HD Video And Telemetry Link Uses Standard WiFi Hardware

December 27, 2018 by Donald Papp 3 Comments

[GlytchTech] decided to implement his own Digital Data Link (DDL) for his drone experiments, and by using a Raspberry Pi Zero and some open-source software, he succeeded in creating a mostly self-contained system that delivers HD video and telemetry using an Android phone as a display.

USB tethered Android phone used as a display and touch interface.

The link uses standard WiFi hardware in a slightly unusual way to create a digital data link that acts more like an analog system, with a preference for delivering low latency video and a graceful drop-off when signal quality gets poor. A Raspberry Pi Zero, Alfa NEH WiFi card, external antenna, battery, and a 3D printed enclosure result in a self-contained unit. Two are needed: one for each end of the link. One unit goes on the drone and interfaces to the flight controller, and the other is for the ground station.

A companion android app allows for just about any old Android phone to serve as video feed, on-screen display of telemetry data, and touchscreen interface.

The software is DroneBridge (GitHub repository) and it implements Wifibroadcast which uses WiFi radios, but without the usual WiFi functionality. A Raspberry Pi is the usual platform, but there’s also an ESP32 port. The software is capable of even more, but so far suits [GlytchTech]’s needs just fine, and he was able to refine his original Watch_Dogs-inspired hacking drone with it.

Cortex 2 Is One Serious 3D Printed Experimental Rocket

December 25, 2018 by Donald Papp 9 Comments

Rocketry is wild, and [Foaly] is sharing build and design details of the Cortex 2 mini rocket which is entirely 3D printed. Don’t let that fool you into thinking it is in any way a gimmick; the Cortex 2 is a serious piece of engineering with some fascinating development.

Cortex 1 was launched as part of C’Space, an event allowing students to launch experimental rockets. Stuffed with sensors and entirely 3D printed, Cortex 1 flew well, but the parachute failed to deploy mainly due to an imperfectly bonded assembly. The hatch was recovered, but the rocket was lost. Lessons were learned, and Cortex 2 was drafted up before the end of the event.

Some of the changes included tweaking the shape and reducing weight, and the refinements also led to reducing the number of fins from four to three. The fins for Cortex 2 are also reinforced with carbon fiber inserts and are bolted on to the main body.

Here’s an interesting details: apparently keeping the original fins would result in a rocket that was “overstable”. We didn’t really realize that was a thing. The results of overstabilizing are similar to a PID loop where gain is too high, and overcorrection results in oscillations instead of a nice stable trajectory.

Cortex 2 uses a different rocket motor from its predecessor, which led to another interesting design issue. The new motor is similar to hobby solid rocket motors where a small explosive charge at the top of the motor blows some time after the fuel is gone. This charge is meant to eject a parachute, but the Cortex 2 is not designed to use this method, and so the gasses must be vented. [Foaly] was understandably not enthusiastic about venting hot gasses through the mostly-PLA rocket body. Instead, a cylindrical cartridge was designed that both encases the motor and redirects any gasses from the explosive charge out the rear of the rocket. That cartridge was SLA printed out of what looks to us like Formlabs’ High Tempurature Resin.

Finally, to address the reasons Cortex 1 crashed, the hatch and parachute were redesigned for better reliability. A servo takes care of activating the system, and a couple of reverse-polarity magnets assist in ensuring the hatch blows clear. There’s even a small servo that takes care of retracting the launch guide.

The rocket is only half built so far, but looks absolutely fantastic and we can’t wait to see more. It’s clear [Foaly] has a lot of experience and knowledge. After all, [Foaly] did convert a Makerbot printer into a CNC circuitboard engraver.

Lenses For DIY Augmented Reality Will Get A Bit Less Unobtainable

December 25, 2018 by Donald Papp 13 Comments

You may remember that earlier this year Leap Motion revealed Project North Star, a kind of open-source reference design for an Augmented Reality (AR) headset. While it’s not destined to make high scores in the fashion department, it aims to be hacker-friendly and boasts a large field of view. There’s also an attractive element of “what you see is what you get” when it comes to the displays and optical design, which is a good thing for hackability. Instead of everything residing in a black box, the system uses two forward-facing displays (one for each eye) whose images are bounced off curved reflective lenses. These are essentially semitransparent mirrors which focus the images properly while also allowing the wearer to see both the displays and the outside world at the same time. This co-existence of both virtual and real-world visuals are a hallmark of Augmented Reality.

A serious setback to the aspiring AR hacker has been the fact that while the design is open, the lenses absolutely are not off the shelf components. [Smart Prototyping] aims to change that, and recently announced in a blog post that they will be offering Project North Star-compatible reflective lenses. They’re in the final stages of approving manufacture, and listed pre-orders for the lenses in their store along with downloadable 3D models for frames.

When Leap Motion first announced their open-source AR headset, we examined the intruiguing specifications and the design has since been published to GitHub. At the time, we did note that the only option for the special lenses seemed to be to CNC them and then spring for a custom reflective coating.

If the lenses become affordable and mass-produced, that would make the design much more accessible. In addition, anyone wanting to do their own experiments with near-eye displays or HUDs would be able to use the frame and lenses as a basis for their own work, and that’s wonderful.

Epoxy Too Thin? Use Wood Flour As A Thixotropic Filler

December 23, 2018 by Donald Papp 27 Comments

The world of glues is wide and varied, and it pays to use the right glue for the job. When [Eric] needed to stick a wide and flat 3D printed mount onto the back of a PCB that had been weatherproofed with an uneven epoxy coating, he needed a gap-filling adhesive that would bond to both surfaces. It seemed like a job for the hot glue gun, but the surface was a bit larger than [Eric] was comfortable using with hot glue for. The larger the surface to be glued, the harder it is to do the whole thing before hot glue cools too much to bond properly.

What [Eric] really wanted to use was a high quality two-part epoxy that he already had on hand, but the stuff was too runny to work properly for this application. His solution was to thicken it with a thixotropic filler, which yields a mixture that is akin to peanut butter: sticky, easily spread to where it’s needed, but otherwise stays in place without dripping or sagging and doesn’t affect bonding.

3D printed pad stuck to back of PCB with thickened epoxy.

Common thixotropic fillers include ground silica or plastic fibers, but [Eric]’s choice was wood flour. Wood flour is really just very fine sawdust, and easily obtained from the bag on his orbital sander. Simply mix up a batch of thin two-part epoxy and stir in some wood flour until the sticky mixture holds its shape. Apply as needed, and allow it to cure.

Thanks to this, [Eric] was able to securely glue a 3D printed pad to the back of his animated LED snowflakes to help mount them in tricky spots. Whether for small projects or huge installations, LEDs, PCBs, and snowflakes are a good combination.

Manual 3D Digitizer Works A Bit Like 3-Dimensional Measuring Tape

December 21, 2018 by Donald Papp 26 Comments

Digitizing an object usually means firing up a CAD program and keeping the calipers handy, or using a 3D scanner to create a point cloud representing an object’s surfaces. [Dzl] took an entirely different approach with his DIY manual 3D digitizer, a laser-cut and 3D printed assembly that uses rotary encoders to create a turntable with an articulated “probe arm” attached.

Each joint of the arm is also an encoder, and by reading the encoder values and applying a bit of trigonometry, the relative position of the arm’s tip can be known at all times. Manually moving the tip of the arm from point to point on an object therefore creates measurements of that object. [Dzl] successfully created a prototype to test the idea, and the project files are available on GitHub.

We remember the earlier version of this project and it’s great to see how it’s been updated with improvements like the addition of a turntable with an encoder. DIY 3D digitizing takes all kinds of approaches, and one example was this unit that used four Raspberry Pi Zeros and four cameras to generate high quality 3D scans.

Vintage IBM 1403 Printer Problem Evades An Easy Fix

December 20, 2018 by Donald Papp 15 Comments

The Computer History Museum in Mountain View has two operational IBM 1401 mainframes, which use IBM 1403 high-speed printers. They aren’t some decades-old notion of “high speed” that barely looks sluggish today, either. These monsters slam out ten lines per second thanks to a rotating chain of type slugs and an array of electromagnetic hammers. Every 11.1 microseconds, a character in the chain would be lined up with a hammer, and if the control circuitry identified it as a character that needed to be printed, the hammer behind the paper would drive the paper into the print ribbon and the slug, putting an imprint of the character onto the paper. When one of these printers failed with a sync error, it kicked off some serious troubleshooting to diagnose the problem.

The IBM 1403’s type chain has a repeating set of characters that spins around at high speed. Unlike a typewriter or label maker, the hammers are not inside this unit. The hammers are on the outside, and work by pressing the paper onto the type slugs as the required characters line up.

Investigation of the problem ultimately led to an intermittent connection in a driver card due to a broken PCB trace, but by then some fuses had been blown as well. In the end the printer was brought back online, but possibly with a slightly damaged coil on one of the hammers.

[Ken]’s writeup on the repair process is highly detailed and walks through the kind of troubleshooting and repairs involved when solving problems with vintage electronics. Electrical fundamentals might be the same, but a deep understanding of not only the architecture but also the failure modes of vintage hardware is needed in order to troubleshoot effectively.

If IBM 1401 mainframes and fixing 1403 printers sounds familiar, it’s because a printer fix has been done before. That was due to a different problem, but still a challenging task to narrow down and fix.

Behold The WT-220: A ‘Clever’ VT-220 Terminal

December 20, 2018 by Donald Papp 27 Comments

[John Whittington] failed to win a bid for an old VT-220 serial terminal on eBay, so he decided to make his own version and improve it along the way. The result is the Whitterm-220 (or WT-220) which has at its core a Raspberry Pi and is therefore capable of more than just acting as a ‘dumb’ serial terminal.

Rear of the WT-220 with paint-filled laser engraving and all necessary connectors.

The enclosure is made from stacked panels of laser-cut plywood with an acrylic plate on the back for labels and connectors, where [John] worked paint into the label engravings before peeling off the acrylic’s protective film. By applying paint after laser-engraving but before peeling off the film, it acts as a fill and really makes the text pop.

Near the front, one layer of clear acrylic among the plywood layers acts as a light guide and serves as a power indicator, also doing double duty as TX/RX activity lights. When power is on, that layer glows, serving as an attractive indicator that doesn’t interfere with looking at the screen. When data is sent or received, a simple buffer circuit tied to the serial lines lights up LEDs to show TX or RX activity, with the ability to enable or disable this functionality by toggling a GPIO pin. A video overview is embedded below, where you can see the unit in action.

Continue reading “Behold The WT-220: A ‘Clever’ VT-220 Terminal” →