Have now designed and ordered the main PCB for my solder feeder. It's entirely unlabeled because my SVG outline didn't appear for some reason but that won't stop this prototype from happening.

This PCB layout looks chaotic as hell, but it'll make sense once stuff's been soldered to it - it's double-sided and has to situate stuff in precise places in a very small box.

For my own benefit in a couple of weeks: Pen buttons are pins 2/3, main buttons are pins 4/5, servo is on pin 6.

Success! Very small PCBs have arrived.

The long ones are 0.8mm thick - I think I could snap one if I wanted, but it's not going to happen by accident.

Also there are five of each and I bought plenty of everything else, so if anyone else in Australia thinks they might like a solder feeder pen, sing out.

Not my best, not my worst - but for a V1.000 of something that is unreasonably crammed into as tiny a space as I could imagine, it's come together pretty well.

This is the mainboard for my solder feeder - the 3-pin socket near the Pro Micro's USB port is the connection out to the pen itself, the 3 pins at the back end go to the continuous rotation servo that actually moves stuff, and there are four buttons - the two on the main body will be for fast feeding, the two on the pen for slow.

Well, the electronics seem to work correctly - all four buttons can make the servo run back and forth - but now there is a new problem: the servo suddenly looks and sounds underpowered, moving much more slowly than it did when everything was on the breadboard.

No code changes, and the extra electronics are just input-pullup buttons. Not sure why it'd behave differently now. A new spare servo acts exactly the same, and it's the same whether powered from VCC or RAW.

Now witness the power of this fully armed and operational solder feeder!

The power issue I think came down to a crappy Pro Micro clone board - another successfully flashed once and then permanently died, so the one in this short clip is actually the third one I've tried. It also sometimes cuts out until it's power cycled - I think it's overloading a cheap 5V regulator, version 2 will be powered from RAW rather than VCC.

This Works. Time to finish the enclosure. #3DPrinting #Electronics

The code for this thing is dead simple - it reads from four buttons, then an if-then-else tree says what the servo does.

A regular servo gets commanded to positions expressed in degrees - 0 is fully one way, 90 centred, 180 fully the other. A continuous rotation servo like this one takes the same commands, but those values become relative speeds - so zero and 180 are full speed one way or the other, and 98 and 85 are the closest values I could get working for the slow feed option. #Arduino

My completely arbitrary obsession with stuffing the entire workings of this thing into an equally arbitrary 4x4x4cm cube because that's how big a square I drew in Rhino 3D on day one of this project is... resulting in a pretty nice little clockwork-looking object.

I'm not loving the "Iron Grey Metallic PLA Metal" filament as much as I would for this though, that might be prettier for more monolithic parts but I think I want something more matte for this project in the end. #3DPrinting

Holy crap I got the entire servo's cable in there - and it's powered up and spins when I press the buttons.

It spins the *wrong way*, but that's okay - that USB socket there is the Pro Micro so it's very easily reprogrammable. #Arduino

The fully assembled solder feeder, now attached to the front of my #SolderingStation. It replaces the simple little stand I previously used to just bend the end of my solder wire over - that's still attached next to it in case I ever want it again.

I think the trick with this filament is to print with super low layer height, and enable ironing - the top of this thing looks like machined delrin or something.

Last step is finishing the feeder pen itself.

This is a failed print (the other half of the pen lost bed adhesion and became plastic spaghetti), but enough of it worked to confirm I now have the negative space cutout for the feeder pen's electronics correct enough to spend time refining the outside shape.

It started with a roughly cylindrical cross-section, but that immediately made it tougher to print - and it already is a technical print because of the thin walls and large hollows everywhere. Pen v2 will have a simpler form. #3DPrinting

Went for the smarter option of "look at the internals, then design an enclosure around them" rather than "try and stuff everything into an arbitrary tiny shape" and ended up with this extremely functional part. Still need to make the button insert, but this thing absolutely works - press the button, more solder wire appears.

Could I actually be close to - dramatic gasp - finishing a project!? Better yet, as an infrastructure project, this'll make it easier to finish *other* projects as well.

I am now extremely bored of making sub-millimetre adjustments to this tiny button part, but I am pleased to report: it works! Press the button, solder comes out; press the other button, solder goes back in.

The tube is held rigidly and almost completely straight - there is one bend near the nozzle because it has to be off-centre to run underneath the PCB, but I think that might actually help to keep it straight as it deploys.

The solder feeder pen has reached v1.0. #3DPrinting #Arduino #maker

Hocus pocus, abracadear; I would very much like the sponge that came with my soldering iron to re-appear

I guess this is a sign it's another thing I should make myself and attach to my soldering station?

I usually try not to toot my own horn too much, but I will do so when I'm proud of making something good - and my solder feeder fucking slaps.

This is the first real test done in anger, and there are definite changes to make. It needs a narrower nozzle to get closer to the workpiece, a longer one to straighten the wire better, and a more comfortable handle.

But I will never go back to just feeding solder in by hand. This is 1000% better in every way.

#3DPrinting #Electronics #Arduino #maker

Thinking about what modifications might be needed in order to hand one of these to someone else to use. And what would need to change to suit different sizes of solder - mine is 0.8mm, but I've heard 0.5, 0.7 and 1.0 are common.

The 3D printed tube anchors could be simplified to work with multiple sizes, and the TPU wheels probably don't even need a groove when the wire is guided by tubing - so maybe the only things that needs to swap for different sizes are the PTFE tubing and the pen.

From day one, I designed the thing to screw to the base of my own #SolderingStation project, which includes a spindle for a roll of solder. I could make a basic version of that to go with the feeder.

It's not hard to hook a roll of solder onto something, but the feeder will work best when it's pulling right off the top of a roll, so a bespoke 3D printed part to put things in the right position won't be much of a chore.

There's also been questions about speed control. I didn't get quite the range of adjustment I was expecting out of a continuous rotation servo - the "full" speed is maybe 1.5-2x what the pen buttons do currently - but maybe there's a beefier servo option, or scope for a more traditionally motorised version.

The wheels being as big as they are also have the effect of gearing the mechanism up, and I based their size around the bearing I had sitting in a drawer, so that could be adjusted too.

I removed the notch from the wheel model, found a really good profile for printing TPU that I've used before, and cranked the layer height down to 0.08mm for these new wheels - new ones on the left, old ones on the right.

Those are gorgeous prints, given they are only 1cm tall and 2cm across.

The notch isn't necessary because the wheels just provide linear motion in one direction or the other -keeping the wire aligned is the job of the PTFE tubing either side. #3DPrinting

This test print nozzle is firmly in the category of "things I didn't expect to go well". This is super thin - the top 10mm or so has a 3mm outside diameter, with a 1mm channel running through it - and it was printed upright, at 0.08mm layer height. It's surprisingly strong... but even if it broke, this would be trivial to re-make.

It's been hand-drilled out to 1mm, which is still very sloppy for my 0.8mm solder wire - guess I need to buy a set of very small drill bits. #3DPrinting

Part of what I'm calling power issue might actually be a traction issue - since removing the notch from the wheel design, I'm noticing the idler wheel sometimes doesn't turn when the mechanism is moving, and the solder wire is slipping as a result.

A simple fix there might be to gear the two wheels together, forcing both to turn mechanically - currently the idler is just driven by friction with the servo's wheel. That won't be too difficult to introduce I don't think.

Okay, this might be the coolest thing I've ever designed and printed. The wheels for my solder feeder are now geared, so one will drive the other, while retaining the flat zone (outsized by 0.1mm so it's the main contact point between the two wheels) in the middle for gripping and pushing the solder.

This is printed in TPU using this profile, at 0.08mm layer height: https://www.printables.com/model/552337-p1p-tpu-profile-configurations-black-overture-tpu

Will this actually work? Time will tell, now I have to rebuild the feeder base to fit these. #3DPrinting

The new tyres have, however, introduced a new issue - there's now so much power and resistance in the mechanism that the wheel rim attached to the servo is *unscrewing itself* after a few turns.

Fortunately I foresaw this forever ago and added a hole for a tiny grub screw to press up against the servo's splines! That's been there since v1 of that part. Problem solved, I think. #3DPrinting

That might be it for now, until the new servo arrives or I want to change something about the pen. Here's my current #SolderingStation setup, the solder feeder mounts to the front and the spring clips hold my Pinecil and the feeder pen when not in use so the whole thing packs up vertically and is carry-able with that handle.

I didn't actually need to put headers on this Pi Pico for any particular project, but - I'm sure you'll understand - I needed to solder *something*.

#3DPrinting

With apologies to the Whitlams:
🎶 Chunky chunky servo motor now
🎵 Chunky chunky servo motor

Some specs I found on the big one suggest it should be able to rotate about once per second under full power. With 2cm diameter drive wheels, that's about 6 centimetres of solder wire delivered into a workpiece every second - does that sound like enough for your purposes @jpm ?

Successfully designed, printed and assembled this one all in one go. Keyboard for scale.

There's a slot for each drill bit so they sit fully below the top face of the tray. Those are 3x10mm rod magnets pressed into place and they hold the bits securely enough that I could drop this tray on the ground and they wouldn't jump out. And there's enough room to intentionally dislodge and grab a bit out by hand to actually use it.

Hooray for spontaneous workshop infrastructure projects? #3DPrinting

I want a girl with a short skirt and a looooong jumper cable.

Also - ever wondered why servo motors often have a brown, red and orange cable? There wasn't some international committee that decided this based on usability or research - it's just the first 3 wires in a mass-produced ribbon like this, and that rainbow colour pattern is "borrowed from the standard resistor color code": https://en.wikipedia.org/wiki/Ribbon_cable

#Electronics

Today's task is refining the screw-on part.

Getting a nice fit is easy if the parts are printed vertically - I use BoltGen to make me an M8 screw thread, and the inner bit gets scaled down by 0.1mm in two dimensions for a really satisfying connection.

I want the pen to print horizontally, though, and be relatively flat, so I need threads I can print sideways on the bed without distortion from elephant's foot or other print artifacts. This will take several iterations to get right. #3DPrinting

Here's a behind-the-scenes on the process.

In the same way a woodworker might use a jig on a table saw to get consistent, repeatable cuts, I use a virtual jig to get consistent and adjustable subtractions on my model. The magenta thing in this screenshot is that jig - I just generate a bolt, cut it with the subtraction jig, and I have the shape I need to print.

That jig is made up of six different objects I can tweak independently to make iterative changes to the design.

#3DPrinting #Rhino3D