More people working from home has had an impact on the cost and availability of USB webcams, so [Jeff Geerling] got around the issue with a DIY solution that rang in around $100. It consists of a Raspberry Pi and HQ camera module acting as a USB webcam, and there is no messy streaming of ffmpeg over the network masquerading as a camera device or anything. It works just as a USB camera should.
[Jeff] chose a Raspberry Pi Zero and HQ camera module for his unit, making a tidy package that might not be quite as small as commercial webcams, but is certainly perfectly respectable as a USB camera. That being said, there are a few drawbacks, namely the lack of a microphone or autofocus, latency issues at higher resolutions, and the need to shut down the Pi cleanly.
Check out the GitHub repository for everything needed to set up your own, including a complete hardware list and some options for mounting. [Jeff] also tested whether the camera would work with the new keyboard-embedded Raspberry Pi 400, and it absolutely does. Embedded below is a video walkthrough and demonstration of the whole project, so check it out.
Lacking a DVD drive, [jg] was watching a TV series in the form of a bunch of .avi video files. Of course, when every episode contains a full intro, it is only a matter of time before that gets too annoying to sit through.
Chapter breaks reliably inserted around the intro, even when it doesn’t always occur in the same place.
The usual method of skipping the intro on a plain video file is a simple one:
Manually drag the playback forward past the intro.
Oops that’s too far, bring it back.
Ugh reversed it too much, nudge it forward.
Okay, that’s good.
[jg] was certain there was a better way, and the solution was using audio fingerprinting to insert chapter breaks. The plain video files now have a chapter breaks around the intro, allowing for easy skipping straight to content. The reason behind selecting this method is simple: the show intro is always 52 seconds long, but it isn’t always in the same place. The intro plays somewhere within the first two to five minutes of an episode, so just skipping to a specific timestamp won’t do the trick.
The first job is to extract the audio of an intro sequence, so that it can be used for fingerprinting. Exporting the first 15 minutes of audio with ffmpeg easily creates a wav file that can be trimmed down with an audio editor of choice. That clip gets fed into the open-source SoundFingerprinting library as a signature, then each video has its audio track exported and the signature gets identified within it. SoundFingerprinting therefore detects where (down to the second) the intro exists within each video file.
Marking out chapter breaks using that information is conceptually simple, but ends up being a bit roundabout because it seems .avi files don’t have a simple way to encode chapters. However, .mkv files are another matter. To get around this, [jg] first converts each .avi to .mkv using ffmpeg then splices in the chapter breaks with mkvmerge. One important element is that the reformatting between .avi and .mkv is done without completely re-encoding the video itself, so it’s a quick process. The result is a bunch of .mkv files with chapter breaks around the intro, wherever it may be!
The script is available here for anyone to play with, and the project page is a good learning reference because [jg] kindly provides all the command-line options used for each tool. Interested in using audio fingerprinting in your own projects? Remember to also check out Olaf, the Overly Lightweight Acoustic Fingerprinting method that can be implemented in embedded systems and web browsers.
[Jesse]’s modification doesn’t affect the laser beam itself; it is an improvement on the air assist, which is the name for a constant stream of air that blows away smoke and debris as the laser burns and vaporizes material. An efficient air assist is one of the keys to getting nice clean laser cuts, but [Jesse] points out that a good quality air assist isn’t just about how hard the air blows, it’s also about how smoothly it does so. A turbulent air assist can make scorch marks worse, not better.
3D-printed nozzle to promote laminar air flow on the left, stock nozzle on the right.
As an experiment to improve the quality of the air flowing out the laser nozzle, [Jesse] researched ways to avoid turbulence by creating laminar flow. Laminar flow is the quality of a liquid having layers flowing past one another with little or no mixing. One way to do this is to force liquid through individual, parallel channels as it progresses towards a sharply-defined exit nozzle. While [Jesse] found no reference designs of laminar flow nozzles for air assists, there were definitely resources on making laminar flow nozzles for water. It turns out that interest in such a nozzle exists mainly as a means of modifying Lonnie Johnson’s brilliant invention, the Super Soaker.
Working from such a design, [Jesse] created a custom nozzle to help promote laminar flow. Sadly, a laser cutter head carries design constraints that make some compromises unavoidable; one is limited space, and another is the need to keep the laser’s path unobstructed. Still, after 3D printing it in rigid heat-resistant resin, [Jesse] found a dramatic improvement in the feel of the air exiting the nozzle. Some test cuts confirmed a difference in performance, which results in a noticeably cleaner kerf without scorching around the edges.
One of the things [Nervous System] does is make their own custom puzzles, so any improvement to laser cutting helps reliability and quality. When production is involved, just about everything matters; a lesson [Nervous System] shared when they discussed making the best plywood for creating their puzzles.
[Niklas Roy] rolled his own photo diary, because he found the core functionality of something like instagram attractive, but didn’t want the social network baggage that it came with. His simple system is called my own insta ;) and it consists of some javascript and PHP to create a nice progressive web app photo diary and backend that can be accessed just fine from a mobile device. It is available on GitHub for anyone interested in having their own.
This project came up because [Niklas] sometimes found himself working on small projects or experiments that aren’t destined for proper documentation, but nevertheless could benefit from being shared as a photo with a short description. This dovetails with what many social networks offer, except that those platforms also come with other aspects [Niklas] doesn’t particularly want. His online photo diary solves this by having a simple back end with which he can upload, sort, and caption photos in an easy way even from a mobile device.
[Jan Mrázek] is no stranger at all to home-grown improvements with his Elegoo Mars SLA 3D printer, and there is a lot going on in his experimental multi-LED upgrade which even involved casting his own lens array. In the end it did speed up his prints by a factor of three to four, though he cooked an LCD to failure in the process. Still, it was a fun project done during a COVID-19 lockdown; as usual there is a lot to learn from [Jan]’s experiences but the mod is not something he necessarily recommends people do for themselves.
[Jan] started by wondering whether better print quality and performance could be obtained by improving the printer’s UV light source. The stock printer uses a single large UV LED nestled into a reflector, but [Jan] decided to try making a more precise source of UV, aiming to make the UV rays as parallel as possible.
Custom LED array molded in clear epoxy.
To do this, he took a two-pronged approach. One was to replace the single large UV LED with a 4×7 array of emitters plus heat sink and fans. The other was to make a matching array of custom lenses to get the UV rays as parallel as possible.
Casting one’s own lens array out of clear epoxy was a lot of work and had mixed results, but again, it was a lockdown project and the usual “is-this-really-worth-it” rules were relaxed. In short, casting a single custom lens out of clear epoxy worked shockingly well, but when [Jan] scaled it up to casting a whole 4×7 array of them, results were mixed. Mold deformation and artifacts caused by the areas between individual lenses robbed the end result of much of its promise.
More success was had with the array of UV emitters, which enabled faster curing thanks to higher power, but the heat needs to be managed. The stock emitter of the printer is about 30 W, and [Jan] was running his new array at 240 W. This meant a blazing fast one second exposure time per layer, but the heat generated by the new lighting was higher than anticipated. After only ten hours the LCD failed, probably at least in part due to the heat. [Jan] halved the power of the array down to 120 W and added an extra fan, which appears to have done the trick. Exposure time is two to three seconds per layer, and it’s up to 150 hours of printing without problems.
Again, it’s not a process [Jan] necessarily recommends to others (and he definitely recommends buying lenses if at all possible instead of casting them) but as usual there is a lot to learn from his frank sharing of results, both good and bad. We’ve seen 3D-printed lenses as well as adding WiFi connectivity to one of these hobbyist printers, and it’s great to see the spirit of hacking alive and well when it comes to these devices.
3D printing can be great for making enclosures, and following some simple guidelines can help the whole process go much smoother. 3D Hubs has an article on designing printed enclosures that has clear steps and tips to get enclosures coming out right the first time. 3D Hubs offers 3D printing and other services, and the article starts with a short roundup of fabrication methods but the rest is a solid set of tips applicable to anyone.
The first recommendation is to model the contents of the enclosure as a way to help ensure everything fits as it should, and try to discover problems as early as possible during the design phase, before anything gets actually printed. We’ve seen how a PCB that doesn’t take the enclosure into account risks needing a redesign, because there are some issues an enclosure just can’t fix.
The rest of their advice boils down to concrete design guidelines about wall thickness (they recommend 2 mm or more), clearances (allow a minimum of 0.5 mm between internal components and enclosure), and how to size holes for fasteners, clips, or ports. These numbers aren’t absolute minimums, but good baseline values to avoid surprises.
One final useful tip is that using a uniform wall thickness throughout the enclosure is general good practice. While this isn’t strictly necessary for successful 3D printing, it will make life easier if the enclosure ever moves to injection molding. Want to know more? Our own Bob Baddeley has an excellent primer on injection molding, and his been-there-done-that perspective is invaluable.
3D printers are amazing things, but if one judges solely by the successes that get showcased online, it can look as through anything at all is possible. Yet in many ways, 3D printers are actually quite limited. Because success looks easy and no one showcases failure, people can end up with lopsided ideas of what is realistic. This isn’t surprising; behind every shining 3D print that pushes the boundaries of the technology, there are misprints and test pieces piled just out of sight.
If you have ever considered getting into 3D printing, or are wondering what kinds of expectations are realistic, read on because I am going to explain where objects come from, and how to recognize whether something is a good (or bad) fit for 3D printing. The important thing to understand is that printers have limitations, and to get a working idea of what those limitations are. The result will be a better understanding of what they can do, and what problems they can reliably solve.
3D Printers Have Limits
I recently had a talk with someone who wanted to know if a 3D printer could help with a problem they had. As I listened to them describe their needs, I realized I had in a way heard it all before many times.
My colleague actually had a fairly good idea of what printers could do, in theory. But they had very little grasp of what printers did not do, and that disconnect left them a bit adrift when it came to practical applications. To help address this gap, here are some tips that can give anyone a working understanding of the things 3D printers do not do well. Continue reading “3D Printering: The Things Printers (Don’t) Do”→
[Jeff] chose a Raspberry Pi Zero and HQ camera module for his unit, making a tidy package that might not be quite as small as commercial webcams, but is certainly perfectly respectable as a USB camera. That being said, there are a few drawbacks, namely the lack of a microphone or autofocus, latency issues at higher resolutions, and the need to shut down the Pi cleanly.
