Tiny Arduino

Details

Published: Tuesday, 11 August 2015 02:49

Written by Kyle Granat

You know me, I'm not powered by pure whimsy like some people. But I do really enjoy making things small (as evident by my quest for a Tiny Hexapod. So now I've got this crazy idea about making a tiny Arduino. Not an Arduino Mini or a Tinyduino, but a Tiny, to scale arduino.

Of course this is a bit impractical. There's no real benefit of having a smaller scaled Arduino, at lease in the prototyping sense, but it could be fun, so that's enough for me. So for a basis, I'm thinking that I could use 2mm headers instead of 2.54 mm headers .2 .54mm headers are used on a lot of standard prototyping hardware, since it's .1 inch - so everything from breadboards to components use this. 2mm tends to be for a little smaller stuff lie XBees. So just from the headers I've got a 22% size reduction - see the image for a mockup. That puts the total board length at 2.1 inches. If I cut off a little here and there, I could probably get it down to 2 inches. But can I get it smaller?

Digikey has these headers that are 1.27mm pitch - half the size! a Tiny 1.54 inch Arduino would be crazy. But at that point, do I actually have enough room to get all of the electronics on the board? I mean I can't just 'shrink' the circuitry like I did in the mock-up. I have a couple of things on my size (small surface mount chips, smaller usb, the fact that there's a lot of 'empty' room on the Uno) but it's still going to take some work. Not to mention finding eagle parts for those tiny connectors and making sure I can get them in smaller quantities, as they're somewhat less common.

My mom has a STEAM workshop next month, I'm thinking these could be cool badges. at 2 inches across, it's a little more reasonable to use them as actual Arduinos, at 1.54 it might be hard to find the right size pins to connect.

Really, I just need to spend some time in my PCB software, Eagle, and see how compact I can comfortably make the board, then pick my pins from there. Based on the Arduino mini and pro mini, I think I can get pretty small. I figure I'll go with a USB micro connector to keep things small. I might need to lose the ISP connector and just use wires to the side headers for programming the bootloader, but I'm not sure.

More importantly, what will I call it?

My New Felt MP

Details

Published: Monday, 03 August 2015 19:34

Written by Kyle Granat

A couple years ago my Lead Engineer Andrew bought a Felt MP Cruiser off cragislist. I was insanley jealous. I absolutley loved the look of it (much more than the newer version of the Felt MP). He also bought a Skyhawk Bicycle Motor Kit which he installed with the help of our friends over at Mad Lab Industries. The motor worked at one point, but I'm not sure it still had an issue or if my lead engineer just didn't have room for it by the time it was finished. Earlier this year my lead engineer moved to start our second office, and he decided tot to take the bike. Knowing that I loved his bike, he gave it to me!

Over the next couple of months Don helped me to get the bike running again. And by helped, I mean did almost all the work. I'm slowley learnining more about this little 2-storoke engine, but I'm no mechanic. We had to clean out the fuel lines and get it a new spark plug, but finally the bike roared to life!

Until it broke.

The weld where the motor mounted to the bottom frame of the bike failed. I ordered a mounting bracket from King Motor Bikes, hoping it would get me up and running, but I didn't take into account that there was basically no clearance between the exhaust on the bike and the bike frame. I was also having some clearance issues on the bolts coming out from the motor itself. Which just serves as a $20 reminder that I need to measure and actually think about how the peice will fit, not just use the order and pray method.

Don, like a true pro came through with a new weld job. He welded a small flat plate to the bike frame to cover the hole /reinforce the frame. Then he wlesed a standoff to that plate. He used his plasma cutter to make a small plate with three holes, one to bolt into the standoff and two to bolt into the bolts coming off the motor. And just like that I was good to go!

Since then I've had a couple of 'adventures'. When I got it one of the pedals was ruined and I was impatient, so I bought a cheap pair of pedals at target. Never. Again. The pedal cracked right down the middle on an uphill climb. I also managed to lose a master link on my engine chain, but I'll carry a spare from now on just incase.

I kinda had my doubts about the motor when I first heard about it, but it's really a stupid amount of fun. And being able to switch between motor and bike can be really handy when getting around town. For my commute to work I try to use the motor sparingly so that I still get some amount of exercise, but If I'm going across town to pick up some last minute groceries I'll just motor it all the way there.

So what's next? I retro-fitted the headlight with an 1 watt LED but I think I'd prefer something brighter. I don't know if I'll try to keep the original enclosure or not. It would be easy if I wanted to go with a LiPo battery, I just don't know how much mainatnce / circuitry I want to do. Once I figure that out I'd like to try to do break lights and turn signals. Until then I'll just use a light up blinker / reflector.

I'm also looking to add a couple of saddlebags to the bike that I got from the American Science & Surplus store. If everything goes well, I'll have it looking something like this

And just for fun, here's a hasty scan of my bike using the structure sensor.

Say Hello to Heelo

Details

Published: Monday, 27 July 2015 03:43

Written by Kyle Granat

I want to make an adorable robot. I've wanted to make one ever since I saw this video of the robot Mira. So the quest began.

Ideally, I want to make the robot out of RobotGeek parts, as they're plentifully available to me, are low cost, and this robot could eventually become a kit. I decided that I wanted the robot to be able to sit in place, but be able to spin freely, so I started looking at the RobotGeek Continuous turn servo. I'm hoping that it's up to the task - if not I'll need to look at an Pixy camera so that the robot could track color blobs and color codes.

After a couple of terrible sketches, I enlisted the help of Wade to see what we could do. We played around with a couple of options for the tilt mechanism, eventually settling on a 180° RobotGeek Servo. Since we don't have any 'back' brackets, we had to so some clever assembly to get it working. I came up with the idea of using the 4C's idler to attach to the 180° servo, and Wade figured our we could brace two side frames together with a plate. We still need to cut a proper plate, but we used so spare plates from other projects to make a prototype. We

After our initial built, wade made some improvements to the robot, making a better mounting rig for the camera and adding a buzzer. He started working with the Pixy's code and I helped him getting the servo settings right to get some basic color tracking down.

We're tentatively calling the robot Heelo after Kat recommended we call it something that sounded like Heliotropism . We started at Helios, which turned into Helio, which begat Heelo.

We've still got a long way to go. We're evaluating the RobotGeek servos to figure out if they'll give us good enough control to make the robot lifelike. We've also toyed with the idea of using a raspberry pi as a main controller and doing some more advanced vision processing, though I think we want to keep this at a little lower level, at least for now. And then there's the question of shells. We'll probably start by designing and 3d printing some shells, but we don't quite know how the robot will look. Time will tell I guess!

Hexapod Toy

Details

Published: Monday, 20 July 2015 04:00

Written by Kyle Granat

My quest to create a PhantomX Hexapod Toy continues! After a couple of days thinking I had killed my printer (PLA all over teh extruder, no I don't want to talk about it right now but thaks for asking) I've got a couple new prints. I've been playing with super basic joints using rods and holes. You can find the models of my attempt here. I took the femur and tibia from my previous model and separated them. Then I added in a cylinder at the middle of the 'servo' for each part, and a hole big enough for the cylinder on the "F1 Bracket". I figure I'd try this method after I saw how well some of the joints on Kevin's Golsm 3d prints I made worked out.

The holes don't print perfectly, though I wonder if they'll be any better if I use a higher resolution (I only did 200 microns on this print). Also, I could use a file to round out the holes a bit.

This time the rods turned out pretty well, previously I had made them too long. This means there's a lot of sag and when I try to fir the bracket over it I end up breaking the bracket. I'm thinking about thickening the bracket walls up just a little bit.

As it stands, all that's keeping the joint from moving under gravity is the imperfections of the holes and rods. Kat suggested I do something with cotter pins, and I'm also just thinking about striaght gluing the prts together. But If I can figure out some reliable way to get

Gif made with the awesome Gif Rocket app. As an added bonus for this blog, here's a gif that's just full of derp. We had a Quadruoed user load the wrong code onto their robot which resuted in this derptastic IK.

3d Printed Hexapod Shells and More

Details

Published: Monday, 13 July 2015 03:24

Written by Kyle Granat

So I've been having fun with the Buccaneer 3d printer, making all sorts of things. You can see an album of all the random things I've printed here

Renee from the forum posted some really cool Tibia Shells and I had to take a shot at printing them with my Buccaneer and my Glow in the Dark Filament.

I tried one using the Buccaneer's auto supports, but it used a lot of material and was difficult to remove. So I took a shot at printing without supports, and I was actually pretty impressed with the results.

There was an issue with the fr side of the shell, without a support my printer just couldn't make it work. But even with the mess, I managed to clean up the print pretty well.

Since forever I've wanted to have tiny toy models of some of our robots, so one of the te things I've been working on are tiny hexapods! Initially I actually tried printing based on some of our 3d production files for the hexapod just to see how things would work. It turned out...poorly. The thickness of the brackets wasn't enough at the scale I was making, so I had a body and tibias, but no femurs.

Eventually I brought some of the plates (Tibia, body plates) into Tinkercad and just did some basic extrusion. Even this had some issues, so brought the .svg vector files for the plates into illustrator and deleted all the un-needed holes (I left some of the signature holes, but got rid of all the bolt holes and smaller mounting holes). I used the side profiles of F2/F3 brackets and uses some primitive cubes to build some thick brackets. I was actually pretty happy with the simplified hexapod design and it turned out better than I hoped. It was still a pain to get the supports off the model, and the bottom is marred by the supports.

So my next goal in mini hexapod is to make something that is actually joint based. If I can make the body separate from the rest of the robot, then It should cut down on supports and I'll have a movable hexapod. Of course that'a all much easier said than done.

I also printed out some of the models from KevinO's Golem Hexapod. I printed them just to see how they would come out, and I certainly didn't expect them to be great ( I expected a similar experience I had with the PhatontomX Hexapod) but surprisingly they turned out pretty well.

Printing out the parts from Kevin's Hexapod gave me some ideas about how I can make some joints using rods and holes. I started playing with some idead, but I've got a bit to go bfore I've got a working prorotype. If I make the holes to small, then the slop in the printer makes it so the rods won't fit, and if they're too big, there's slop in the joint. That, and if the rods are too long, then they won't fit in the hinges without cracking.

That's all for now., more prints soon!