Project Logs

Collapse

• Keebs

  Tinfoil_Haberdashery • 09/27/2022 at 06:36 • 0 comments

  With the RC wings almost done, I buckled down and designed the keyboard wings.

• Core finished, starting on the wing modules.

  Tinfoil_Haberdashery • 09/24/2022 at 02:58 • 0 comments

  Getting into crunch time, the core of the deck is finished and fully functional.

  Now working on the wing elements. The dovetail connectors and cam locks are functioning beautifully. Material selection is relevant, since ABS layer adhesion is questionable for such a thin-wall part.

  Also, dimensions on the screen hinge are off. Got to reprint in PETG or PLA.

• Bringing it together

  Tinfoil_Haberdashery • 09/17/2022 at 10:39 • 0 comments

  After getting a fair few of the case components printed out, I built the central portion up and set to soldering up the rat's nest of a wiring harness. Narrowing down where everything fits and what's required to keep it there is a bit challenging in a case this small, and I definitely wish I were more confident in my ability to design a PCB to reduce the sheer wire mass and mess in here.

  In any case, the features in this version are crystalizing nicely. We have a QC3 fast charge board charging the 37wh battery, as well as providing power output. Input will come either from a USB C input, or via a 6-36v buck converter. In either case, max power input is 20 watts, not half bad. There will also be ad direct 3.7v port on the exterior, allowing me to use an RC battery charger to push up to 40 watts, much better than the usual 10w you'd get via USB.

  Power supply to the rest of the device will be via two 2A 5V boost converters that are separate from the main charge board. I've found that if you use one of these QC3 boards to output 5v, it limits the input to 5v as well, potentially reducing charging speed if the device is running. By separating the power supply and charge controller boards, I can get full-speed charging all of the time, regardless of load voltage.

  While I've yet to print out any of the modules, having the frame...pretty much  together has let me test a mockup of the dovetail system and it seems very robust! Tolerances also seem quite adequate.

• Chasis Construction and Fitment

  Tinfoil_Haberdashery • 09/12/2022 at 12:26 • 0 comments

  Well, I'd hoped to complete this project with just a membership at the local makerspace, but...3D printing takes a long time, y'all, and having to commute an hour each way by bike and train for every iteration of every part was...time-prohibitive. So, I bought myself a 3D printer. Yes, I own other 3D printers, but...those are on a different continent, which is even more of a commute than the makerspace.

  Anyway, the nice thing about this build is that the components are small, so a relatively inexpensive printer can do the whole thing...in theory.

  fig. 1: the primary chasis
  

  First results show excellent fit of the various pieces, but I'll need to break the aformentioned pieces up even more than I already had in order to prevent improper fit due to overhangs and support material.

  The excellent news is, one of the most vital pieces to fit, the dovetail assembly, does fit! and very nicely...at least when printed in the most optimal orientation, perpendicular to the final print orientation...time will tell if it works when printed vertically. It does seem like a delightfully robust design for a relatively low-profile locking mechanism.

• Wireless Audio and Tiny Toggle Switches

  Tinfoil_Haberdashery • 09/02/2022 at 13:03 • 0 comments

  Using 5.8ghz for video is, surprisingly, easier than using it for audio in a number of ways. Most FPV pilots don't really care much about the audio coming through on their video feed, so most VTXs don't really focus on that. Indeed, most these days only use their audio channel for "SmartAudio" commands from the aircraft's flight controller, and many VTXs (like mine) don't transmit any actual audio coming in via the SmartAudio pad, using it only for digital data.

  My VTX did have a built-in microphone, though, so I desoldered that and used a voltage divider and coupling capacitor to take the headphone signal coming out of the Rpi and drop it down to mic level signal.

  ...which turned out to be pretty unnecesary, somehow. Messing around on my breadboard, I found that with a proper voltage divider the audio was far, far too faint to overcome the noise on the audio channel. Just pumping line level into the mic pad on my VTX worked fine. Even the coupling capacitor didn't seem necessary, though it also didn't seem to make things worse, so I left it in. The resistors are just to help crush the stereo down to the single channel I have at my disposal.

  Fig. 1: messing with the audio circuit on the breadboard

  Anyway, I now have the option of getting audio and video through the wireless signal, plugging headphones into my goggles.

  Another challenge I've run into with this compact build (and even my larger one) is the sheer massive size of commercially available switches. Sure, they're not big for like...a car, and you can get truly tiny SMD switches, but they're never designed to be on the exterior of a device and usually lack any detent or a toggle you can actually grip without the help of a switch part designed into the larger device.
  

  After searching for hours and hours on every site I could think of, I gave up on finding a toggle switch that was low profile and would do what I needed...then saw a USB hub on a website I was shopping that had 7 of them.

  These USB hubs are cheap and ubiquitous. Presumably they just know where to find these switches; I don't, other than "cheap USB hubs".

  Fig. 2: The smallest rocker switch I could find anywhere, next to the ones built into this no-name USB hub.
  

  Fig. 3: I don't know where you came from, tiny switch, but I love you.

• Direct I/O connections for a more compact build.

  Tinfoil_Haberdashery • 08/27/2022 at 07:23 • 0 comments

  One of the difficulties with using the Raspberry Pi in a build like this is that major ports are on adjacent corners, meaning unless your board is on a specific exterior corner, you won't have access to major inputs or outputs. Additionally, lots of the cables required stick out a LONG way from the board. The stock MicroHDMI cable, for example, is difficult to use in a way that doesn't double the effective width of the board.
  

  fig. 1: with strain relief, the micro HDMI nearly doubles the width of the board.
  

  This can partially be dealt with by switching the Raspberry Pi to composite output, a strategy which will indeed be integral to this build but which will be discussed later. Still, things like power input and the TRRS cable that carries composite video are still annoying, not least because while I have 2 displays that take TRRS composite video input...both of them take it in a different conductor pattern than each other, or the pi.

  I will therefore be soldering to the underside of the board for most of the relevant connections, including power, video, audio, and USB2.0.

  Some probing with a continuity meter reveals relevant I/O connections and their associated test pads on the underside of the board. At some point in the future I'd like to design a circuit board that uses pogo pins to distribute signals from these contacts, but for the initial build I'm planning to simply solder wires that serve that purpose.

  Preliminary tests are promising. Hooking up a 5.8ghz VTX to the composite output on the pi lets me get wireless video to my outboard screen.

  
  

View all 6 project logs