Basics: How Not to Solder

While there are a great many guides that teach you to solder (here is one fine example), we have found that there is a surprising lack of guides to help you with the opposite skill: How not to solder. This guide shows you some wonderful examples of how well circuits can come out when you disregard all of those other guides. Let’s get started!

Untitled

To begin with: Make sure that your circuit board is “generally messy.” A messy board might have leads trimmed to various lengths and/or extra little bits and blobs of solder and flux everywhere. Not only will the extra little bits of solder occasionally cause short circuits, but the disarray will help to hide other issues that might be lurking, making them nearly impossible to diagnose.

Untitled

Zooming in on that same example, we can find incomplete solder joints like the one close to the rubber foot. A joint like this may look like it’s making an electrical connection, when in reality it may or may not be. These kinds of joints really are the best, because they can lead to intermittent connections that usually work. Intermittent connections are also a great way to prank anyone who likes to debug electronics. Think of it as making your own Annoy-a-tron!

Untitled

While most of the solder joints shown here have a clean, smooth meniscus, there are also two fine examples of connections that have gaps in the solder joint. Gaps like these are essential to ensure adequate ventilation of the electronic components on the other side. Some people may tell you that joints like these may crack (or break off entirely) over time, but don’t listen to them.

026 - Version 2

When soldering components from the bottom side of your circuit board, you can sometimes — if you feed enough solder into your solder joint for long enough — wick enough through the holes to form blobs of solder on the top side of the circuit board.  You can see these blobs here on four pins of the chip, as well as on some of the resistor leads. These blobs are highly desirable because you can make a “trick” circuit board where all of the solder joints look good from the bottom side, but there are actually short circuits on the top side of the board.

If you hone your skill well enough (or just get lucky), you may even be able to create an “invisible” short circuit between two pins of a component, fully hidden beneath the component. We’ve seen “secret” shorts like these under both chips and discrete components like capacitors and LEDs.

As an added bonus, it usually takes quite a while to wick this much solder through a joint. Most soldering guides recommend that you limit the time that you heat a component to just a few brief seconds. If you ignore that to get this much solder in the joint, you may have the added outcome of overheating the component and damaging it beyond functionality. That way, even if someone were to find and remove the short circuit, the component still wouldn’t work.

Untitled

Expert mode! Going one step further, if you solder a given location on a circuit board for long enough or with enough pressure, you can actually delaminate the printed copper pad (trace) from the circuit board. The pad is usually a thin ring of copper around the hole with the pin that you’re trying to solder, or (on surface mount boards) simply a rectangle or oval of exposed copper that you solder to.

If you can manage to remove pads from a circuit board, then you remove the ability for a component to make electrical contact with the circuit board there. Sometimes, depending on the circuit, one can manually add a repair wire to fix the board. But in other cases, tearing off just a pad or two can destroy the circuit board beyond repair. (It’s also possible to break components this way, by overheating their leads.)

On single-sided circuit boards, you merely need to look at the pad once too many times to make it fall of. But on multi-layer (e.g., two-sided) circuit boards, pads tend to be resilient, so you’ve got to either heat them for quite a while or use pressure with the soldering iron to dislodge the pad. Again, this is “expert mode” territory, but the two most common techniques that we’ve seen for delaminating pads are (1) using a “cold heat” soldering iron (for which you may need to heat the joint for a very long time to get it to melt) and (2) repeatedly soldering and desoldering components at the same location.

In the photo above, the pads have been torn off of the circuit board at two of the solder points (both ends of one resistor). Rather than having the solder flow down to a smooth meniscus there, the solder forms a blob that sits above a mysterious dark circle at those two points— the exposed circuit board substrate.

Untitled

Here’s another example of what can happen when you heat a board for long enough. The two wires (red and black) from a battery holder are coming up through wide clearance holes next to the “8×8″ marking, and then are soldered back down to the VCC_IN and GND_IN locations in the “Batt. In” section.

The insulation around the two wires has been melted back (almost back to the wide clearance holes) from long overheating, and the wires themselves have been frayed until there are just a couple of fine strands making all of the electrical connections. Added bonus: Stray strands like these can help to cause intermittent short circuits, when the wires get bumped.

Untitled

Soldering guides will often try and steer you away from making “blobby” solder joints with excess solder, but there’s clearly no good reason for this.  If a little solder is good, surely more is better!

Some of the blobby solder joints (like those at at the lower left) are shaped like onions grown over the integrated circuit pins, making it impossible to see how (or even, if) the joint actually contacts the circuit board. Others — like the giant gravity-defying inverse silver teardrop in the center — seem to hover in mid-air above the circuit board, deftly managing to avoid contact with the plated through-hole of the circuit board.

image

Keeping the component leads long and using blob-style solder construction can also help you to protect your intellectual property, by obscuring your circuit design from prying eyes. Spaced along the left edge of this circuit board, you can see that there are eight LEDs wired up… or are there? By lumping the two pins for a given LED under a single blob of solder, no one will ever be certain! (Also worth noting: this technique may have some side effects on the functionality of those LEDs.)

The very lowest solder joint on this circuit board is where the power and ground wires (red and black) are attached from the battery holder. Note that these two wires have been soldered together. “Shorting” the power and ground together like this is a classic technique to protect a circuit from damage due to unwanted charging or static discharge. (Note, however, that if the battery is switched on with its leads shorted together like this, the battery itself will also discharge quickly, get very hot, and possibly even explode.)

023

Here’s another related technique: If you solder together multiple pins of your microcontroller, you can connect to all of those pins at once, ensuring that no one pin steps out of line, and that all of the pins will work together in perfect digital harmony.

Got any other favorite examples of “novel” soldering techniques? Let us know in the comments or in the flickr group, and we’ll do another roundup sometime!

We would like to sincerely thank the helpful individuals who kindly granted us permission to use their photos and also those who allowed us access to their boards for photography.

Podtique: podcast player in an antique radio

Podtique: Antique Podcast Player

Our friend Rick stopped by to show us his latest project, which he calls Podtique: a podcast player built into an antique radio cabinet. Using the original knobs, you tune in on “stations” which play different podcasts, with realistically generated static in between.

Podtique

The whole thing is run on a BeagleBone Black, and uses NeoPixel backlighting behind the dial. He’s written up the build on his blog, posted his code on github, and shared a heap of build photos in an album on flickr.

A Vintage Melody Synthesizer IC

Untitled

Leigh Klotz, author of Ham Radio for Arduino and PICAXE, gave us this interesting chip from the 80′s to play with: a UM3482A “Multi Instrument Melody Generator” IC.  While not quite rare, it is a bit of a vintage curiosity these days, and we wired one up to see what exactly it does.

Untitled

The vintage Radio Shack Archer package it came in was minimalist (though not quite this minimalist as it came to us– imagine that the chip were still there in the bubble). It promises not just 12 tunes, but circuit diagrams as well.

Untitled

This Archer-branded datasheet (including the promised circuit diagrams) came stapled to the back.  You can download a readable copy of the manufacturer’s original datasheet here.

“A mask-ROM-programmed multi-instrument melody generator, implemented in CMOS technology. It is designed to play the melody according to the previously programmed information and is capable of generating 12 songs with 3 different effects: piano, organ, and mandolin.”

Amongst the twelve musical selections are London Bridge, Row Row Row Your Boat, Oh My Darling Clementine, and (of course) Happy Birthday.  Suggested applications included toys, door bells, music boxes, and telephones.  (File under: Customized ringtones of the 80′s?)

Untitled

The three little graphs in the upper right show the three different timbres (the “instruments,” in the phrase “multi-instrument”) in terms of amplitude versus time. The other diagrams show how to wire it up to a speaker and how to configure the various inputs to select which song to play and so forth.

Untitled

We breadboarded up a sample circuit from the datasheet, substituting with parts we had on hand, including a little magnetic buzzer as the speaker. A 2xAA battery holder is connected up to the power and ground rails.  There is a momentary button switch to select the next melody in the set, and a row of DIP switches to set the configuration options.

And sure enough, it plays melodies.

From the mailbag: DIY holiday lights

Chris wrote in:

I wanted to say thank you for writing your blog and the products you’ve created.  I used both to make my christmas lights this year.  Couldn’t have done it without you.

 

I used an ATmega xx8 mini dev kit programmed with an ISP shield to control a series of WS2811 LED pixels to make beautiful light. The controller is designed to be standalone, not part of a bigger system. I used a BCD thumbwheel switch to select up to 10 looks.

 

We can give credit to the FastLED Library team for the heavy lifting.  The case is from your friends at Adafruit. The target boards don’t quite fit in the Adafruit cases.  Thankfully ya’ll had the schematic PDFs posted and I saw I could literally cut corners to make them fit.

Back to why I appreciate what you do so much, growing up my dad tried to teach me about electronics.  He was putting motorola 6802 CPU’s in Sperry New Holland hay bale wagons.  Unlike how easy we have it today with Arduino he had to program in assembly…

 

I got a degree in Theatre and loved doing sound and lighting design.  So I really love the art you put into your technology. I’ve built many of your kits.  My favorite are the interactive LED panels.  Blinking lights!  I like to put them behind the ikea glass “white boards” so my ideas really shine when I write them down.  I’m looking forward to seeing what you do in 2015!

Flickery Flame Soldering Kits!

FlickeryFlame

In October, we released our Solderless Flickery Flame project, based on a tiny breadboard with six red and yellow candle-flicker LEDs, to give a fun and semi-realistic flame effect. Today, we’re releasing two new Flickery Flame Soldering Kits along the same lines, each of which has 6 candle-flicker LEDs, a little circuit board, and a battery holder.

FlickeryFlame

The Yellow/Red kit has the same mix of yellow and red candle-flicker LEDs that works so well in the breadboard kit.  This one will look great in a jack-o-lantern, luminaria, or scale-model fireplace.

FlickeryFlame

On the other hand, the White/Warm White kit has a mixture of (cool) white and warm white LEDs that give a modern wintery flame effect that has at least as much charm, but won’t be mistaken for a natural fire. This one will look great in all kinds of winter holiday decorations, luminarias, and props.

FlickeryFlame FlickeryFlame

Both the Yellow/Red and White/Warm White kits are fun, low-cost, self-contained, and easy soldering kits, which will be right at home both as stocking stuffers and as bite-size first projects for soldering workshops.

Interactive LED Cats

We were sent a picture of handsome cat Gandalf with an array of green Octolively kits all built up.

Mr. Pumpernickel also looks great in the green glow. Both Gandalf and Mr. Pumpernickel are continuing in a longstanding tradition of cats and interactive LEDs.

Harley Cat, Before frosting glass

Harley Cat (who passed away a few years ago) helped test our very first interactive LED project: our Interactive LED Dining Table.

Jellybean helped demonstrate a later project: our interactive LED coffee table. She is featured at about 48 seconds into this video.

LED Pez Menorah, this time with instructions!

Joyce wrote in this year with her latest Pez Menorah (using our Deluxe LED Menorah kit), featuring a Star Wars theme.

She also wrote up a thorough set of instructions on tumblr, which looks like a remarkably easy way to publish a step-by-step tutorial without using a platform like instructables. It may take some getting used to reading chronologically, but it is effective.

Previous Pez Menorah posts:

LED Metronome


After seeing our Larson Scanner kit, Martin shared this LED metronome project with us. Martin says:

It was designed as a “Visual Metronome”  So a learning music student could help see the timing by watching the green light.  There was to be an optional clicking sound  by using a small solenoid for the ticking – I chose that in place of a speaker for a more authentic sound.

The timing is a standard 555 timer which is fed to 7442 BCD to DECIMAL counter.  Next chip is a 74193 UP/DOWN counter.  When the count hits the last number, it sends a pulse to reverse the count or start over – depending on the toggle switch on the side.

There is also a pot on the 555 to control the speed.  All this was made in one night while I was working the graveyard shift.

The entire LED display was hand wired using a manual wire-wrap tool.

The chip pin labels on the back of the perf board are a particularly awesome relic of a different era of electronics assembly. Thanks for sharing your project photos and video with us!

Introducing WaterColorBot 2.0

WaterColorBot 2.0

We are very pleased to introduce something that we’ve been working on for most of this year: WaterColorBot version 2.0!

WaterColorBot 2.0

The WaterColorBot is our collaboration with Super Awesome Sylvia: A friendly art robot that moves a paint brush to paint your digital artwork onto paper, using a set of watercolor paints.

Version 2.0 brings it to the next level with some greatly improved hardware. First and foremost, the carriage that holds the brush has been completely redesigned:

WaterColorBot 2.0

The carriage on the original WaterColorBot was made from laser-cut plywood, with nylon bushings and two simple delrin strips that formed the vertical flexure translation stage. (You can read more about the original carriage here and here.)

The new carriage consists mainly of two pieces of metal. The center block of anodized aluminum is CNC milled, and houses crossed linear roller bearings. Wrapped around that is a laser-cut and formed aluminum part that mounts the brush-lift motor, cable guide, and the flexure stage.

WaterColorBot 2.0

The new flexure stage is built with two custom flex circuit boards, used in this case as mechanical flexures. Each board consists of a very thin (0.1 mm, 4 mil) Kapton sheet with a thin fiberglass (G10/FR4) stiffener on its center section. With the two ends of each sheet clamped rigidly and the stiffener in the center, each flex circuit is to flex only along two well-defined lines. And with two boards, it forms a neat parallelogram linkage, without the slop that one might encounter in multi-part hinges. The net effect is that this new flexure stage has remarkable stiffness compared to the old design.

WaterColorBot 2.0

That stiffness, combined with the improved performance of the linear ball bearings makes this a more precise WaterColorBot. Not that you could even detect the improvement with a fat brush and watercolor paints, but things are looking quite good even with using ultra-fine point drawing pens, as you can see above.

WaterColorBot 2.0

The second major change is to the system of Spectra cords that the stepper motors control in order to move the carriage. Previously, the cords were guided around 11 plain bearings (stainless steel solid rivets) and 3 ball bearings.  We’ve simplified this into an arrangement of just 8 ball bearings— four for each motor. The ball bearing pulleys have also been updated to use wide V-groove bearings that are easy to wrap the cords around.

Which brings us to the third (and last) major change. Thus far, WaterColorBot kits have shipped “some assembly required” — with all the major components built, but the cord lacing left to the end user. As of 2.0, WaterColorBot kits now come fully assembled and tested. That doesn’t make them any less hackable, but it does mean that you can get up and running faster.

WaterColorBot 2.0

Version 2.0 includes the same CNC machined aluminum winches that we introduced back in August. Tiny detail: we’ve carved a subtle indentation into the wood around the winch that makes them a little easier to turn by hand.

WaterColorBot 2.0

The new WaterColorBot kits will begin to ship right after Thanksgiving. And a bonus present for the holiday season: Version 2.0 is priced the same as the previous version, it’s just a whole lot more awesome per dollar.