Monday, November 14, 2011

Screenshot Assault

So I've been working on two designs simultaneously. I'm trying to finalize both of them so that I can then decide which one to actually follow through with bringing to life, since I'm too poor to do both right now. I have two options: improve an existing design, or do something no one has ever done before (I think).

Option #1:

New suction carrrr!

You may notice that for the time being, I've named it Parkour, because I can't think of anything else and Parkour is a lot better than "suction...vehicle...thing". If anyone thinks of any better names then there's a comments section. Anyways, here's a better idea of what Parkour looks like:

Ohhhh yeah, going for two fans this time. It's still going to be powered by Pololu's micro metal gearmotors, since they've proven they're worth their weight in gold, and will grip everything via Pololu's awesome wheel/tire sets.

So this thing is very very wide because when a vehicle is turning like a tank, it simply performs better the wider it is relative to it's length. As far as the huge 90mm tires are concerned...well, this screenshot says it all.

Parkour is designed to be able to go around the outside edge of corners. The huge tires and substantial ground clearance will also aid it in conquering just about everything else too, of course, from deep grooves in the sides of buildings to interior 90 degree transitions, which suction car v1.0 needed a couple extra angles to conquer.

Besides opening the doors to pretty much any terrain, I'm also working to improve upon all of the other minor little issues with suction car v1.0 that bugged me. To start off, Parkour operates off of one battery.

One. Battery.

Instead of, you know, three. I plan to use one of these. This will give me the same 4S power and twice the capacity, which will be awesome since I'm shooting for at least a 10 minute run-time this time around. Further improvements include actually having enough chassis space for all of the components, and using hacked servo boards to control the drive motors, instead of shitty R/C airplane controllers that decide to stop working sometimes, and won't reverse. So now, besides being able to reverse out of bad situations, I'll be able to actually turn in place, with the outside tires moving forward and the inside tires moving backward, instead of just having the outside tires moving forward and the inside tires doing nothing. Maneuverability was always a bit of an issue with SCv1.0.

Oh, everyone stop for a second and look at how much effort I put into modeling the Pololu 90mm tires/wheels for this project. I don't know how many grooves there are around the real tires, but mine have 210 little grooves all the way around. I feel like a god, knowing my creation so well.


Option #2:

Yes, it's a gun. No, it's not designed to kill people.

...and yes, I have contemplated mounting it on Parkour and calling it one project.

It's basically a baseball pitching machine, except scaled way down and designed to shoot elongated projectiles through a barrel, hopefully at a much higher speed and accuracy. I've actually been toying with this idea for a very very long, years.

Everything's pretty compact in this design and therefore hard to see but I'll try to explain as best I can.

The projectiles are going to be made out of something cheap, readily available, easy to work with, and 1/4" in diameter. Sounds like wooden dowels to me. Here's a screenshot showing the loading device. There's just a tall gravity-fed ammo rack that holds the wooden dowel segments, and at the bottom a servo will work an arm (not in the drawing...yet) that pushes the lowest dowel into the giant spinning wheel which then propels it to deadly high speeds out of the barrel.

Maybe that's a better view. Here you can also see the second standard size servo I have mounted underneath the entire assembly, which will be used for the tilting portion of the pan/tilt mount that I want this gun to be mounted upon.

The reason the projectiles are going to be so long (two inches, actually) is pretty simple. The giant rubber tire that will be propelling them is going to be spinning at somewhere around 40,000rpm as fast as I can get it to go, and no matter how grippy the silicone tire is, it's going to slip. Also, no matter how powerful the motor driving the tire is, it will bog down when it suddenly comes in contact with a projectile that is stationary (being reasonable here...if I drive the wheel with a turbocharged V8 engine, it's pretty safe to say that there will be no bogging, I know). The longer the projectile is, the more time the giant rubber wheel will have time to get a grip on it and accelerate it to an even higher speed. It's kind of the same idea as having a longer barrel on a real gun to allow the explosion going off behind the bullet more time to accelerate that bullet.

That being said, the barrel on this thing actually needs to be the opposite of a real gun...very short; just long enough to make sure the projectile is moving straight. After the wheel is no longer accelerating the dowel segment, the only thing that projectile is doing is slowing down, and will do so at a higher rate if it's traveling through a barrel with friction.

So why'd I only go with one tire to accelerate the projectiles, and not two? Well...I didn't find it necessary for one, since the projectiles are elongated and won't start spinning like a ball would if spun by a single giant tire. Baseball pitching machines need two wheels to be any good because they're pitching...balls. Also I went with one tire for the sake of being compact.

Oh yeah:

On the other side of the giant wheel are ball bearings, so that eliminates the huge amount of friction that would have been present if the projectile was being pressed against the other side of the barrel as the drive-wheel propels it.

I guess that's pretty much it for the wheel gun. When I first began designing it, I had a second drive motor at the end of the barrel that was angled ever so slightly, to put a spin on the projectile, but that turned out to be too much of a pain in the ass. Also, right now it's semi-automatic. I really want it to be fully automatic because that would be awesome but I'm still working on figuring out the best way to make that dream come true.

So...both are still only about 90% finished, but...Parkour, or wheel gun? I can't decide.

Saturday, October 29, 2011


Well I've been a little busy but I'm back, with the promised pictures of my balance project's pretty brass weight that this guy was making for me. I still don't have a video because the ESC/receiver decided to die just recently...I'm not sure what's wrong with it but it's done this before and then magically come back to life later, so maybe I'll get lucky again. But anyways...

It looks so much better than a big roll of lead tire weights, in my opinion. Especially when the backdrop is a toaster oven. It's much heavier too. 14.1oz vs. 8oz.

This is kind of neat too. Alex designed it so that I could epoxy a steel bolt into the carbon rod of the balance car and then screw different brass weights on if needed.

That's all the...presentable material I have for now. I'm working on designing a new and very improved suction car (I'm thinking I'll name it and a kind of weapon that I've been wanting to experiment with. You'll see.

Saturday, October 15, 2011

Giving Gravity a Big Middle Finger Once Again

So I received my laser-cut parts from Ponoko a few days ago, rendering my balance project juuuuust about complete. At a glance, I wasn't as impressed with these parts as I was with the acrylic chassis for my suction car, but considering the material this time is over twice as thick, the parts look pretty awesome. I was too excited to take a picture of them as I was popping them out, but here's the aftermath:

Yes, I know, those two pieces in the middle look like little androids.

After realizing that I might have made the tolerances on these parts a little too tight, followed by having to hammer all the parts together as a result, a little balance vehicle was born.

I'm extremely happy with how the laser-etching looks. I'm going to have to do that with every subsequent project now. Also, this thing is SOLID. I designed it as if it would need to handle drops off of really tall buildings. Oops. Oh, and here are the gears:

So...not as pretty and precise as I would have liked. It's understandable though. This is the thickest material that Ponoko offers for laser cutting. And they will work, they just might have to be broken in a lot.

So at the time I got all the laser cut parts in the mail, I had everything else I needed to put this whole project together. Everything but the drive shaft, that is. You know, the main component that I need to make this thing move. I needed a steel shaft that was 170mm long and 5mm in diameter, but the problem was that I needed to buy it already down to size, since it's really hard to machine something that long down to that small of a diameter without it bending. I couldn't find a 5mm shaft anywhere for a while, but then I remembered that my favorite website ever, SDP-SI, sold shafts in metric sizes. I gave their site a visit and sure enough, I could buy a 180mm long, 5mm diameter stainless steel shaft for about $7, which I was very happy about. So I ordered one, but it didn't arrive until two days after I had everything else! What was I to do in the meantime?

Well...the night I had put everything together and was waiting for my 5mm shaft to arrive, I discovered that Walmart sells knitting needles, and that knitting needles come in metric diameters. I'm sure you can see where this is going.

So here we see a picture of a knitting needle cut to length and used as a drive shaft. Hey, I was desperate to test this thing out! The needle was actually solid aluminum too, which was kind of nice. I was certain it would be hollow when I bought it.

Completely assembled at last! Well, almost. Now all that was left was to wrap a half-pound of lead tire-balancing weights at the very top of the carbon tube.

There we go.

I definitely don't plan to leave these weights wrapped around like this. Mr. Machinist over here is going to make me a nice brass weight to serve the purpose that this huge roll of ugly lead is currently serving. I should have pictures of that up in a couple days.

So...initial test runs were tough. It was a lot harder to balance than I thought it would be. It's not super fast, so once the vehicle tilts more than about 5 degrees from vertical, it's going down. If you mess up a little and don't keep it almost completely vertical, it's game over. This is the reason the bigger 70mm wheels are mounted on it in all the pictures: I needed more speed. Also, a little problem that needed correcting was that the vehicle isn't exactly wide enough to prevent it from rocking from side to side. This results in the thing randomly turning towards some unknown direction very quickly and engaging some unsuspecting audience member in a surprise jousting match. And of course, the laser-cut spur gear does work, but not nearly as smoothly as it needs to. All of these things made balancing a little harder than it should have been.


In true human style, Alex and I were able to adapt to all of these issues and within a few short hours had the vehicle balancing for upwards of 20 seconds. It just took practice, really. The main issue with the vehicle wanting to rock from side to side was somewhat solved with a combination of being really careful not to jerk the throttle too much, driving it on soft carpet, and luck.

Then of course, a couple days later my stainless steel drive shaft arrived. I installed it and, since it was a little longer than I needed, I used that to my advantage and used some extra hubs I had ordered as spacers to widen the whole vehicle by 10mm, to help with the rocking issue.

Bam, done.

In the past couple days I've practiced a little more with this thing and despite how incredibly challenging it was to balance at first, on average I can keep it balanced for about 30 seconds now. One time I had it balanced for about a minute. It just takes a ton of practice, but that's what makes it fun.

I'll have a video posted soon.

Tuesday, October 4, 2011

I'm Alive I Promise

Well it's been a while. I've been busy with uhh...being poor as a result of buying one of these. In case you don't know anything about R/C rock crawlers, or if the name isn't self-explanatory enough, these trucks are freaking beasts and are designed with super low gear ratios (this one is 38:1), super awesome solid axle suspension, and super powers, allowing them to conquer just about any terrain. Exhibit A:

But of course, as with pretty much anything R/C, there are competitions that require rock crawlers to actually crawl...over rocks...ok that was a no-brainer but you know what I mean. They actually have a purpose that matches up with what their name implies, as opposed to just crushing people and making them beg for mercy. I don't know...I think it's pretty neat. Check it out.

Anyways, with that out of the way, I have a new project I've begun work on that was actually conceived right around the time I blew all my money on an Axial XR10, but the new toy actually helped my new project become more of a tangible idea, since I didn't really know if the project was...possible. This needs explaining.
Ok so I always found it fascinating that it's easier to balance something tall if it's...taller, and is heavier at the top, like a yardstick or wooden dowel, as opposed to a pencil or something. It seems like it would be harder to balance a taller object, and it looks more impressive, but those guys in the circus balancing ten-foot poles on their foreheads would have a hell of a time doing the same with a ruler.
After doing no research whatsoever and thinking about it for like 10 seconds, I figured out why taller things are easier to balance. I thought it was kind of cool, so I made a little presentation explaining the phenomenon.

Figure 1

As you can see in Figure 1, we have an object that is 300mm tall, and if it falls all the way to the ground, the very top of that object will have to travel about 471mm.
Figure 2
Now in Figure 2, we can see that whatever lazy asshole drew these pictures simply scaled Figure 1 up to get Figure 2. The object is 1200mm tall, and the very top of that object has to travel about 1885mm before it hits the ground.

Even though the object in Figure 2 is scaled up and looks just like Figure 1, gravity is constant. Therefore, since the object in Figure 2 has a longer arc to fall across, it will take more time to fall to the ground. Since a taller object takes more time to fall, it is easier for a human to balance that object on their hand or head or finger or whatever, because there is more time for our slow reflexes to react to whatever that object is doing.
By the way, I promise I'm going somewhere with this. Stay with me.

My little illustration explains why taller things fall slower, but ignores weight altogether, which is very important to this whole...idea. What actually matters is not just how tall an object is, but where the center of gravity is. My illustration works if both of those objects, let's call them "poles", are of uniform mass everywhere. In that case, both poles have a center of gravity right in the middle of them, which still allows the taller pole to take longer to fall.

Now, if you were to put a lot of weight at the top of the shorter pole in Figure 1, bringing the center of gravity very close to the top, you can get it to take longer to fall to the ground. In addition, if you put a lot of weight close to the bottom of the taller pole, lowering the center of gravity, you can get it to fall much quicker.
So...height and center of gravity are both of equal importance when talking about how easy a tall object is to balance. Optimally, it should be as tall as possible and have as high of a center of gravity as possible.

Anyways, now that we're all good and learned, the project idea that I had in mind that exercised this...concept was a two-wheeled vehicle with a long pole extending out the top, with a weight on top, and the operator must keep it balanced to drive it around. At about the same time I had this idea, I was screwing around with my newly acquired R/C rock crawler. One thing led to another and eventually it ended up having two wooden dowels zip tied together, with a heavy piece of steel zip tied on the end, all zip tied onto the axles so that this big long top-heavy pole stuck straight out the front. I wish I had a picture. This setup allowed the truck to be balanced on end for upwards of 20 seconds and made for a pretty fun game, and ultimately proved that it was possible for a human to balance a vehicle like that.

Ok, so we've made it through physics and history class. Here's what I'm currently working on building:

That carbon fiber rod coming out the top extends out of the frame about an extra 940mm or so, and will have a heavy weight at the very top. The design has actually changed just a little bit, but it wasn't necessary to update in the CAD I didn't. All I changed before I sent the drawings to Ponoko was the orientation of the motor, which will be mounted facing the opposite direction so that the spur gear is closer to one of the bearings, minimizing stress on the drive-shaft, and I added some fancy cutouts to the chassis pieces to make it look awesome reduce weight. So here's exactly what I'll be receiving from Ponoko, cut out of 6.6mm Delrin:

Seriously, Ponoko is the best. I highly recommend their services. I only had to pay $17 for them to cut all of this stuff out. On top of that was $30 for the 6.6mm Delrin but...that's because it's 6.6mm Delrin.

I put an extra gear on there because why the hell not/just in case/I like gears and I wanted another one to show off and play around with. I should have just put like six of them on there. The word "balance" will actually be laser-etched. The laser-etching, along with laser-cut gears and working with delrin at all, are all things I've never done before, so this should be a huge learning experience for me, which is always fun.

You can probably notice that the whole assembly is actually clamped together by four huge standoffs running along the width of the chassis. I'm lucky enough to have an awesome friend that will make those kinds of things for me in exchange for gifts such as chocolate chip cookies.


Moving right along. I got all the Pololu stuff I needed. Wheels, motor, hubs, all that. Even the little Team Losi Micro Rock Crawler ESC I'll be using for this thing. It doubles as a DSM receiver which is why I chose it. It's super compact.

I purchased 60mm and 70mm wheels just because I wasn't completely sure on the gear ratio. This thing needs to be geared down low enough to allow precise adjustments, but it can't be incredibly slow to the point where it can't catch up if it tilts more than 1 degree either. I'll have to mess around with it once I get it all put together. I'm hoping to have my Ponoko order within the next week and a half.

Monday, August 29, 2011

Shenanigans, Light Sabers, etc.

Well with the first week of school happening and me now possessing an R/C vehicle that drives on freaking walls and ceilings, I've been quite distracted. I've been working on designing a little motor-driven pan and tilt mount that holds one of these unsafely powerful lasers light sabers that Alex will be spending way too much money on though. If you don't know already, these lasers cost thousands of dollars because...well, let's just say light saber is literally a better name for them, rather than laser. I'm pretty excited to carve my name into unsuspecting people's arms from afar see it in action. So here's v0.2 of the pan and tilt mount:

Don't ask about v0.1, it was just a bad idea altogether.

So I designed this thing with my new love in mind. It's like 90% laser-cut acrylic, because when your suction car which has an acrylic chassis falls from a 10ft ceiling three times and everything but the chassis breaks (I don't want to talk about it), acrylic suddenly becomes the new steel. I think it will look awesome with clear acrylic anyways...oh speaking of which, I can actually apply that material to parts in 123D.

Instant cloak of invisibility.

The gears will be laser cut from the wonder-plastic that I've been dying to work with, Delrin. You don't even know how excited I am to design my own gears, by the way. Oh and then the little aluminum main mount will be machined by Alex because he loves me (although I guess this whole design isn't really for me anyways), and because these people lied and apparently I can't use the UCF machine shop after all. Oh well.

Then of course the actual pan and tilt functions will be powered by these little gems that I love so very much. This of course gives me an excuse to design some extremely simple motor controllers and maybe learn a little more about those silly little electrons. I ordered some of these incredibly awesome little joysticks partly to control this pan and tilt thing, and partly just because...LOOK AT THEM! You want some too, don't even try to lie. Anyways...looks like I'll be designing my first H-bridges. I've hardly even thought about that yet...I'll get around to it.

Saturday, August 20, 2011

Drum Roll Please

So over the past couple days I've been conducting some research on this new little project of mine that totally sucks, trying to optimize its performance and such. I settled on 40% throttle for the fan as the lowest I'd ever go. That setting works perfectly for walls and ceilings and there is minimal traction loss. At that setting, the car's ability to stick to surfaces is however all based on suction, meaning that if I run over an obstacle too big and the chassis gets raised off of the surface too much, there is not enough thrust to keep the car on that surface until suction can take over again. At 40% throttle, battery life for the fan also seems to be around five to six minutes, which I say is pretty decent. Battery life for the drive motors, which all run off of the tiniest, cutest little 3S lipo ever, is probably like 15 minutes...or something. I haven't even tested it but after a five minute run the 3S pack's voltage only drops about 0.4 volts.

Anyways, now that I'm back at UCF and reunited with this guy, we took some video. And without further ado, here it is, the feature presentation.

Tuesday, August 16, 2011

It Lives!

Well this is exciting.

The Ponoko laser cut chassis has arrived! The final piece to this seemingly never-ending puzzle. It's really nice too. Once again, I definitely recommend laser cutting. Time to put stuff together!

This is so exciting. As a result of the excitement overload, from this point on I kinda didn't take any intermediate pictures. So here's the finished product with everything stuffed onto the chassis.

If you look carefully, you can see the gobs of epoxy holding the ducted fan in place. I just really did not want a brushless fan with 540 grams of thrust coming loose. And now for another view.

I call this one "OH MY GOD WIRES".

Oh umm...anyways, keeping things short and sweet, it works.

Ignore the shirt.

It's extremely loud, like you don't even know. It sounds like a jet engine. It also works on ceilings (although I don't have a picture at this very moment), and the ducted fan is only running at 50% throttle for both wall and ceiling use. It absolutely killed the batteries the first run, lasting for about two minutes total, but I don't even know if they were fully charged. It was just a quick test run to see if it worked. I'm pretty sure I can reduce the throttle to 40% or even 30%, and I'm really hoping to squeeze five minutes of run time out of it.

SO, coming soon to a blog near you (...this one), I'll have test results for battery life when I have the amount of throttle absolutely necessary from the fan tuned in, a video, and ramblings about my ideas for v2.0. I might even actually organize the wiring on this thing. I'll post again within the next few days for sure.