Hand soldering of surface mount components is a bread-and-butter task for anyone working with electronics in 2024. So many devices are simply no longer available in the older through-hole formats, and it’s now normal for even the most homebrew of circuits to use a PCB. But how do you solder your parts? If like us you put a blob of solder on a pad and drop the part into it, then [Mr. SolderFix] has some advice on a way to up your game.
The blob of solder method leaves a little more solder on the part than is optimal, sometimes a bulbous lump of the stuff. Instead, he puts a bit of flux on the pad and then applies a much smaller quantity of solder on the tip of his iron, resulting in a far better joint. As you can see in the video below, the difference is significant. He starts with passives, but then shows us the technique on a crystal, noting that it’s possible to get the solder on the top of these parts if too much is used. Yes, we’ve been there. Watch the whole video, and improve your surface mount soldering technique!
We’ll be honest with you: we’re not sure if the use of “LED stud” in [mitxela]’s new project refers to the incomprehensibly tiny LED matrix earrings he made, or to himself for attempting the build. We’re leaning toward the latter, but both seem equally likely.
This build is sort of a mash-up of two recent [mitxela] projects — his LED industrial piercing, which contributes the concept of light-up jewelry in general as well as the power supply and enclosure, and his tiny volumetric persistence-of-vision display, which inspired the (greatly downsized) LED matrix. The matrix is the star of the show, coming in at only 9 mm in diameter and adorned with 0201 LEDs, 52 in total on a 1 mm pitch. Rather than incur the budget-busting expense of a high-density PCB with many layers and lots of blind vias, [mitexla] came up with a clever workaround: two separate boards, one for the LEDs and one for everything else. The boards were soldered together first and then populated with the LEDs (via a pick-and-place machine, mercifully) and the CH32V003 microcontroller before being wired to the power source and set in the stud.
Even though most of us will probably never attempt a build on this scale, there are still quite a few clever hacks on display here. Our favorite is the micro-soldering iron [mitxela] whipped up to repair one LED that went missing from the array. He simply wrapped a length of 21-gauge solid copper wire around his iron’s tip and shaped a tiny chisel point into it with a file. We’ll be keeping that one in mind for the future.
When it comes to home-lab reflow work, there are a lot of ways to get the job done. The easiest thing to do perhaps is to slap a PID controller on an old toaster oven and call it a day. But if your bench space is limited, you might want to put this compact reflow hotplate to work for you.
There are a lot of nice features in [Toby Chui]’s build, not least of which is the heating element. Many DIY reflow hotplates use a PCB heater, where long, thin traces in the board are used as resistive heating elements. This seems like a great idea, but as [Toby] explains in the project video below, even high-temperature FR4 substrate isn’t rated for the kinds of temperatures needed for some reflow profiles. His search for alternatives led him to metal ceramic heaters (MCH), which are commonly found in medical and laboratory applications. The MCH he chose was rated for 20 VDC at 50 watts — perfect for powering with USB-PD.
The heater sits above the main PCB on a Kapton-wrapped MDF frame with a thermistor to close the loop. While it’s not the biggest work surface we’ve seen, it’s a good size for small projects. The microcontroller is a CH552, which we’ve talked about before; aside from that and the IP2721 PD trigger chip needed to get the full 60 watts out of the USB-PD supply, there’s not much else on the main board.
This looks like a nice design, and [Toby] has made all the design files available if you’d like to give it a crack. Of course, you might want to freshen up on USB-PD before diving in, in which case we recommend [Arya]’s USB-PD primer.
There’s no denying how useful surface mount technology is, and how enabling the ability to make really small circuits has become. It comes at a price, though; most of us probably know what it’s like for the slightest wrong move to send a part the size of a grain of sand into another dimension.
To help make testing these parts a little easier, [IMSAI Guy] has come up with this clever little SMD test fixture. It’s designed to hook up to another custom board, which in turn connects to a wonderful old Hewlett-Packard 4275A LCR meter. The jig is based on two pogo pins mounted directly across from each other on a scrap of single-clad PCB. The spring-loaded contacts, which short together when not in use, are pulled apart to load an SMD part, like the 1-μH inductors shown in the video below. The pins hold the component firmly and make good electrical contact, allowing hands-free testing without the risk of an errant touch of the test probes sending it flying.
While the test fixture works well for larger SMDs, we could see this being a bit fussy for smaller parts. That would be easy enough to fix with perhaps some 3D-printed arms that retract the pogo pins symmetrically, holding them open until the part is loaded. A centering fixture might help too, as would a clear shield to contain any parts that get the urge to go for a ride. But, for the tactical application [IMSAI Guy] has in mind, this sure seems like enough.
When you say “recapping” it conjures up an image of a dusty old chassis with point-to-point wiring with a bunch of dried-out old capacitors or dodgy-looking electrolytics that need replacement. But time marches on, and we’re now at the point where recapping just might mean removing SMD electrolytics from a densely packed PCB. What do you do then?
[This Does Not Compute]’s answer to that question is to try a bunch of different techniques and see what works best, and the results may surprise you. Removal of SMD electrolytic caps can be challenging; the big aluminum can sucks a lot of heat away, the leads are usually pretty far apart and partially obscured by the plastic base, and they’re usually stuffed in with a lot of other components, most of which you don’t want to bother. [TDNC] previously used a hot-air rework station and liberally applied Kapton tape and aluminum foil to direct the heat, but that’s tedious and time-consuming. Plus, electrolytics sometimes swell up when heated, expelling their corrosive contents on the PCB in the process.
As brutish as it sounds, the solution might just be as simple as ripping caps off with pliers. This seems extreme, and with agree that the risk of tearing off the pads is pretty high. But then again, both methods seemed to work pretty well, and on multiple boards too. There’s a catch, though — the pliers method works best on caps that have already leaked enough of their electrolyte to weaken the solder joints. Twisting healthier caps off a PCB is likely to end in misery. That’s where brutal method number two comes in: hacking the can off the base with a pair of flush cutters. Once the bulk of the cap is gone, getting the leads off the pad is a simple desoldering job; just don’t forget to clean any released schmoo off the board — and your cutters!
To be fair, [This Does Not Compute] never seems to have really warmed up to destructive removal, so he invested in a pair of hot tweezers for the job, which works really well. But perhaps you’re not sure that you should just reflexively replace old electrolytics on sight. If so, you’re in pretty good company.
While for some of us it’s a distant memory, every serious electronics hobbyist must at some point make the leap from working with through-hole components to Surface Mount Devices (SMD). At first glance, the diminutive components can be quite intimidating — how can you possibly work with parts that are literally smaller than a grain of rice? But of course, like anything else, with practice comes proficiency.
It’s at this silicon precipice that [Larry Bank] recently found himself. While better known on these pages for his software exploits, he recently decided to add SMD electronics to his repertoire by designing and assembling a pocket-sized CO2 monitor. While the monitor itself is a neat gadget that would be worthy of these pages on its own, what’s really compelling about this write-up is how it documents the journey from SMD skeptic to convert in a very personal way.
A fine-tipped applicator will get the solder paste where it needs to go.
At first, [Larry] admits to being put off by projects using SMD parts, assuming (not unreasonably) that it would require a significant investment in time and money. But eventually he realized that he could start small and work his way up; for less than $100 USD he was able to pick up both a hot air rework station and a hotplate, which is more than enough to get started with a wide range of SMD components. He experimented with using solder stencils, but even there, ultimately found them to be an unnecessary expense for many projects.
While the bulk of the page details the process of assembling the board, [Larry] does provide some technical details on the device itself. It’s powered by the incredibly cheap CH32V003 microcontroller — they cost him less than twenty cents each for fifty of the things — paired with the ubiquitous 128×64 SSD1306 OLED, TP4057 charge controller, and a SCD40 CO2 sensor.
Whether you want to build your own portable CO2 sensor (which judging from the video below, is quite nice), or you’re just looking for some tips on how to leave those through-hole parts in the past, [Larry] has you covered. We’re particularly eager to see more of his work with the CH32V003, which is quickly becoming a must-have in the modern hardware hacker’s arsenal.
Elliot Williams and Al Williams got together again to discuss the best of Hackaday for a week, and you’re invited. This week, the guys were into the Raspberry Pi 5, CNC soldering, signal processing, and plasma cutting. There are dangerous power supplies and a custom 11-bit CPU.
Of course, there are a few Halloween projects that would fit in perfectly with the upcoming Halloween contest (the deadline is the end of this month; you still have time). OpenSCAD is about to get a lot faster, and a $20 oscilloscope might not be a toy after all. They wrap up by talking about Tom Nardi’s latest hardware conversion of DIP parts to SMD and how TVs were made behind the Iron Curtain.
Did you miss a story? Check out the links below. As always, tell us what you think about this episode in the comments!
