Hobbyist's Experiment Creates a Self-Soldering Circuit Board (hackaday.com) 114
Long-time Slashdot reader wonkavader found a video on YouTube where, at the 2:50 mark, there's time-lapse footage of soldering paste magically melting into place. The secret?
Many circuit boards include a grounded plane as a layer. This doesn't have to be a big unbroken expanse of copper — it can be a long snake to reduce the copper used. Well, if you run 9 volts through that long snake, it acts as a resistor and heats up the board enough to melt solder paste. Electronics engineer Carl Bugeja has made a board which controls the 9 volt input to keep the temperature on the desired curve for the solder.
This is an interesting home-brew project which seems like it might someday make a pleasant, expected feature in kits.
Hackaday is impressed by the possibilities too: Surface mount components have been a game changer for the electronics hobbyist, but doing reflow soldering right requires some way to evenly heat the board. You might need to buy a commercial reflow oven — you can cobble one together from an old toaster oven, after all — but you still need something, because it's not like a PCB is going to solder itself. Right?
Wrong. At least if you're Carl Bugeja, who came up with a clever way to make his PCBs self-soldering.... The quality of the soldering seems very similar to what you'd see from a reflow oven.... After soldering, the now-useless heating element is converted into a ground plane for the circuit by breaking off the terminals and soldering on a couple of zero ohm resistors to short the coil to ground.
It's an open source project, with all files available on GitHub. "This is really clever," tweeted Adrian Bowyer, inventor of the open source 3D printer the RepRap Project.
In the video Bugeja compares reflow soldering to pizza-making. (If the circuit board is the underlying dough, then the electronics on top are the toppings, with the solder paste representing the sauce that keeps them in place. "The oven's heat is what bonds these individual items together.")
But by that logic making a self-soldering circuit is "like putting the oven in the dough and making it edible."
This is an interesting home-brew project which seems like it might someday make a pleasant, expected feature in kits.
Hackaday is impressed by the possibilities too: Surface mount components have been a game changer for the electronics hobbyist, but doing reflow soldering right requires some way to evenly heat the board. You might need to buy a commercial reflow oven — you can cobble one together from an old toaster oven, after all — but you still need something, because it's not like a PCB is going to solder itself. Right?
Wrong. At least if you're Carl Bugeja, who came up with a clever way to make his PCBs self-soldering.... The quality of the soldering seems very similar to what you'd see from a reflow oven.... After soldering, the now-useless heating element is converted into a ground plane for the circuit by breaking off the terminals and soldering on a couple of zero ohm resistors to short the coil to ground.
It's an open source project, with all files available on GitHub. "This is really clever," tweeted Adrian Bowyer, inventor of the open source 3D printer the RepRap Project.
In the video Bugeja compares reflow soldering to pizza-making. (If the circuit board is the underlying dough, then the electronics on top are the toppings, with the solder paste representing the sauce that keeps them in place. "The oven's heat is what bonds these individual items together.")
But by that logic making a self-soldering circuit is "like putting the oven in the dough and making it edible."
Though-hole through and through (Score:1)
Almost anything I'd be willing to do at home in terms of complexity I can find through hole components or at least through hole sockets for.
I can (and have) hand-soldered some larger surface mount dips and qfp devices like pic24s. A buddy of mine used to solder smaller bga chips with a heat gun.
But anything beyond that seems like it would be enough complexity in the design to justify spending money anyway. Prototype it with a dev kit and a breadboard and have it made up and populated by a contract pcb manuf
Re: (Score:2)
I've read enough about BGA that I'm not interested. I won't even buy BGA chips, if I can avoid it.
Re: (Score:1)
Re: Though-hole through and through (Score:2)
It's true but you can do things in order to make it painless.
Do all the through hole mount chips first and you can just rest the board on them. Then do the resistors and diodes and long lead stuff and it'll hold itself in place. Then any protruding caps or transistors last.
What you lose in trimming leads you gain back in how fast the work is relative to small surface mounts.
Again, we're talking about one off stuff, not a production line.
Re: (Score:2)
For almost anything, that is true and fair enough. But when you need very small runs it isn't cost effective to have a manufacturer populate the board for you and/or there are some components you'll need to do yourself.
I've run into a few dozen cases over the years where I needed something one off that someone had a schematic for, didn't lend itself to microcontroller usage or was a board to be driven by the same, etc. Ironically enough, one example was when I went to DIY a reflow oven from a toaster oven.
Re: Though-hole through and through (Score:1)
I get my boards made professionally. *I* don't have to drill a thing. The same fella I mentioned with the heat gun dabbled in making his own boards photochemically. I don't really have a way of disposing of those chemicals so I got to pcbexpress pr whatever the hell there is now and wait a few days.
Re: Though-hole through and through (Score:2)
Clever but not really useful (Score:2)
It's like watching a dog walk only on hind legs. It's amazing for the first 15 seconds then ...
Re: (Score:2)
Would this be useful at end of life ? An easy way of removing components from a board so that they can be recycled.
Re: (Score:2)
Typically you do that with a fancy hair dryer called a "reflow station." Or if you wanted even less human intervention, an oven.
I prefer wire wrapping (Score:2)
I honestly prefer the way electronics used to be made with wire wrapping, and larger components.
Re: (Score:2)
You can still do that. Wire wrap wire, dip sockets and etc are still available.
Last time I had something made, a lot of the factory test kit was cobbled together with wire wrap. Surprised me do so that testing a surface mount board full of qfns.
Re: (Score:2)
Well, yes and no.
More and more components are not made anymore as DIP. There simply is no market for them anymore, what you get as DIP chips is mostly older components, and even there you sometimes run into the problem that only the SMD version is actually readily available.
Re: (Score:2)
Well, yes and no.
Indeed! Even quite a lot of THM stuff isn't .1" any more. Like a decent buck/boost module I used a while back.
More and more components are not made anymore as DIP. There simply is no market for them anymore, what you get as DIP chips is mostly older components, and even there you sometimes run into the problem that only the SMD version is actually readily available.
True, but there's ways around that. To be sure: you can't survive on wire wrapping alone any more, or any .1" tech. I did some
Re: (Score:2)
It seems like there are preprepared carriers for common packages, are there just too many different ones to bother using them?
Re: (Score:2)
It seems like there are preprepared carriers for common packages, are there just too many different ones to bother using them?
As usual, it depends, also my information is probably out of date. My experience was that they were also quite expensive which is fine, but you need quite a lot. SOP, SOIC, maybe SSOP, in the two sizes each, which is OK, but for the non interchangeable ones, there's about a million QFNs of various pin counts and pitches and the oddball one like low pincount LGA packages.
In practice I
Re: (Score:2)
I wire wrap things now and then when prototyping. Sometimes you just have components that don't lend themselves nicely to a breadboard.
One example was a project to convert an old scanner into a pcb uv exposure unit. I got ahold a bunch of uv resist film I could stick to PCBs, there is a wet version of this as well but the film always gives an even coat and applies sort of like sticking a screen protector on a cell phone screen.
In any case, I needed a uv light panel which not only put out enough light but al
Re: (Score:2)
Re: (Score:2)
... I never really got the appeal of SMDs for hobbyists either. I prefer regular components and soldered wires.
Similar experience level with electronics here, both as a hobby and as a career. There are now a lot of semiconductors used regularly even by hobbyists, that simply aren't available in through-hole packages. And even things like resistors are now cheaper as 0805 packages than as leaded.
Once you've made the adjustment, assembling and re-working surface-mount boards manually isn't that much of a pain. And when you're designing, the ability to make boards smaller and put components closer together is a game c
Re: (Score:2)
Resistors etc are WAY cheaper if you buy the surface mount versions. You can get a reel of 1000 quality, name-brand resistors for under ten bucks from digikey or Farnell (none of this aliepxress, amazon or ebay shit). Buy a few reels in a few common values (300 ohm for LEDs, 1k for transistors, 10k for pull-ups/pull-downs, 100k for weak pull ups/downs, capacitor drains etc) then you'll rarely have to buy a resistor again, certainly not for your first proof of concept design. I find that 0603 resistors are d
Re: (Score:2)
You can get a reel of 1000 quality, name-brand resistors for under ten bucks from digikey or Farnell
You're better off buying a reel of 5000 for $13 than you are buying 1000 and paying the reeling fee. You can get 1000 0603 resistors for $3 or $4 from Digi-Key if you're willing to just have cut tape. Most hobbyists have no use for reels.
Re: (Score:2)
You're right, that was sloppy language on my part. I usually just buy the long tapes. Cheaper and a hobbyist doesn't usually need the actual reel.... although DIY automated pick and place could be a fun rabbit hole to go down...
Re: (Score:2)
0603 imperial, I hope. Else I wonder whether you enjoy soldering under a microscope.
Personally, I think that's no fun. Doable, but certainly nothing you want to do for longer than you have to.
Re: (Score:2)
Yeah, 0603 imperial. 0603 in metric are way too small and easy for me to lose!
Re: (Score:2)
Don't breathe. :)
Re: (Score:2)
Re: (Score:2)
I love discussions like this. I've been using KiCad for about as long as you have, though I haven't been active in a while so I'm a good number of revs behind.
As for SMD passives, 0805 is kind of a sweet spot size-wise. I've used that size of resistor and capacitor on prototype boards with holes and square pads on a 0.1" grid and with a ground plane on one side. 0805 components solder easily onto the pads, and the grid size allows for a wide variety of leaded components including DIP packages. Even SOT23 pa
Re: (Score:2)
I think I settled on 0603 after reading a datasheet that promised terrible things if I used decoupling caps that were too physically big. It was probably overkill as I really don't do any super high speed stuff.
That protoboard sounds useful - do you have a link?
Re: (Score:2)
Resistors may be cheaper as SMD, but resistors are not expensive in the first place and the SMD ones are annoying to work with. It sucks when a chip is not available in a DIP or at least SOIC package though.
I am not making more than one or two copies anyway, most of the time just one copy and a perfboard is much faster/easier than designing a PCB (and having worry about the size of capacitors I'll get etc).
Re: (Score:2)
Wire wrapping is useful in all sorts of situations you would not expect. Get yourself a tool, try it, and then put it away with the thought "OK, I tried that, I'll never use it." and over and over you're going to think wait... let's wire-wrap this onto that. Connections are better and stronger when you solder after wire-wrapping then when you just lay things together or hand-twist them. For example, in-line resisters and other components in impromptu spider-webs of components are great wire-wrap targets.
Re: (Score:2)
Yup. The project that got me wire wrapping was a large size LED light panel. It just wasn't practical to get a copper clad board as large as I needed but with wirewrapping I just needed cardboard cut to size and could poke leds through wherever I needed them.
Re: I prefer wire wrapping (Score:2)
Eh. The laziest way I think is to:
1. Use an eval board or something for the main IC. Don't even try something like a BGA Nvidia Jetsons.
2. Get PCB way to make a circuit board for your analog circuits/power control/whatever you are doing.
3. Use throughhole when available, use the biggest smd size when not
4. Just use tweezers and hot air. I found for soldering on surface mount this to be the easiest. If you do this a lot buy robots to do this step for you.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I was the same 'til I tried it. Wrapping has one unbeatable advantage over soldered prototyping: It's far, far easier to tweak the design or remove components. Ever spent a few hours salvaging a soldered chip because you didn't want to trash it? Well, with wire wrap you can keep the chip AND the board.
It's my in-between step between breadboard and finished PCB, where I can try out different chips and configurations to see whether the timings align before I commit to solder.
Re: (Score:1)
surf or solder (Score:3)
I love the smell of rosin flux in the morning...
Re: (Score:2)
Re: (Score:2)
I too love that smell. It reminds me of ye olden days when 6040 solder flowed smoothly like quicksilver, without guilt or derision...
No derision from this quarter - I still unapologetically and unashamedly use resin-core 60/40 lead-based solder for everything. It's the easiest and most reliable option; and with all the mega-scale toxic dumping still going on in the world, the amount of lead contributed to the environment by millions of folks like me, is about equivalent to spitting in the ocean.
Re: (Score:2)
I have a spool and a half left over from when radio shack still sold it, and every time I have to use something else it's like torture
Nice trick, but I doubt it's really useful... (Score:5, Informative)
Makes PCB more expensive, and worst of all, it's an EMI/EMC nightmare.
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
Would mod up if I had points. Ground planes are continuous for a bunch of really good reasons, putting a bazillion slots into them defeats most of them.
I had the same thought when I read the story on HaD. But for low-speed boards it's still a good technique. And depending on the number of layers and the size of the board, you could sandwich the heating layer between a ground plane and a power plane. That would largely eliminate EMI / EMC and crosstalk concerns.
Re: (Score:2)
I think, for this to be common, you'd really want to figure out how to leverage one of the existing layers. You might do it differently than he does, but adding a whole new layer to eliminate a toaster over is too wasteful. I think this is a REALLY neat idea, but it needs to be worked out using an existing plane.
Re: (Score:2)
I think, for this to be common, you'd really want to figure out how to leverage one of the existing layers. You might do it differently than he does, but adding a whole new layer to eliminate a toaster over is too wasteful. I think this is a REALLY neat idea, but it needs to be worked out using an existing plane.
I think you're right for the most part, but trying to think up a rebuttal lead me to a potentially interesting use case.
Suppose there's a fairly high likelihood of rework being required, either for repair or for modification during early development. Then think of what a PITA hot air rework can be - applying Kapton tape for protection, blowing components around that need to stay in place, melting connector shrouds, etc. You can't really do rework while a board is in the oven, and trying to do it while it's
0? (Score:5, Funny)
Re: (Score:2)
Re: (Score:2)
Indeed, nothing is zero ohm. I am pretty sure close to absolute zero superconductor must have some kind of residual resistance as well.
So, nothing is 0 ohm and just like nothing is absolute zero, there is always residual traces.
That being said, they do sell 0 ohm resistor which are just like a wire. It apparently saves on the manufacturing of the board:
https://en.wikipedia.org/wiki/... [wikipedia.org]
This format allows it to be placed on the circuit board using the same automated equipment used to place other resistors, instead of requiring a separate machine to install a jumper or other wire.
Re: (Score:2)
It's amazing how few people understand significant digits.
Re: (Score:2)
It's also amazing how people believe in absolute truths because they only take into account significant digits. Most of the time, they aren't even aware of the fact that they're using significant digits only! :)
Re: (Score:2)
Indeed, nothing is zero ohm.
Huh? What's your point? Are you saying a 1 inch 22awg jumper's resistant is significant in any electronics? Because the resistance of a 1 inch 22awg copper jumper (I could not find tinned copper's information) is 0.001345 ohms. That's pretty much zero with respect to any circuit that uses a 1 inch jumper. Hell the resistance of a 1 inch run of copper on the circuit board is probably an order of magnitude higher than that.
Re: (Score:2)
Re:0? (Score:5, Informative)
In laying out PCBs you often find you need to overlap connections, especially if it's a one-sided (copper only on one side) board. So you often need jumpers. Sometimes wire jumpers are used, but depending on your assembly process, wire jumpers might be very laborious and time consuming.
Long ago someone came up with zero ohm resistors that are easily handled by humans and machine insertion, for both through-hole and surface mount.
Now there are machines that can auto-insert wire jumpers, so zero ohm resistors are much less common in through-hole boards, but I still see them a lot on SMT.
Another major reason and use for them: if you need to jump over another trace, you run a huge short-circuit risk if you use a piece of bare wire. A zero ohm resistor is an insulated wire.
Re:0? (Score:5, Informative)
Zero ohm resistors? I know I'm getting old, but it seems to me we used to call those wires.
There are "zero ohm" resistors (they are not exactly zero of course). They are for when your PCB is planned to have optional resistors in series, but after testing you find out you don't need additional ohms. You still need to put a component there, otherwise it would interrupt the circuit, so you put a zero-ohm of the correct footprint.
If you solder manually you don't do that of course, you can just short-circuit with solder or a bit of wire. The zero-ohm components are useful when you have access to a reflow oven or assemble through a factory, where they won't do manual tricks, they just want component reference to program their machines.
Here are zero ohm resistance category (it has 575 products) of a popular electronics supplier: https://www.newark.com/c/passi... [newark.com].
Re: (Score:2)
Re: (Score:2)
Even funnier is they have a specified tolerance.
Re: (Score:2)
For cylindrical resistors, the ohm rating uses stripes with invisible ink :-)
Actually I think they use just the tolerance alone, because there's no way to do "0" easily. Maybe black-black-black?
Re: (Score:2)
IIRC a through-hole zero-ohm resistor has a single black stripe.
Re: (Score:2)
So, basically, they're a piece of wire with a resistor-shaped handle? Seems logical, and needed in some circumstances.
Re: (Score:2)
They're shorts essentially. For manufacturing time configurations and the like, use them instead of DIP switches, etc. Also use to isolate components used only for development, leave the 0 ohms off for production to reduce noise. Leave it off to make boot sectors read-only. Stuff like that. I think they're actually a bit easier to use than wires at times, though I can't solder either worth a damn.
Re: (Score:2)
Zero ohm resistors? I know I'm getting old, but it seems to me we used to call those wires.
It's possible you're either joking or trolling, but in case you're not I'll explain.
Even in the days of through-hole components, it was common to use zero-ohm resistors instead of wire jumpers because some automated assembly lines couldn't handle wire jumpers. Also - much less commonly - a circuit might have been designed to allow the addition of a 'tuning' resistor whose range of possible values included zero ohms.
On an SMD assembly line, adding a wire jumper is a lot more expensive than adding a zero-ohm
Re: 0? (Score:2)
I've seen surface mount zero ohm 'resistors' (with a 0 printed on them) which were really jumpers in a SMR package.
Re: (Score:2)
It depends. You can get actual zero ohm resistors - they are resistor shaped wires, basically. You can even get them in standard axial packaging where it's usually a single black stripe to indicate it's zero ohms.
More typically they're standard SMD resistors with a 0 on them. They're used a lot - often for things like configuring a board for features or to enable and disable pathways.
For example, you might have an I2C
Old news (Score:3)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Person comes in with a faulty board and asks you to fix it
Ok. You lay the board on table and apply power. Proceed to close your eyes and slowly move on open hand over the board murmuring softly. After a few seconds you point to a specific component and declare it’s faulty. After replacing the component the board works great and they are now convinced you have super powers to repair stuff. In reality it’s more fun than telling them yo
Re: (Score:2)
Sometimes that's a great way to fix something. Years ago I had a difficult repair- moderate size board with a couple dozen chips- mostly TTL. The system's power supply couldn't power it- the short was too good. So, what else do you do? You get a beefier power supply and force the issue. The bad chip actually blew a hole, which made it very obvious. Very lucky my eyes weren't in the trajectory. A few similar repairs (if you can call it that) caused smoke to shoot out of some of the leads, and discolore
Re: (Score:2)
Re: (Score:2)
That's awesome, thanks. I wish I could mod this funny.
Besides the great fun of that, I'd love to tackle fixing the circuit. Did you come up with a solution?
Re: (Score:2)
Re: (Score:2)
... just how richly purple the smoke from some mosfets can be.
Does the purple smoke create a Purple Haze? Jimi Hendrix lives!
Damn. I really am an old fart.
Re: Old news (Score:2)
When I was a kid, and playing around with electronics I really should not have been having access to, I nearly made molten lava out of a transformer from an old digital alarm clock.
I got that puppy red hot and glowing, though not at all by intention.
Re: (Score:2)
zero ohm resistors? (Score:1)
If resistance is zero they are not resisting so not resistors.
Re: (Score:2)
They are used for future circuit insertions. In manufacturing, it would take a separate process to put in a junction wire so they make these resistors to they can be placed mechanically in the circuit without adding manufacturing cost. Later the resistor can be replaced with future circuits that extend functionality.
Re: (Score:2)
Re: (Score:2)
I thought Resistance was Futile, no zero
Re: (Score:2)
Resistance is not futile. It's voltage divided by current. ... at least according to my workplace coffee mug. (My corresponding T-shirt is for the Mordor Annual 5K Fun Run: one does not simply walk.)
Re: (Score:2)
Blowing hot air on YT (Score:3, Informative)
It's not self-soldering, it's just self-heating. Getting the right amount of solder paste in the right places is the difficult and messy part. Blowing hot air is the easy part.
Re:Blowing hot air on YT (Score:5, Funny)
Blowing hot air is the easy part.
Spoken like a true Slashdotter!
Re: (Score:3)
Getting the right amount of solder paste in the right places is the difficult and messy part. Blowing hot air is the easy part.
I have found that most PCB fabrication shops will supply a laser-cut stencil for solder paste for minimal cost. Most PCB software (I recommend Kicad which is open-source and works great on Linux Mint) will generate the solder paste layer, just make sure to upload this to your fab shop* and they'll send you a lovely laser-cut solder stencil for a couple of dollars added to your PCBs. Now apply solder paste with a squeegee, add your components, then bake in an oven.
It may take a few tries to get your oven set
Pretty brilliant (Score:2)
that's why (Score:2)
they don't make double-sided pizza
Self destruct ability? (Score:2)
If combined with something that can send a voltage high enough to keep the resistant hot, this might be a way for a board to self-destruct, although someone good at reflow soldering could get things replaced. However for somewhere remote, it is a good way to ensure the board can't be used without physical access.
Re: (Score:2)
Nope. Useless as self-destruct because not reliable and actually dangerous.
Pretty much bullshit (Score:2)
Yes, it can be done. Is it original? No. Is it useful? No. You can just place the PCB on a heating plate for the same, crude effect with a lot less effort.
Re:Pretty much bullshit (Score:4, Insightful)
Yes, it can be done. Is it original? No. Is it useful? No. You can just place the PCB on a heating plate for the same, crude effect with a lot less effort.
Is it useful in the context of everyday electronics work? Possibly not. Is it cool and inventive? Personally, I think it is, and I've worked and played in electronics for about 6 decades now.
Also, it's a very good example of the hacker mindset. It's that kind of explorative and speculative "Can we do this? Let's try!" attitude which has lead to a lot of unexpected innovation. As opposed to the "Why would we do that? It's pointless and silly" attitude that short-circuits creativity and does jack shit for the cause of innovation and discovery.
Long snake ground (Score:4, Informative)
From the summary:
Many circuit boards include a grounded plane as a layer. This doesn't have to be a big unbroken expanse of copper — it can be a long snake to reduce the copper used.
This is wrong. You might get away with a "long snake" ground for a basic low-speed circuit, but if you are doing anything high-speed, you're gonna have problems.
Re: Long snake ground (Score:2)
I'm thinking this 'snake' would act like a transmitting antenna, transmitting on the same wavelength as the length of the individual long lines of the 'snake' (assuming my radio theory here is correct).
Depending on what radio frequencies your project is shitting on, you will likely get a nasty letter from the FCC, or a visit by the men in black.
Re: (Score:2)
Indeed. Reducing RFI is a major reason for a big flat ground plane that should fill an entire layer.
Some vias and a few short bridge hops are acceptable, but you don't want any long traces that break up the plane.
A good strategy on a 6-layer board is to make the two middle layers a ground plane and a power plane (typically 3.3 volts). The adjacent planes will act as a low-impedance capacitor, quieting any jitter on the Vcc and Gnd pins.
GC (Score:2)
" You might *NOT* need to buy a commercial reflow oven â" you can cobble one together from an old toaster oven,"
Re: (Score:2)
And below a certain price point (i.e. under $4-5k), those converted toaster ovens work surprisingly well. With appropriate modifications and a purpose-built controller, they're probably the best low-end solution for handling the reflow task.
Re: (Score:2)
+1.
A modified toaster oven is a great DIY project. You can learn a lot of control theory (PID is your friend, when dialled in properly) but you don't need to worry about difficult sub-second timing precision, like you would with some other common DIY projects like self-balancing robots, flying machines etc. Not to mention that when you're finished, you'll have a useful tool which will make future builds more efficient!
Just don't cheap out on the parts that you need to connect your new reflow oven to the mai
Re: (Score:2)
I said you don't need crazy timing precision for a toaster oven, then I mentioned some other popular DIY projects that do need it. I think we are saying the same thing :-)
Just learn to solder (Score:2)
Applied Science (Score:2)
Applied Science on YouTube did this in his video on making high temperature PCBs from alumina sheet. Here's the direct link to the video and time..
https://youtu.be/kxXEI0Ce6C0?t... [youtu.be]
It's a neat trick.
That this point in the video is that exact number of seconds in was unintentional and makes me happy.
Delamination (Score:2)