Hey everyone! I’m doing a bad job!

I missed one of these monthly updates earlier this year, and I said I was going to stop missing them. Guess what?! I missed July’s update post!

Let’s just jump right into things.

I was a guest on The Create/Invent Podcast again!

I was sad to learn that Max Maker has decided to no longer be a part of The Create/Invent Podcast, but I was excited when Jeremy Cook asked if I would be his guinea pig for his first attempt at a video podcast.

In the past, The Create/Invent Podcast has focused on the people being interviewed. Jeremy has decided to shake things up a bit. Instead of focusing on me and how awesome I may or may not be, he asked me to come on the show to talk about how to get started flying FPV drones.

I’m pretty good at talking about things, especially things I’m interested in or knowledgeable about, so this sounded like a great idea to me.

As of this writing, the podcast hasn’t been published to YouTube. I already have the link, though, but I’m a bit nervous about watching it. I’m hoping I did a good job.

If I didn’t do a good job, though, things will be OK. I used the podcast as an excuse to write a more up-to-date version of my So You Want To Get Started Flying FPV Quadcopters blog post. Things have changed for the better for new pilots since writing the original post last year, so this was definitely a good excuse to knock the cobwebs off!

I got a new 3D printer!

Way back in 2014, I bought a used MakerFarm Prusa i3 3D printer. These were sold as kits, but there was one already assembled listed on Craigslist. It was less than 10 miles away, it was a few hundred dollars cheaper than the kit, and I wouldn’t have to wait for shipping! I also didn’t have to choose a hot end, filament diameter, or even which printer to buy!

The MakerFarm printer was poorly assembled, so it needed some work, but it has been an amazing workhorse of a printer for the past six years.

The connector for the MakerFarm’s heated bed melted in May. It is an easy fix, but this seemed like a good excuse for an upgrade.

I had my eye on the Prusa MINI. It would be a downgrade for me in speed, build volume, and material capabilities. However, it would give me automatic mesh bed leveling, whisper-quiet stepper motor drivers, and a general overall quality of live upgrade for only $350.

The Prusa store said the MINI wouldn’t be shipping until some time in September, but they said the preassembled Prusa MK3S would be shipping in 5 or 6 weeks. I figured paying an extra $750 was worth having a printer two months earlier and all the upgrades that come with the MK3S.

It actually wound up taking 10 weeks to ship, but it is finally here, and it is awesome. The loudest things on my Prusa MK3S are the fans!

I had my first print running through OctoPrint in about 20 minutes. When I got home later that evening, I had my first successful TPU print without much effort too.

Do you know what’s amazing about the Prusa MK3S compared to my old printer? I rarely even open the OctoPrint interface. When I send prints to OctoPrint from PrusaSlicer, I just tick the “start print after upload” box, and everything just works. It hasn’t failed on me yet!

I range-tested my Gotway Tesla electric unicycle

I started with a little InMotion V5F EUC. It was a fantastic little machine to smash up while I learned to ride a unicycle, and being able to get a refurbished unit for $399 was amazing. On my longest journey, I managed to push the V5F to just over 14 miles with about 15% charge left in the tank.

I knew the Gotway Tesla would do better. It has three times as much battery capacity, but its motor is also more than three times as powerful. I’m riding the Tesla a good bit faster. I didn’t expect three times the range, but I expected a big bump! How did it work out?

You can read the blog post about it, but I’ll just tell you right here. The Gotway Tesla made it 34 miles with about 15% left in the tank. I was averaging 2 or 3 mph more than I ever did on my InMotion V5F, and 15% of the Tesla’s 1,020 watt-hour battery is a lot more juice than 15% of the V5F’s 320 watt-hour battery, so I do have more in reserve there.

The Tesla has been a fantastic upgrade, and a terrific value. I hope I never have to push it much farther than 34 miles on a single charge. I’m sure it can do it, but lithium-ion batteries can’t deliver as much power when they’re drained. You don’t want to run a self-balancing vehicle too close to empty, because you might wind up flat on your face!

I made a Starbucks undertow

The undertow is a hidden menu item at Starbucks. I’d never had one before, but I saw Sprometheus’s version of the undertow on YouTube, and it sure looked like the sort of drink I could make at home!

Mine is very similar to Sprometheus’s undertow: 2 oz half and half, 1/2 oz simple sugar, and a double shot of light-roast espresso on top. The layering is tricky for me, but I seem to be catching on quickly.

I made several with some rather boring Ethiopian beans. Those beans make a nice coffee, but there’s nothing special about it. I thought the undertows came out quite nice, though!

I didn’t have any glasses or mugs that were the right size for this drink. You really need to be able to hold the cup at about a 45-degree angle while you drink to make sure you get both hot espresso and cold half and half into your mouth at the same time.

I ordered a set of Sweese 4-oz double-wall glasses. I love my 12-oz Bodum double-wall glasses, but Bodum didn’t seem to have any in this size, so I figured I’d try the cheaper brand! I’ll have to write a post about these at some point. I don’t want to get too far off track here, but these seem to be the perfect size.

The Sweese glasses feel a bit more dense that the Bodum glasses, but that might just be due to their diminutive size. They seem well made, but some of the Sweese glasses have a tiny bump or two in the glass that I’ve never noticed on my Bodum glasses.

I made one undertow with a super fruity Ethiopian yesterday, but I didn’t get to drink it. My wife drank it, and it was gone in about 30 seconds. She said it was much better with the fruitier coffee. I’m excited about trying one!

That’s the bummer about the undertow for me. The calories, fat, and prep time are nearly identical to those of a latte. I can sip on a delicious latte for 15 minutes, while the undertow is gone in three sips!

I’ve been tweeting Team Fortress 2 video clips

This isn’t terribly exciting, but I find it interesting! I’ve made a few attempts to stream out almost nightly games of Team Fortress 2 on YouTube. I don’t expect anyone to tune in and watch us play for hours on end, but I figured it was a way to coax some of our other friends into joining us.

When I stream, I save a local copy. At that time, I was saving them at 720p with a reasonable bitrate, and 3 hours of footage was only 4 or 5 gigabytes. That’s less than one FPV flight on the GoPro! I’ve bumped it up to 1080p and increased the bitrate accordingly, and I’m only up to 9 gigabytes per night.

Now I just have to figure out what snippets are worth posting to Twitter!

I just cut and flew the latest revision of my Falcon FPV freestyle frame

My 3” Kestrel and 5” Falcon FPV freestyle drone frames are open-source, but I don’t like to push updated code to GitLab as I’m working on it. So much of what I’m doing is experimentation, and I don’t like making the code available until I’ve cut the parts and I get to see them go together on my own.

I’ve been flying my last Hyperlite Flowride and a sort of hybrid version of the Falcon for a while. The hybrid Falcon is the original prototype fuselage with an updated set of arms. They have the same shape and upgrades as my new frame’s arms, but the positions of the mounting points have moved quite a bit, so they aren’t actually compatible.

A month or so ago, the 4-in-1 ESC in my Falcon prototype exploded, and in that time, two of the arms on my Flowride have been getting softer and softer. I’ve been avoiding cutting a fresh frame for the Flowride hardware because I have some new ideas I want to try out in the design. I couldn’t wait any longer, though, so I just cut what I had ready!

The unique part of my Falcon design is the arm-mounting system. Most frames use at least two screws per arm. I’m doing my best to use only one screw per arm with a carbon fiber connector piece tying everything together in the center. It is working pretty well on the first Falcon prototype, but it is a bit less rigid than I would like, so I decided to add a fifth screw hole in the center of the frame to hold the dog bone in place.

This is significantly more rigid. Unfortunately, my adjustments to the arm geometry goofed things up a bit. You’re supposed to be able to remove a single screw to swap out a broken arm, but there is no way to wiggle an arm free with the new layout. The arm is stuck between the dog bone and one of the stack screws!

I’ve beefed up the arms a bit. I’ve beefed up the top and bottom plates quite a bit. I think I’ve overdone it. Between adding an extra stack in the rear of the frame and all this extra carbon for strength, it looks like I’ve added at least 20 or 25 grams to the weight of this thing!

This is the fattest quad I’ve ever flown. With the battery and GoPro HERO7 Black, I am up to 730 grams! A few other changes besides the frame have added additional weight too.

The frame design is going to need to go on a diet for the next iteration!

I made a corner bracket hold-down thing for my Shapeoko CNC

I’ve been wanting to optimize my cutting workflow on the Shapeoko for a while. I mostly cut carbon fiber plates, and the thickest plates I work with are currently 4 mm thick.

I use wood screws and washers to secure the carbon fiber to the wasteboard. The carbon fiber plates have their fibers rotated 90 degrees with each layer. This is important, because I need the long fibers to run the entire length of the arms of my drones.

I do my best to line up the carbon plates with my machine, but I’m sure I’m off by a bit every time, and lining it up correctly takes extra time. I want the alignment to happen automatically.

My plan was to cut strips of wood to screw down to the wasteboard. I wanted those strips to run the entire length of the wasteboard. My plan was to cut a chamfer on the underside of the wood. That way I could just push the plate into the corner, and the chamfer would hold my work down on two sides.

Usually you would use your CNC to make those edge pieces square, but I couldn’t come up with an easy way to assure that the chamfer would be square. I figured my new Prusa MK3S could bail me out here!

I designed and printed a 7 mm thick corner bracket. It is reasonably square, and it has a big notch taken out of the corner to make sure the bracket never gets in the way of the material.

I used the CNC to bore the mounting holes in the wasteboard, so the bracket should be quite close to being square to the tool on the X and Y axes.

It is working well on the thickest plates. I drilled a hole in the opposite corner of a 4 mm plate, and my bracket and a single screw kept everything in place.

We’ll see how it work out on thinner sheets of carbon. They’re not as rigid, and they tend to flex a bit when the tool retracts if there isn’t a screw holding the material down nearby.

The OoberLights software is making progress!

I feel like I need to include bits of the elevator pitch for the OoberLights every time I talk about them, because I don’t truly expect anyone to know they exist yet!

The OoberLights are made up of concentric rings of RGB LEDs. The prototypes have two sets of LED rings, but we plan on including a third ring in the finished product. The board is sized to fit in a 5.25” optical drive bay, and your computer or server can communicate with the OoberLights via USB or WiFi.

The idea here is to have a cool blinkenlights display on your computer. It will be great for displaying things like disk utilization, network throughput, and other things of that nature.

The prototype hardware is working. We are able to update the firmware via WiFi. We have basic control of the various animations and status displays via USB now. We also have the beginnings of a compositor up and running, so we should now be able to layer animations on top of each other.

We’re almost at the point where a simple guy like me can start creating new animations. I feel like we’re almost to the point where I can send prototypes to people I know, but our developer wants to iron out some rough edges first.

We are miles away from fully utilizing the hardware that we have, but we are mere inches away from reaching the point where we have the product I originally envisioned. I am extremely excited about this, because it means we’re also inches away from ordering the next batch of prototype hardware, and there is a chance of those prototypes being the finished product is extremely high!

Of course I’ll be excited when we have a super fancy web interface, and dozens or maybe even hundreds of animations to choose from, but once we can indicate and update disk usage and network speed over USB, we will have an actual product in our hands that we can start selling. Isn’t that exciting?!

What’s up for the next month?!

This is a good question. I’m not sure I even know myself. At this point, it seems like more of the same stuff, except I’ll be doing it better. I hope!

I’m working on faster printing profiles for the Prusa MK3S. I’m already approaching the speeds I was getting on my MakerFarm printer while still maintaining most of the MK3S’s excellent print quality. There’s probably a bit more performance to squeeze out, but I’m sure I’m most of the way there. I’d like to share the process and final profiles with everyone.

With the replacement 3D printer here, I can finally cut some more carbon fiber ducts for my Tindie store. It is going to be quite hot over the weekend, so I don’t expect to start cutting until sometime next week.

I have a handful of mouse-sized micro switches here. I have an idea for some thumb switches for a custom keyboard. I’m not quite sure how I want to tackle this design, but I have a few ideas. Maybe I’ll have the first prototype printed before next month!

What do you think? How am I doing? Did I choose the right 3D printer? Should I keep playing Team Fortress 2? Should I be more careful when I send my miniquads up on their maiden flights? Let me know in the comments, or stop by the Butter, What?! Discord server to chat with me about it!