Driving more than a handful of LEDs from a microcontroller is often a feat that takes tedious wiring, tricking the processor, or a lot of extra external hardware. Charlieplexing is perhaps the most notorious of these methods, and checks two of those three boxes. This library for the Teensy 4.0 checks all three, but it can also drive a truly staggering 32,000 LEDs at one time.
The TriantaduoWS2811 library is able to drive 32 channels of LEDs from a Teensy 4.0 using only three pins and minimal processor resources. It uses the FlexIO and DMA subsystems of the i.MX RT1062, the particular ARM processor on the Teensy, to drive four external shift registers. Together, the system is able to achieve 30 frames per second on with 1,000 LEDs per channel, for a total of 32,000 LEDs. Whoah.
[Ward] aka [wramsdell] wondered what one would do with all of the horsepower of a Teensy microcontroller when he first saw its specifications, and was able to build this project to take advantage of its features. What’s surprising, though, is that it doesn’t use nearly everything the processor is capable of, so you can do other tasks at the same time as driving that giant LED display.
[Ben Cox] found some interesting USB devices on eBay. The Epiphan VGA2USB LR accepts VGA video on one end and presents it as a USB webcam-like video signal on the other. Never have to haul a VGA monitor out again? Sounds good to us! The devices are old and abandoned hardware, but they do claim Linux support, so one BUY button mash later and [Ben] was waiting patiently for them in the mail.
But when they did arrive, the devices didn’t enumerate as a USB UVC video device as expected. The vendor has a custom driver, support for which ended in Linux 4.9 — meaning none of [Ben]’s machines would run it. By now [Ben] was curious about how all this worked and began digging, aiming to create a userspace driver for the device. He was successful, and with his usual detail [Ben] explains not only the process he followed to troubleshoot the problem but also how these devices (and his driver) work. Skip to the end of the project page for the summary, but the whole thing is worth a read.
The resulting driver is not optimized, but will do about 7 fps. [Ben] even rigged up a small web server inside the driver to present a simple interface for the video in a pinch. It can even record its output to a video file, which is awfully handy. The code is available on his GitHub repository, so give it a look and maybe head to eBay for a bit of bargain-hunting of your own.
The Axiom motor controller was a winner of the bootstrap contest and is a Finalist in the 2019 Hackaday Prize. The driver aims to deliver 300A continuous at 400V all day long. Which is a very impressive amount of power from a board that appears to be quite compact.
The brains of the device is an ice40 FPGA from Lattice running software based on the VESC Project. Its open source roots will certainly allow for some interesting hacks and an increasingly stable platform over time. Not to mention the existing software tools will aid in the sometimes cumbersome motor-driver tuning process.
The board designs are available, but we agree with the team that the complexity of assembly is likely going to be high (along with the price). The amount of research and skill going into this complicated kit is a bit mind-boggling, but we hope it will really enable some cool hacks, from cars, to ATVs, and maybe even an electric flyer.
When deadlines loom and your future is on the line, do what top college students through the ages have always done: procrastinate! [Simen] and [Amund] did that in grand style by starting a YouTube channel, delightfully and aptly named “Applied Procrastination”, wherein they plan to avoid their responsibilities as long as possible in favor of making a large-scale ferrofluidic display panel. (Video, embedded below.)
We suppose we should encourage them to hit the books, but honestly they look like they’re having much more fun and learning more than they would in class. The idea isn’t new; we’ve seen ferrofluid clocks before, after all. [Amund] and [Simen] have grander plans for their display, but they’re wisely starting small with basic experiments. They had an early great idea to use a double-pane window as a tank for their display, but coatings on the inside of the glass and the aluminum frame conspired to cloud the display. They also did some tests to make sure they can control 252 electromagnets safely. They did manage to get a small test display working, but really the bulk of the video is just them playing with magnets and ferrofluid. And again, we’re OK with that.
It looks like this is going to be an interesting project, with hopefully regular updates to the channel now that summer break is upon us. Unless they find something else to do, of course.
While it might be tempting to start soldering a circuit together once the design looks good on paper, experience tells us that it’s still good to test it out on a breadboard first to make sure everything works properly. That might be where the process ends for one-off projects, but for large production runs you’re going to need to test all the PCBs after they’re built, too. While you would use a breadboard for prototyping, the platform you’re going to need for quality control is called a “bed of nails“.
This project comes to us by way of [Thom] who has been doing a large production run of circuits meant to drive nixie tubes. After the each board is completed, they are laid on top of a number of pins arranged to mate to various points on the PCB. Without needing to use alligator clamps or anything else labor-intensive to test, this simple jig with all the test points built-in means that each board can be laid on the bed and tested to ensure it works properly. The test bed looks like a bed of nails as well, hence the name.
There are other ways of testing PCBs after production, too, but if your board doesn’t involve any type of processing they might be hard to implement. Nixie tubes are mostly in the “analog” realm so this test setup works well for [Thom]’s needs.
Despite appearances, [This Old Tony]’s latest series has little to do with CNC-ifying an Etch A Sketch. Although he certainly achieves that, more or less, automating the classic toy is just the hook for a thorough lesson in CNC machine building starting with the basics.
Fair warning: we said basics, and we mean it. [Old Tony]’s intended audience is those who haven’t made the leap into a CNC build yet and need the big picture. Part one concentrates on the hardware involved – the steppers, drivers, and controller. He starts with one of those all-in-one eBay packages, although he did upgrade the motion controller to a Mach4 compatible board; still, the lessons should apply to most hardware.
By the end of part one, the Etch A Sketch is connected to two of the steppers and everything is wired up and ready to go for part two, the first part of which is all about inputs and outputs. Again, this is basic stuff, like how relays work and why you might need to use them. But that’s the kind of stuff that can baffle beginners and turn them off to the hobby, so kudos to [Old Tony] for the overview. The bulk of the second part is about configuring Mach4 Hobby, with a ton of detail and some great tips and tricks for getting a machine ready to break some end mills.
For someone looking to get into a CNC build, [Old Tony]’s hard-won CNC experience really fills in the gaps left by other tutorials. And it looks like a third part, dealing with making all this into something more than an automated Etch A Sketch, is in the works. We’re looking forward to that.
Building an LED matrix is a fun project, but it can be a bit of a pain. Usually it starts with hand-soldering individual LEDs and resistors together, then hooking them up to rows and columns so they can be driven by a microcontroller of some sort. That’s a lot of tedious work, but you can order an LED matrix pre-built to save some time and headache. You’ll still need a driver though, and while building one yourself can be rewarding there are many pitfalls and trade-offs to consider when undertaking that project as well. Or, you can consider one of a number of drivers that [deshipu] has outlined in detail.
The hangups surrounding the driver board generally revolve around the issue of getting constant brightness from LEDs regardless of how many in the row or column are illuminated at one time. Since they are typically driven one row or column at a time, the more that are on the lower the brightness each LED will have. Driver boards take different approaches to solving this problem, which usually involve a combination of high-speed scanning of the matrix or using a constant-current source in order to eliminate the need for resistors. [deshipu] outlines four popular chips that achieve these purposes, and he highlights their pros and cons to help anyone looking to build something like this.
Most of these boards will get you to an 8×8 LED matrix with no problem, with a few going a few pixels higher in either direction. That might be enough for most of our needs, but for something larger you’ll need other solutions like the one found in this 64×32 LED matrix clock. There are also even more complicated drivers if you go into extra dimensions.