Archive for July, 2010


Hub Motors on Everything, Part II: RazEr rEVolution

Jul 24, 2010 in Project Build Reports, RazEr rEVolution

One of the compliments that RazEr always gets is that the vehicle is so freakin’ small. Its outline is the very image of a stock Razor (™,®,©, what-have-you) scooter, and most people do not notice the hub motor until I show them. They tend to try to look under it or behind the rear wheel to see where the motor is. The other reaction I get is more along the lines of what the hell was that, aroused when I fly past a group of unsuspecting and well-meaning pedestrians.

Novelty value aside, the size of RazEr is also one of its worst flaws, and the biggest engineering headache. The stock Razor A3 frame is just small enough to not really fit anything. The latest iteration of the scooter actually featured a fully custom underside so I could put meaningful amounts of battery in. The addition of this drop-frame compromised the stuctural integrity of the scooter significantly. The motor width is constrained to fit between the existing forks, which puts a limit on how much torque it can physically make, as well as makes it a pain to service or remove.

Given that, what would I change if I could start from a blank Inventor file and design the scooter frame from the ground up, while maintaining the pudgy kick scooter look as much as I could?

Before I answer that, let’s go back in time a bit to see what necessitates this total redesign. In April, I “rebooted” RazEr from a state of suspended animation (and total motor wreckage). The 2nd iteration hub motor was pretty much totally destroyed, save for the stator, which I recycled back into version 3, the present one.

During what was supposed to be a routine test ride just a few days later, this happened.


At the time, I was purposely riding over rough cobblestone to see what would break first. I had expected that perhaps the bearings would crunch again, or I’d vibrate a magnet loose. It turns out that the whole motor shifted on its bearings (which means I did an excellent job making sure that bore was a press fit…), and the windings started running into the conductive aluminum side plates. This grinding caused several windings to short and burn out.

The arcing marks and smoke residue on the side plate is a giveaway as to the failure mode.

Afterwards, I hastily rebuilt the core using a spare harvested copier motor that had a stack 5mm shorter than the magnets in the can (-20% torque). I was out of 22 gauge wire, and only had 20 gauge, so I rewound the core using dual 20 gauge wire. The problem was that I could only fit maybe 18 turns on each tooth, which was significantly less than the 25 turns per tooth before – another loss of 33% or so. All things considered, this rebuilt core was complete shit.

Then I discovered during testing that I wired up the B phase backwards. So, I had to cut the termination and jump wires around until that phase was correctly hooked up again. The result after all this turmoil was a motor which looked like this:


Okay, so it actually worked just fine, despite being terrifying. The loss of torque was definitely felt during rides, when acceleration was weak and current consumption was much higher than before. The controller’s runup to full motor start took longer and was less consistent. Things got hotter.

I was pretty unsatisfied, but that’s how RazEr’s been for all of May and June.

RazEr rEVolution

Enough was enough. I’ve been brooding over the build of a fully custom scooter for a while now. It would address all the volume and rigidity issues of stock kick scooters, while fitting my custom designed components with no compromises. Specifically, tt would carry a much torquier hub motor and enough A123 DeWalt drill cells to match or exceed the watt hour capacity of the existing 5Ah lithium polymer packs. It would be more heavily built and stiffer.

Yet I’ve always put the idea aside when it came up because I figured it was more building than I actually wanted to do.

Well, forget that now, because I’m sick of bumming around with toys. And after building the \M/ETALPAXXX, I’m a little less averse to the idea of making more battery packs.

Let’s begin. I’ve titled this rebuild RazEr rEVolution because….


I always start with a very basic flat geometry study. Here’s the new frame at a pretty early stage. It’s still box shaped like the stock Razor A3 frame, and even has the front chamfer.

But what’s less visible just from a screenshot is the increased dimensions. The underbody is a full inch wider than a stock A3. The whole body, deck included is 2 inches thick (compared to around 0.75″ stock, and the 1.75″ depth of the appended underframe of RazEr).  This whole frame section measure 2 feet long from front to back, too.

Oh, and the wheelie bar is taller and more integrated. That’s important too, but basically, it’s what the A3 would be if it actually filled its outline instead of shrinking away from it in a cost-cutting induced anorexia.

Note the cutout in the top plate of the frame – this is to ensure available space to fit the 12S2P LiNP battery pack, which will clock in at a cool 38 volts and 4.6Ah.

Now I get to the fun part – laying out where the joints of the puzzle will go. I’m going to continue on my spree of tabs-slots-and-T-nuts construction, which ought to make this thing come off fast on an Awesomejet. The t-nut slots haven’t been placed yet in this picture.

I’m only designing the business half of the scooter. The front end will be a stock A3 steering neck and folding joint, since that has worked fine in RazEr from the very start.

One of the issues with mating your own back ends to existing front ends is the rigidity of the joint in the middle. Razor scooters are generally pretty well designed in this area, but once you append your own design to it, all bets are off.

To promote structural integrity, I designed in this 1/4″ nut plate in the dead space in front of the batteries. It has the stock folding joint bolt pattern in it, and is otherwise physically interlocked to the frame.

To address the whole “Hey, you’ll be stepping on your battery packs with that big cutout in the middle there!” problem, I’m going to overlay the frame with a nonstructural cover. The cover is shown rendered in carbon fiber, but that’s alot of fucking carbon fiber and I’ll most likely take the cheap way out of either 1) Garolite or 2) polycarbonate. Or the same sheet of 1/8″ aluminum if I’m that lazy.

Here’s the  side text, modified after insertion to be “waterjet friendly”. Closed R’s, A’s, and O’s just make for a n00bly mess when cut out.

The “EV” in “revolution” is cleverly capitalized to be clever.

new rimz… again

So the version 3 hub motor looks kind of cheesy in the screenshot of the new frame. It appears atrophied and undersized, kind of like putting 15″ wheels on a luxury SUV.

With the new frame width (three whole inches of between-the-forks space!), I’m definitely going to design a new motor to use it.

Here’s the tire!

It’s a 5 inch diameter, 2 inch wide polyurethane-tread, aluminum-core fork truck wheel from McMaster-Carr. Did I mention it’s 2 inches wide?

Originally a wheel investigation for BWD, it will now be used with the new new new! RazEr motor:

It looks like the V3 motor if I pulled on the ends really hard, and essentially it is. The one exception is the flanged endcap, which was a design investigation.

The stator will be 3 (count’em: THREE) stacked 67.5mm diameter, 17mm thick copier stators that I binged off Ebay a few months ago. The total lamination height is designed to be 46 +/- 1 mm, to account for lost laminations during the joining process.

That looks so much better. It fills the space like a giant custom hub motor should do. Now I’m talking like

Holy crap, what if this thing had HUB MOTORS in those wheels?

Of course, one of the hazards of designing your motor before you actually get ahold of the wheel to measure it is


The tread depth turned out to be shallow enough on the new wheel such that if I cored the thing out all the way to the rim (past the relatively thin web center), it would just fall right through my motor!

The wheel could be bored to a diameter of 3.6 inches, but the motor is only 3.25 inches at the threaded portion, and 3.5 inches at the flange. I designed it to use my existing Deathrunner and V3 motor stock, which is a 3.5″ steel pipe.


So here’s the 5AM hackaround… just shove that motor inside a plushy, shiny aluminum case. I ordered a 4″ diameter, 0.5″ wall aluminum tube to make the locking ring from anyway, and decided against ordering a huge 4″ steel pipe just to make the motor casing from. It would be unnecessarily heavy and would have taken very long to machine.

Instead, I could just chop a small section of the pipe I  had already and make it a magnet ring. From the 4″ diameter tubing would come the motor casing and side plates, as well as the threaded ring. The magnet assembly can then press or adhesive-fit into the casing.

The magnets were big enough this time (2 full inches wide, of course) such that it was out of the range of SuperdupermagnetGeorge‘s stock. I cruised the intertubules for a while, and settled on Applied Magnetics and their 2″ x 0.5″ x 0.125″ N42 chunks. A Gobrushless calculation shows that these are an optimal fit.

With the side plates, tire, and threaded ring loaded in, the whole thing looks ridiculous.

I went back to the cross-drilled holes method due to familiarity. I had 17 cap screw on that flange at one point in time (because more cap screws makes a project more hardcore), but fortunately returned to sanity just long enough to realize that I would only have the patience to machine 9 of them.

So there are nine.

And now the whole assembly slammed onto the chassis model. I’m essentially going to be riding a small steamroller.

preliminary calculation

Let’s plat the nibbler game with the most recent dimensions. It’s a crude, unsound, faulty, but convenient way to ballpark the torque production of a BLDC motor. The dimensions are:

  • N, turns per tooth. I’ll use the (v2, working) RazErmotor number of 25 turns per tooth. Assuming I wind all the teeth, it means 8 are active at any point in time. Each tooth contributes two runs through the magnetic field, so the total number of “active passes” is 8 * 2 * N, or 400.
  • B, the magnetic flux density in the motor, in Tesla. Essentially “strength of magnets”. N42 magnets have a surface Br rating of 13,000 Gauss (1.3T). Add in the airgap factor t / (t + g) where t is the magnet’s thickness (1/8″, i.e. 3mm) and g is the airgap length (designed to be 0.3mm, but I more realistically shoot for 0.5mm), and the reuslt is Br * [t / (t+g)] for 1.1T in the airgap.
  • L, the active magnetic length of the motor, which is the stator height: 45mm, or 0.045 meters.
  • R, the active radius of the stator, which is 33.75mm, or .03375 meters.

The motor constant Kt is then really roughly 400 * 1.1 * 0.045 * .03375 = 0.668 Nm/A. That’s almost absurd, and is probably 33% or so higher (via past experimentation) than what it will end up being.

Still, that’s insane. RazEr’s v2 stator had a measured (not roughed out) Kt of 0.25 Nm/A. By direct scaling of the stator length, i.e. if I stretched that stator out to 45mm and kept all other parameters the same, it would have a Kt of 0.45 Nm/A. I suspect this is a more realistic number.

when do i get to ride it???

I ordered everything yesterday, scheduled for easy-peasy laid-back ground shipping. This means I can probably start plugging on the motor right after Otakon concludes (and the August robot season begins). Look out for this thing at Dragon*Con in September, since it’s small enough to throw into the same box as the robots.

Knowing me, though, I’ll just blitz the entire frame on Monday and shove the existing RazEr motor in there.

Hub Motors on Everything, Part I: The RazErBlades Contingency Plan

Jul 24, 2010 in Project Build Reports, RazErBlades

Reason #1 to not engineer things at 5 in the morning: You think that putting hub motors on inline skates is actually a good idea.

Reason #2 to not engineer things at 5 in the morning: You forget how many magnets each hub motor needs, and like a total dumbass, only order half the number you need.

Well, guess who is guilty on both counts. Late in June, I put in a reorder of the custom arc-segment magnets that I got for the first two skatemotors from SuperduperfabulousMagnetGeorge. Each motor takes 7 “north” magnets and 7 “south” magnets, where the designations just describe which pole is on the inside face of the arc. So, I ordered 16 magnets in total, 8N and 8S, so I have 2 spares in case I break something.

wait, what?

If you’re keeping track, the left RazErBlade has 2 motors. That means I only ordered enough magnets to make 3 wheel drive skates. By the time I discovered this minor oversight, it was already two weeks ago, so I hurriedly put in an appended order. The custom magnet service has a minimum turnaround time of 3 weeks, and there was (at that point) 3 weeks left until Otakon. Now there is one, and I’ve been informed that my appended order will ship next Thursday.

You know, when I leave for the con. Clearly, this was not going to work at all.

And so I deployed the backup, pain-in-the-ass-but-it-would-get-them-moving plan, and dropped some more dimes on a set of rectangular magnets for the left side motors.

Using GoBrushless’ excellent rotor magnet placement calculator, I discovered that SMG’s stock 20mm x 5mm x 2mm magnets were a good fit for the can if I doubled them up side-by-side. They would require some spacing games, but I was used to playing that with RazEr anyway.

And here they are! I got the shipment notice 90 minutes after I entered my order – that’s essentially on par with McFaster-Carr. Due to the miracle of express shipping, they were in my mailbox the day after.

I printed out the generated magnet placer graphic to use as an epoxying guide. Step one is to put in the “keystone” magnets, the first 14 of alternating poles. Trying to jiggle too many magnets next to eachother, I have found, always results in unsatisfactory placement and a dent or two in the workbench from my forehead.

After the first 14 magnets set in each motor, I crammed their complements in next to them. Putting 2 magnets of the same pole orientation next to eachother means they tend to force themselves apart. To combat this, I wedged little plastic spacers into the horizontal gaps as I placed each new magnet.

I made the spacers using a handy-dandy sheet metal notcher tool and some strips of thin unknown plastic.

new rimz

Usage reports from friends who actually are good at skating have told me that the 72-78A durometer scooter wheels are too soft to perform most skating maneuvers effectively, such as sliding or otherwise breaking traction.  So I wasn’t totally crazy when I thought the ‘blades handled like bricks – they actually do!

Solution: Hop online and find some harder compound wheels. I decide to upgrade one step and go to 85A wheels. Finding 100mm wheels was actually pretty difficult, since the vast majority of inlines use smaller wheels such as 72 or 80mm. Then came the issue of filtering those 100mm wheels to find the ones which can be hollowed out to 2.5 inches on the inner diameter, which was a requirement not met by most.

I finally located these K2 wheels on and had a pack rush-shipped (By this point, I think express shipping has almost matched the cost of parts for this project).

These wheels have a glossy, blank white tread and a black plastic core. Very plain, yet functional, and I was impressed by the quality and finish.

No matter, they’re going on the lathe NOW. I made a quick mandrel to grip them by their bores, since the urethane was actually too slippery to grip with the outer diameter chuck jaws. A flying pass with a boring bar severed the spokes from the outer part of the rim.

Well, mostly. The bar broke through at a place that was not the outer diameter of cut, so now I have these spoke stubs to contend with. When the shops with bigger machines open again, I’ll just knock those out by virtue of gripping the wheel’s OD in a bigger lathe.

final preps

This weekend (and extending into next week, likely, due to laziness) I plan to re-engineer the Skatroller to allow for manual activation of the DEC modules’ electric braking. My spare force-sensing resistor will be hidden under the original wrist-forward trigger point such that it will detect two possible states – willful activation of braking and the palms-open-oh-shit-i-am-about-to-die faceplant mitigation position.

Which, mind you, may possibly be mutually coupled.

I’m also going to switch the analog op-amp circuit to an Arduino Nano based solution, because it’s much easier to throw some if() statements at the two FSRs than try to play the AND/OR/MAYBE game with logic gates and linear components.

Did I just advocate the use of software? Doom.

Past that, I’m going to refine the power system of the Skatroller to use a single lithium polymer cell with a Lilypad boost converter unit. This ought to net me much more efficiency and subsequently battery life, as well as avoid stressing out the XBee by running non-spec voltages.

Non-straightjacketed Agito, theoretically coming to an Otakon near you. Because I'm totally going to be able to stay upright while 95% blind, without the use of my arms, and with motors attached to my legs. Yeah.

The \M/ËTÄLPAXXX (a.k.a The Decline and Fall of the Fankart)

Jul 22, 2010 in Fankart!, Project Build Reports, Stuff

There’s been relatively little site activity lately, mostly because I’ve been diddling with minor and other in-progress projects, which will be detailed soon. I have a lower threshold of build report typing with respect to effort, so unless I’ve done something worth writing about or just stacked up enough work, I don’t update the site.

Testing Fankart has shown me that the battery was the limiting reagent. I’ve been using a 10S (37 volt) lithium ion battery from an old electric bicycle kit that we disassembled long ago at the Media Lab. It was clearly designed for long range, but low power operation, since it had all of 18 gauge leads. This was a battery with 6 18650 lithium ion cells in parallel as a single “cell unit”, so I know it can certainly dump more than enough amps to overwhelm the small wiring. In testing, the battery voltage routinely dropped to 30 volts from a full charge of 42, while flowing 60 to 70 amps in bursts.

Lame. I’ve been in a pinch for batteries lately – this same lithium pack, in fact, is swapped between Fankart and whatever else might need 10S at 9Ah, which includes two of my “minor supporting cast” EVs, the minibike and the heap-scooter (so called because it is a heap of shit). I also had another 10S, 9AH lithium polymer, cube-shaped pack, but that has been put out of commission after one of the cells went completely bad.

So what am I to do? Well, for the past year or so, our EV clique has been sitting on top (and besides, and between) a sizeable quantity of lithium iron-phosphate cells (the DeWalt drill batteries that everyone keeps hacking apart) which were originally a donation to the legitimate MIT Electric Vehicle Team (not to an il-legit experimenter like me, of course). I detailed some of this last year back when LOLriokart first gained unwarranted Internet fame. I had a box full from the “sampler plate” the Media Lab and EVT were originally given. It had been parked on top of my EV parts box since then.

What a box of the aforementioned batteries might look like.

I’ve nicked some of them for Cold Arbor and Überclocker, but the thought of making a large pack out of them was still one relatively easy to suppress, since they were known to not be the highest grade of cells. I’d have to make sure they were matched in characteristics (or just not plain dead like 5 were to start with!) before puttting together a nontrivial pack, else I’d end up with LOLrioKart’s never-ending battery tragedy again, except this time with more alkali metals.

But I finally caved. A resource which most DIY engineers would kill for just sitting in a box is totally not me. The usage experience of peers and The Internet At Large has shown that LiNP batteries aren’t nearly as fragile and temperamental as lithium polymer packs. They don’t like to be babied and trickled charged and balanced like lipolies – or at least, careful battery management system design is less critical for casual use. Plus, my awesome charger can charge and balance up to 10 LiNP cells at once anyway. I knew I bought that thing for a reason.

So I took the dive.

This is \M/ËTÄLPAKKK, so-called because it’s TOTALLY \M/ËTÄL.

I had only one design goal with this battery pack – that it must equal or exceed the watt-hours stored in the electric bike pack. 12S LiNP cells is roughly 10S lithium polymer. Each cell is 2.3Ah, so 4 in parallel would be approximately 9.2Ah. But my charger can’t care and feed for a straight 12S pack, and it would also be unwieldy because of the size. Thus, I decided to split it into two 6S4P modules.

Well – two design goals. These modules should also fit in the minibike. You probably see where this is going.

Above is the pack in mid-construction. Each vertical column is 4 cells in parallel acting as a metacell. Cell selection and arrangement was a quasi-scientific process involving the following:

  • Group the cells by open circuit voltage, in clusters of 8. I made 4 “voltage groups” comprising cells which measure 3.30-3.33 volts, 3.34-3.36 volts, 3.37-3.40 volts, and over 3.4 volts. There was only 1 “over 3.4v” cell, and I had to put it aside as an oddball.
  • Put the 8-cell clusters into a custom cell-holding rig which had balance and charge leads, and wait for the charger to do its magic. The cell-holder was originally designed with 8 cells in mind, which is why I used groups of 8.
  • After the cycle was over, divide the cells into subgroups of 4. The voltage differential between the cells in each subgroup were no more than .01 volts. Legit cells are nice.
  • Cram each subgroup together using Automotive Goop and leave to set.
  • Goop 6 subgroups together

I used my favorite interconnects – grounding braid – to solder the cells together. Soldering these cells is a bit of a shady practice, but by using an 80 watt, huge chisel tipped iron, I was able to keep the “dwell time” per cell under 3 seconds, then move on immediately. It’s never good to park a soldering iron on a  battery, and on lithium batteries in particular, excess heat could melt the polymer separator which keeps your cell from shorting itself internally.

Touch and go, touch and go…

A 6 cell pack needs 7 wires on its balance lead. Luckily, MITERS had a ton of little 7 pin headers that fit perfectly into the charger’s balance port. Add a dab of rainbow wire and you have… I don’t know, gender-diverse batteries?

After adding the balance lead, it was time for closure:

I am fond of the 2 liter soda bottle school of battery pack DIY. In this case, I actually had to buy (and subsquently empty out in a productive if unhealthy fashion) three-liter soda bottles.

The bottles are generally blow-molded polyethylene, so they shrink significantly under the influence of a heat gun. After it cools, the bottle material effectively becomes a hard shell for your pack!

As usual, I added foam rubber padding to the top and bottom before shrinking the bottle section around the cells.

And suddenly, TWO \M/ËTÄLPAXXX. In all, this took about 2 afternoons and evenings of effort. Check the dual 10 gauge leads and 8mm bullet connectors for when I need to sink some more amps.

After I finished each pack, they were put on the charger/balancer and left to ripen for a while.

For common sense reasons, I used only female side bullet connectors on the modules.  But if I were to connect them in series, I’d need to join two of those together.

So, I made this bullet-bridge using 2 male side connectors soldered literally back to back.

What’s next?

put it in fankart

I changed out the prop adaptor bolt to one which was drilled straighter. For the longest time now, the hub has a mild but disturbingly visible wobble because the original prop bolt. The second bolt I made was much straighter because of better machining practice. Past that, the electricals remained the same, and the fan was re-taped to the underside of the cart body.

Best run yet! The \M/ËTÄLPAXXX held at 41 volts under 75 amps of discharge, which beats out the old lithium ion pack by far.

Knowing the no-load voltage and the under-load operating point means that I can sort of calculate the system resistance. The batteries dropped from 44.6 volts to 41 volts (3.6V drop) while flowing 75 amps, so the resistance in the system is 48 milliohms. Not bad.

and then the whole thing


I was just barely out of the Plane of Interdiction before the duct shattered into a few large pieces of PVC shrapnel.

The failure was almost instantaneous and occured at about 75% throttle. I never got a RPM reading on the propeller blades, so all I can say is that they sounded “much faster” this time around.

The impact knocked the two props in line from their original 60 degree offset!

And the aftermath from the top.

Theories about the failure range from deformation of the props themselves at high speed such that they hit the duct (which seems to me like that would result in a prop failure too) to the duct hitting a vibration resonance and deforming cyclically enough to slam into the propellors (my favorite) to the duct being deformed by the change in air pressure alone (seems possible, but out of my realm of comfortable guessing).

Or it could be the fact that the whole thing was suspended with tape.

Lots of circular scrape marks around the perimeter and a weird red powder residue… neither prop nor duct is red in color, so what the hell is going on here?

the end of fankart?

Fankart might return later, built with more engineering and aforethought. But for now, the damage is terminal enough for me to finally put it away and start working on more productive and meaningful things again.  Like finding another home for my 12S4P giant ass-battery… I wonder where it could possibly go.

In the mean time, you should friend Fankart on failbook.

Fankart!!! 3: Better than Fankart!!! 2

Jul 16, 2010 in Fankart!, Project Build Reports, Stuff

Fankart has turned out to be a concentrated form of the kind of projects projects I’m used to building. It was originally built as an engineering joke, it’s only been existence for a week, it’s utterly useless, and it’s already more internet-famous than it deserves. I mean, even LOLrioKart took a full year before people finally noticed how utterly useless it was. Then the Internet fame is at least somewhat warranted.

And RazEr has never been Internet-famous because it’s actually somewhat useful..

I don’t know what I keep working Fankart. Probably because the time delays between when I ordered parts and upgrades to when they actually arrived have all expired within the past week, so I had an essentially continuous stream of resources to keep working on the HFFan. In my opinion, the HFFan in its latest iteration (to be detailed, of course) has almost reached the limit of what I feel like building without actually engineering something. The whole point of the HFFan, originally, was to see what I could pitch together given only McMaster, Hobbyking, and laziness. To that end, it has accomplished more than I had planned.

I got a set of 16×10 propellers from Hobbyking along with my latest impulse purchase. The accidental perspective makes it harder to discern that they’re larger than the 13×8 props in the foreground – but rest assured, they’re bigger.

The idea is to cut these back down to ~13 inches in diameter such that the portion left retains most of the original’s steep pitch.

Oh, about that impulse buy… Remember when I was talking about the Hobbyking 80/100 “HKrunner” before? They are so rarely ever in stock that to actually see a positive number on their stock count is like sighting a carbide-tipped parabolic flute unicorn. But last week, it finally happened – by the time I saw it, there were six left.

Along with motors #3 and 4 for the HFFans, I snagged two HKrunners. The shipment at this point was probably a few kilos short of when it had to go ocean mail, and shipping fees came out to $100 alone.

My bank probably just shit itself and wiped using my checking account. But, now I have two HKrunners.

What ever will I do with them?

HKRunners aside, I began on boring out and trimming the propellers.

Problem: Even the largest lathe I had access to at the time couldn’t…er… swing the props, since they were too large in diameter. And so, I had to cut the tips off using a bandsaw first. The line indicates roughly where “13 inch diameter” is.

Then came the drill-to-12mm process on the heavy \m/etal machine.

The fine trimming process was the same, except this time I used a straight-fluted cutter on the highest speed to avoid the up-and-down flapping of the prop blades that had occurred last time.

The straight-fluted cutter in question was actually a reamer.

… and the new props drop right into place!

The tips were trimmed to less than .02″ clearance this time. Legitimately, I mean – last time, I forgot that endmills have diameters, and came up 0.050″ too short!

Here is the HFFan with the new propeller setup and a new lower mounting position! It turns out that the whole thing fits snugly between the uprights on the cart frame, blocked from forward movement by one of the basket spars.

The same fiber tape I had used to retain the duct in arrangements past sees a return here, pulled as tightly as possible. The whole setup is reasonably stiff, despite not looking like it should be.

Hey, at least now the basket is empty and ready to accept groceries!

The rear view.

Note the scrape marks in the duct – it turns out that even though trimming the props to sub-millimeter clearances was done with the best of intentions, the high-tension tape mounting still causes the thin PVC to deform some.

I decided to not play the tension adjustment game and just let the prop break in its own duct by running it at full bore until the skull-grinding noises stopped. Now that‘s engineering.

The heavier pitch, longer chord, and tighter duct are all reflected in the increased power draw. Clearly, the 18 gauge battery leads are the bottleneck in this system – the voltage drooped to 31 volts under the highest load. Should it have held relatively steady, I would probably have seen the 3 kilowatt mark.

steering the fankart

This is failrudder.

After I got sick of retying the front wheel knots (or having them retied), I started throwing together other possible solutions for steering. One of them was this dead-fish-esque rudder, made by zip tying a cut piece of copper-coad fiberglass board to a servo zip tied to the frame.

No, it did not work. At all.

Next, here is differential failbrake. Real airplanes generally use differential braking force to turn using their propellers or jet engines (for single engined planes), so I figured why not try it on fankart?

Well, it would have  probably worked if the brakes weren’t raw servo horns, the servos weren’t mounted with zip ties and tape, and the wheels didn’t have so much slop they could tilt 10 degrees off axis and still rotate. Those are some seriously worn-ass wheel bores.

All things considered, I just decided to retie the damned servo knot to test Fankart 3.

So here it is – the collection of test video from the past two days or so!

Yeah, that didn’t end well. Something about “night time and shadows obscuring the unforgiving curbside of life”.

There was no major (or expensive) damage to anything. The HFFan just flew off, and surprisingly, all the electronics stayed in the basket.  Maybe I’ll throw it back together and actually try testing during the day some time… New concept, I know.

What’s next for Fankart?  I’m not really sure, but one of things I still want to do is get an actual thrust number on the HFFan. If it’s two digits, I’ll consider making more. If not, it’ll become the next high-five machine. I’m pretty stoked by the fact that the latest HFFan could accelerate the whole thing up a roughly 6 or 7 degree ramp faster than the previous could accelerate on flat ground.


Jul 14, 2010 in Fankart!, Project Build Reports, Stuff

Hello Internets. I don’t know why you’re so fascinated with random things I throw in shopping carts (probably because of LOLrioKart), but I think you should check out my projects which have marginally more value to society (debatable, of course), such as RazEr or its ‘blades derivative. Or the battlebots. Or hell, even my dumb glowing safety goggles.

But anyway, here’s Fankart in its latest mid-engine flying configuration.

After the epoxy joint failed, I busted out my all-time favorite adhesive – Automotive Goop – and rebonded the motor mount to the duct. The new joints are substantially stronger and have the benefit of being mildly flexible, such that squeezing the duct no longer caused brittle failures. The battery was moved underneath out of packaging issues, and the fan now sits in the basket, retained by the pressure of being shoved between the wire basket frame, and by a (difficult to see) fiber tape strap that goes underneath the structure. And by virtue of its own thrust.

No videos of this version yet, nor any thrust numbers from the fan itself. I’m going to build a thrust-measuring rig (scientific rigor to be determined) to see what the specs on it are. Two 16×10 props are also on the way, so expect more  ファンカート!!! action later.