand by my position that old people, no matter how weird they smell, are more hardcore than my current generation will ever be.

The only specs that my 3G-enabled friend could dig up was this milspec, which doesn’t really say anything besides “register for our website”. Anyone know what on earth it is?

At the $60 quote price, I was tempted to buy it just so I could put it in a display case. But a display case hardcore enough for this thing would have to be made from cast magnesium with solid hand-refined fused quartz windows and lit by radioactive phosphorescent compounds.

Anyway, to return to Earth, I got a few things that I could, you know, actually use. Past the supply refills for MITERS, I got this box of giant transistors.

A scan of the datasheet when I returned ousted them not as FETs, but IGBTs. Because the ginormoFETs that I bought at Swapfest the last few times were dubbed Swapfets, these are now Swapbutts, because IGBT is only properly pronounced “igbutt”.

The total count is

3x 1000v, 200A

1x 1200V, 150A

and a single 2000 volt, 300 amp unit. H00t.

I also got a Hot Wheels RADAR gun. This is a real RADAR gun, but with reduced power and sensitivity so small children can bite it.

Maybe now I can actually find out how fast I go? It’s also fairly hackable.

I started at iRobot as a summer engineering intern, tasked with building Terminators.  I was invited along, and subsequently went to, the Buckminster Fuller Challenge award ceremony in Chicago with Smart Cities.

And I finished, broke, finished, then broke, then finished LOLrioKart. I’m waiting on the next “broke” cycle.

Let’s start with a batterygasm.

From the deepest dredges of “Well, they technically never asked for it back”, here is a pile of A123 lithium nanophosphate cells. These are the most ballin’ shit in terms of batteries available today.

The backstory is that A123Systems, through its ancient ties with MIT, donated Over 9000™ cells which failed quality control to the Electric Vehicle Team. The Media Lab also received some for sampling, and being Smart Cities, we quickly snapped them up because… well.

Anyway, I was tasked with the fantastic task of determining what “failed quality control” means. For the most part, it means “smudge” or “wrinkle in the cardboard”, but a few cells were genuinely low voltage or had high internal impedance. Through a series of trials and strictly controlled processes involving a giant power resistor, car battery, and multimeter, I determined that essentially 95% of the cells in each case of 100 were most likely good for our purposes.

Great news for us, better news for me, because A123 probably has more.

Anyway, to equal the 54 volt, 30AH nickel batteries I already have (nominally – these cells are totally fucked and probably return less than 20AH as a pack), I would need about 225 cells – 15 cells in series for a nominal 48 volts, and 15 cells in parallel for about 33AH. The reason I calculated the numbers for 48 volts is because I do have access to 15S chargers for the chemistry, from the car. That’s something like $3500 in batteries if I were to drop some cake for it.

Or two boxes of cells. Come on EVT, you know you want to donate some to The Cause.

In order to resolve the “dude, what the hell is 42 volts doing on my frame?” issue, I stripped the entire electrical system down and pretty much rebuilt it. At the same time, I took apart the back end for cleaning. If you’ve never seen the kart’s running gear, here it is.

While rebuilding the electronics, I decided to try to minimize the footprint of the ginormoFET controller. Here’s the dismantling in progress.

And here’s the completion.  By stacking the busbars, I gained a few square inches more… board space? Deck space? Scrap-of-plywood space? Additionally, it was easier to arrange the wiring to suit the rest of the electricals.

I decided to forego fan attachment until heat was determined to be a major problem.

Contactors donated to The Cause by another MITERer. They are rated to switch 100 amps. In theory, they should never be switching current in my electrical system, merely passing it after closing. Switches tend to conduct far more current in a circuit than they can reliably close or open, so I’m not concerned about melting the contactor.

Also included in the deal is a Hugeasspacitor™. 33,000 microfarads of love at 75 volts.

Most everything in place. I mounted all components with short wood screws this time, which saved alot of drilling and… yes, threading of wood. That was such a dumb idea that I only could have done it at 4am.

The DC key switch is a battery cutoff switch, and isn’t intended to actually turn the kart on and off.

The power system has two stages. First is the closing of the battery switch with the contactor still open. A resistor bypasses this contactor and goes straight to the controller, which allows the Hugeasspacitor™ to charge at a reasonable rate.

The reason for this precharge resistor is, without diving too much into Course 6 theory, that capacitors appear as an instantenous zero-resistance  to a sudden step in voltage. You know, like closing a switch really fast. What that means is if I just threw the capacitor onto the battery, there would be a Big Spark as infinity amps tries to flow into the cap at once. The problem is that if the cap is now charged, there is still infinity amps trying to cram into it. Not good for the capacitor, and especially not good for whatever poor switch gets caught in the middle. It’s easy to weld contactors like this.

So the second stage of the power electronics is powering the contactor, which opens a very low resistance path. I can now drive off into the nearest oncoming semi.

A 12 volt DC-DC converter provides the contactor current. This is placed before the resistor but after the switch so I can run accessories without the main controller being powered.

Alright, a furious night of building and debugging without pictures later, and here’s the system ready for a test run. The night was mostly consumed with debugging my dumb hardware PWM generator that I pledged never to make. Because it’s on a breadboard, there were Over 9000™ things that could have been wrong with it.

It ended up that my comparator was dead. After tearing the whole thing apart to discover that, I threw on a new comparator, coated the board in hot glue, threw it on the kart, and called it a night.

The safest way to perform a test run is of course to strap a 36 volt 10AH lithium polymer pack right under your ass.

A beauty shot, if you stretch the definition of “beauty”. Notice the emergency stop button. This is the contactor controller.

Yeah, yeah, testing… It was late and nobody wanted to grab the camera. So LOLrioKart sails down the hallway successfully. In this configuration, I drove it to Swapfest the next day for some lulzy in-field testing.

So I couldn’t resist. I had to put the voltage-leaking Nicads pack on the kart, because there wasn’t another way to get greater than 36 volts. My DC-DC converter, rated for 48 volts, shuts off exactly at 36. What’s the nominal voltage of the lithium test pack? 36. If I floored the throttle, the kart would turn off.

Not exciting.

Before throwing the packs back in, I thoroughly coated the bottom of each pack in rubber sheeting, just in case I missed a spot and metal touched metal while in the basket.

Despite my efforts, there was STILL live voltage present at the frame. How’s about them electrical gremlins?

However, it was a high impedance leak, so I wasn’t worried about shorting something through the frame. I decided to go ahead with this installation.

Scoping out a(nother) PWM generator problem. I went through FET driver chips like crazy, for reasons totally unexplainable; not even the resident EEs could figure out what I was doing wrong.

The voltage leak was ruled out as a cause because the DC-DC converter provides voltage isolation and no component is frame-grounded. That I know of, anyway.

Out-of-range operation was also ruled out, since I’m running the chips at less than half their maximum voltages. The FET gate had a bleed resistor and the throttle input has an RC filter inline.

Transients were the main suspect, so I arranged some more low-value caps around the important parts.

It seemed to be stable. Time to figure out which way the motor is supposed to be hooked up!

This is what we call “Nope, it hooks up the other way”. I like where this is going.

Alright, so maybe not this picture, because “where this is going” was straight backwards into a shelf.

Giving the nicads a wakeup charge. By virtue of sitting for two weeks, some of the cells have fallen back to zero volts.

That’s how totally fucked they are.

Using the HOLYCRAPWHATISTHAT, I dumped 25 amps into the cells, which is pulling something like 1600 watts.    After an hour and so, they were nice and warm.

Totally the most legitimate throttle pedal ever. That’s a spare bettery switch key zip-tied to a bike hand throttle zip-tied to the frame.

Video time! I decided this was legit enough to take on the streets and have some fun, so I drove to campus and found some testing grounds. It had been raining for two weeks at this point, and I couldn’t really hold it in any more.

Afterwards, I decided this pedal was just not legit enough to keep, and that my scooter needed the throttle back really badly, so I fabbed up a pseudopedal. It’s not really any more legitimate.

I had a “resistive throttle box” already, so I just scrounged some parts from it and assembled them onto a mount I cut out long ago. While it’s functional, the positioning is the least ergonomic thing in existence. I kind of have to side-roll my right foot onto it.

Alright, so it’s good for now. Time to start working on the priorities, like…

…instead of, you know, functioning brakes.

Seriously, the little band brakes have deteriorated to the point that stopping doesn’t really take less distance if I completely step on the brake pedal. They really were not made for the task of stopping a 350+ pound vehicle at n miles per hour, where n is between 20 and 30.

And so this is how the kart was set up for its first ever road trip, from MITERS to the extreme western tip of campus, a distance of roughly one mile each way. No video is available due to it being completely spontaneous

I’m clearly still alive. More work to come, like BRAKES.

While I seem to be in “build season” mode year-round, it is during long breaks with little in the way of academic or life obligations that I get the most done. Last summer, I began work on LOLrioKart and built Überclocker, Pop Quiz 2, and Nuclear Kitten for Dragon*Con.

… which sort of sucked horribly for everything. Except NK, but only by about *this* much.

So what’s coming down the projectubes this summer?

Mostly the same thing. D*C is my biggest bot-celebration of the year, so once again the combat robot fleet takes high priority. Since there’s really just one robot that needs rebuilding, I also have the usual pile of small electric vehicle projects, of which only one is actually urgent.

Übercløcker RЭmiχ

I started redesigning Uberclocker some time in the fall of last year, hoping to get it done by Motorama 2009. Of course, due to scheduling concerns and logistics, this didn’t happen. But what that presented me with was the chance to put it away and not look at it for several months.

This is pivotal. The basic design has already been hashed out, but now I get to return to it after not thinking about it for a while. I am now in the process of analyzing the 3d model for any “impossible objects” that I might have included, or Really Bad Ideas. Such design flaws plagued the real life Uberclocker 1.0 at D*C last year.

Planned upgrades from 1.0? Well, besides EVERYTHING, the primary focus is on drivetrain reliability, center of gravity, and the upper clamp arm.

As a member of the pushybot school of combat robotic thought, I value maneuverability and driving above jawesometacular weaponry. Uberclocker 1.0 had a strange serpentine timing belt setup that seemed like a really awesome idea at 5 in the morning, but… wasn’t.

The robot also suffered from “centrally located center of gravity” syndrome at the event. While a CoG near the geometric centroid of the robot is good in practically every other case, the fact that the bot’s sole purpose was to grab another opponent and lift it off the ground meant that it just sort of faceplanted every time I attempted a lift. Not a very impressive show. The redesign lengthened the wheelbase of the bot, and selective weight reduction moved the CoG back about 4 inches, without additional ballast.

Oh, that’s right, Uberclocker 1.0 weighed in at an incredible 22.5 pounds out of 30 at the event. I’ll fix that too.

What I didn’t really get to (properly, anyway) in the redesign was the upper clamp arm. The previous arm was both weak and structurally unsound. While I think I took care of the “unsound”, I still have my doubts as to the clamp mechanism’s effectiveness.  In the past, clampbots have used pneumatics to actuate the upper half of the clamp. This is advantageous because a pneumatic actuator requires no “holding power”, unlike an electric motor, which has to be continually powered to produce torque. Pneumatics also have a certain amount of spring-back ability that a solidly coupled electric actuator doesn’t.

But robot-heaven forbid that I make Überclocker even more complicated by incorporating a pneumatics system for the one actuator that might need it. Thus, I’m still partial to a (spring-coupled) leadscrew-type mechanism, over the current design candidate’s motor-on-a-weird-gear. Except this time it won’t be driven by a beetleweight motor.

I intend to keep the “Chinese puzzle” frame, and will be refining it for ease of assembly. I devoted a few weeks to just fabricating the frame parts last time – no, never again. That’s what computer-controlled machine tools are for.

Pop Quiz 2√2

Incidentally, 2√2 is about 3. Not quite there, which also describes this planned rebuild of Pop Quiz 2. It’s not quite a complete conceptual revision, but there will be significant upgrades all around.

PQ2 is one of the (if not the) flattest 1lb class robots around that has an active weapon. It hits lower than some undercutters. The problem is that going the extra 1/8″ down in this current design meant that I had to ditch practically all the well-known, battle-proven parts – Sanyo gearmotors, SPEKTRUM 2.4ghz receivers, etc.

It was a fun thought experiment come to life, but the robot had a horrific reliability record, almost no reception due to the FM ground-band receiver, and a 5 minute chopped hack of a master power switch that ended up disintegrating after exactly 1 hit at D*C 2008. Pop Quiz had about 15 seconds in the arena.

Not cool. For ’09, I am INCREASING the height of the bot. Me, making a robot taller. How many times does THAT happen?

The robot height will be increased to about .400″, enough to cram in a set of real Sanyo micro gearmotors. The rest of the robot’s electrical system is sound, and so is the weapon motor. I’ll most likely end up reusing the electronics anyway, minus the cheesy little FM park flyer receiver. Instead, it will be swapped out with the latest Spektrum DSM offering, and I will run one transmitter between all the robots.

There’s no current virtual model for PQ2.8284171, but just imagine the current bot 0.025″ thicker.

Nuclear Kitten 5.1 Digital Surround Sound Edition

I’m actually satisfied with the performance of one of my combat ‘bots for once. NK needs very minor rework to take another run at D*C. The weapon motor needs some magnet reglued, and the weapon pod pivot axle is slightly bent and needs to be made better anyway. Past that, I have a spare blade to replace the faceplant-into-steel-bumper bent blade.

The only point of concern with NK is the drivetrain. Despite having a mechanically isolated weapon, I’m still blowing drive gearboxes, just because the bot is that much more powerful. I might switch to something like the 50:1 Copal motors || redesign the motor mount || use softer wheels.

No frame changes are necessary, since the bot escaped D*C rather unscathed.

LOLrioKart

Since I discovered that the main battery pack was leaking voltage all over the place (somehow, through an eighth inch of rubber?), I stripped down the entire electrical system and tested all the batteries. It turns out that the steel casings of the cells are live, something which I’m fairly certain should never be the case. While it’s fairly common for the battery negative terminal to also be the casing, the errant voltages are always somewhere between 0 and 1 volts.

This case voltage doesn’t seem to have negatively affected the cells, but I’m fairly certain it’s the culprit behind stray frame voltages. Somehow.

The focus for LOLrioKart work will be the electrical system. I intend to complete and test the ginormoFET controller and possibly implement dynamic (or regenerative!) braking using the upper leg of the half-bridge. Mechanically, the kart is fine.

Well, except for the brakes, but they’ve always been undersized and insufficient.

Ultimately the goal is to run it for longer than 1 minute on all 54 volts, or the full pack voltage of whatever eventual power system I might come into. I’m heavily considering crating up LOLrioKart and shipping it down when Dragon*Con comes around, so I can drive it in the parade. This could possibly be the worst idea I have ever thought of.

Project RazEr

It’s been hanging on a utility hook since the last controller fire. Everything works and the batteries are still charged, so all I need is a BLDC motor controller. Since everything still technically “works”, I don’t intend to touch the scooter that much, if at all. Any work on it will be replacing the shell of the wheelmotor with something more substantial (and better engineered, and more reversably built).

Time to get crackin’.