Turbulator - DIY Watch Parts Cleaning Machine

GuiltyGear · May 5, 2026

Hi forum. For the past year or so, I've been designing a watch parts cleaning machine with the invaluable help of @JimInOz . The whole idea was to make something compact that can fit on anyone's desk, relatively simple to build without much experience in electronics, and cheap. What I'm presenting here is the second prototype, one I'm proud to show, but plenty of work remains before it's ready for me to share the build plans. I'm mostly looking to get your feedback and perhaps answer some questions, which may provide insights on how to improve it.

The 3D-printed parts are its core, consisting of about 500 grams of PLA and about 150 grams of PETG for the basket holder and the baskets themselves. Logic is handled by a Raspberry Pi Pico, a couple of TMC2209 to control the stepper motors, and a few components like a button, screen, buzzer, and encoder. The Nema 17 stepper motors handle the carriage movement and spinning, with 2x 12mm linear shafts on linear bearings on the head unit and an 8mm lead screw. The total cost to build, at least as an approximation since I haven't done a detailed BOM yet, should be around $150.

As a single-jar machine, it handles the washing and spin-drying phases automatically. Once done, it beeps and waits for you to change the jar and start the next step. On the final step, it expects the dryer unit, which currently is simply a powerful server fan. Before starting each step, you can update the run time via the rotary encoder.

I have a bunch of improvements for the electronics and software on my list: a custom PCB, PTC heater with a small fan, a larger screen (128x128 instead of 128x64), a redesigned menu system, and an EEPROM chip for saving custom programs and telemetry. These will probably be completed in the following months.

I'll let you watch a video of the machine in action and a few photos. Tell me what you think!

JimInOz · May 5, 2026

Doesn't look too bad for an amateur 😉.

Deafboy · May 5, 2026

That's really nice!

Archer · May 5, 2026

Looks great so far!

Endshake · May 6, 2026

Thank you for investing time in something like that! I love the name😄

GuiltyGear · May 6, 2026

Endshake
Thank you for investing time in something like that! I love the name😄

@JimInOz coined it.

ken_hodson · May 6, 2026

That looks fantastic, I would definitely be interested to build something like that, are you planning a Kickstarter project to buy the plans/print files?

GuiltyGear · May 6, 2026

ken_hodson
That looks fantastic, I would definitely be interested to build something like that, are you planning a Kickstarter project to buy the plans/print files?

Actually the idea is to have all the plans and files free for everyone to use. This is a pet project, not something I'm looking to make me rich. If the project gains some traction and I figure any monetization strategy, then all the better.

For now I'm focusing on making a "final" build, hopefully in the next 3-5 months.

HIXTAR · May 6, 2026

Looks great

HadgeBadge · May 7, 2026

This looks fantastic! I recently bought a very dirty vintage Elma Super Elite that needs all sorts of work like a new basket, rewiring, new jar lids and so on for just over $150.

I expect it to still cost me another couple of hundred dollars at least to get it back into suitable, functioning order. Never mind the time I need to find to take on yet another project.

Your project looks very appealing. Neat, compact, accessible, and cost effective.

Thank you for sharing.

rendo · May 7, 2026

This looks great! I have an L&R Master, but I would definitely build this as an alternative.

A few questions:

- Is the connector on the base of the unit intended for the dryer fan?

- Have you tested PETG's durability when exposed to strong cleaning solutions? It would be easy to reprint a new basket, but I'm wondering if a steel basket inside a PETG frame would also perform better for the drying cycle.

- Related to the previous question, without a lid, I expect the top of the unit (motor housing) to be exposed to fumes during the cleaning cycle. Have you tested how PLA would hold up to this exposure in the long term?

ErichKeane · May 7, 2026

Ok, so there is a TON to love about the design above. I appreciate the automated nature, particularly the spin out. I DO have a handful of suggestions if that is OK.

Jars:
First: TBH, I think you've chosen poorly on the jars. The chemicals (particularly the L&R) STINK, so having jars that seal is very important. While the L&R ones are typically rubber gasket lids (though they DO have screw on lids), the design you have requires a lot of handling of the jars. Spilling/aggitating the fluid is going to happen. I'd suggest seeing if you can find a similar OTS jar that comes with a screw on lid. Also, a sealed lid is a necessity for storage.

Second: for similar reasons above, you NEED to make your moving unit 'seal' against the top of the jar. If it doesn't, you'll agitate the fluid and
it'll smell horrible in your room for days. It also will prevent splash-out in cases where you've messed up 😀

Baskets:
First: I think the 3d baskets are vastly inferior to metal ones. I have had both, and the amount of liquid that gets in/gets spun out in the plastic ones is significantly worse. I'd suggest designing around the inexpensive/available elma-style baskets. I currently have this set (https://www.aliexpress.us/item/3256810257154917.html) which is fantastic, albeit expensive. They are vastly superior to others (the interlocking baskets are REALLY nice). I have some of these on order (https://www.aliexpress.us/item/3256804704686437.html?), 1 of each 'style' (H,X,O) that I think are the same thing, and are about 1/2 the price for a set. I realize this is a huge addition to your BOM, but it might be nice at least to design the basket holder with the right dimensions to allow for an upgrade.

Second: Basket space is ALWAYS the limiting factor. I find even the set of the 3 above is pretty limiting, to the point I do 2 runs per watch on more complicated ones. So when I see you only are 2 inserts high, I can just imagine 2-3 runs per watch being necessary, which is unfortunate. I'd very much see if you can spare a little room to get it a 3rd level high. That said, I DO see the advantage of using so little fluid with only 2!

Drying:
So this is one I never realized until I owned a machine with it, but heated/spin dry is a game changer. I thought I could spin dry only, and that would be enough, but it really isn't. Including a heating element to blow it dry is probably going to be a necessity.

Ultrasonic:
Last thought: The ultrasonic on mine is a HUGE difference maker. I can tell looking at parts whether I've forgotten to turn it on or not. To the point that the ultrasonic on/off is a bigger difference than clean/dirty chemicals. Its a touch of a shame that you've not included one. I don't have a good way of including one in your design without increasing the Z dimension a ton unfortunately though, so this might just be whinging.

All that said: you've got most of a great design there, and I look forward to seeing it progress! The above list is more or less in order of importance IMO.

GuiltyGear · May 7, 2026

rendo
Is the connector on the base of the unit intended for the dryer fan?

Yes. Right now it's magnetic pogo pins, that works great as the dryer unit finds its position almost by itself. But I'm looking to replace it with a non magnetic connector and maybe a sort of simple bayonet mount so it locks.

rendo
Have you tested PETG's durability when exposed to strong cleaning solutions? It would be easy to reprint a new basket, but I'm wondering if a steel basket inside a PETG frame would also perform better for the drying cycle.

That's a very valid concern. PETG is relatively inert, but harsh solvents will make it brittle. Even long exposure to IPA. The alternative is Polypropylene, which is indestructible, but that thing is a pain in the a$$ to print. I can do some tests with what I have, but if @JimInOz can help here in putting some PETG cubes of various sizes in a couple of cleaning solutions, that'd be really great.

rendo
Related to the previous question, without a lid, I expect the top of the unit (motor housing) to be exposed to fumes during the cleaning cycle. Have you tested how PLA would hold up to this exposure in the long term?

I plan on making a lid. Fumes will hurt the PLA and electronics, so it's a necessity. I'll give some more insights when I reply to @ErichKeane message.

GuiltyGear · May 7, 2026

@ErichKeane really appreciate the feedback. Let me respond to your points as well as I can.

Jars. I've gotten some cheap jars with a sealed lid. Tried leaving 2 jars, one with water and one with IPA, for about a week, and didn't notice any evaporation. So the silicone gaskets seems to do the job. I'll give it a try with some naptha, since that thing is relatively stinky, and see if there's any fumes escaping.

Lid. My problem with the lid was that I have a coupler from the motor axis (5mm) into another 5mm shaft, and finally into the basket holder coupler. A stepped approach that doesn't leave space for a lid. I've order a 8x5mm (ODxID) seamless capillary tube, which I plan to cut to length and drill a couple of holes. Then 8mm shaft adapters, one on the motor axis and another on the basket holder (like the current design). That way I have a seamless run from basket adapter to the motor, which will let me do something like the following (hacked it together in Fusion).

Baskets. The basket holder I designed from measurements and references of an Elma basket that @JimInOz provided. Problem is that I haven't validated it against actual baskets, so I may take your advice and order a couple. That's a very good suggestion, since it allows to either buy or 3D print them, depending on budget.

Basket space. I've been struggling with that. Initially I had 25cm shafts which were tight even for a 2-basket holder. Then I upgraded to 30cm shafts on standoffs, like it is now, and that left more breathing room. My jars are 10x12cm, and the 3-basket holder (which I've already printed) has about 42mm in height that needs to be submerged. So it may actually fit. I'll try that out with a longer shaft on the motor and see how that works out.

Drying. Right now it's a "turbine", a 40mm high fan that could be used as a leaf blower. My initial idea was to use a 80w PTC heater with an integrated fan. That thing takes ~10 seconds to come to temperature and [when warm] only pulls about 3 amps. But when cold, it may try and pull 10 amps, which simply tripped my power adapter. Plan is to have a MOSFET drive the heater at 20-30khz pwm (on/off) for the first 10-20 seconds, so it can't pull that many amps, but pulls enough to get warm. If that works out, it'll be great really.

Ultrasonic. I only know that those transducers are huge and they need a big controller on mains power. Any way you see that working here, maybe as a separate module? Ignore the z height.

Rubber feet. You mentioned this on our private conversation, but let me post the photo here. I designed the base lid with some cutouts for 2mm thick (or thicker, doesn't matter) silicone or rubber sheets. I didn't do that great of a job cutting them though. They do help a lot in keeping the machine steady. Weight also helps. The thing is 2.1 kg without jar.

ErichKeane · May 7, 2026

GuiltyGear
@ErichKeane really appreciate the feedback. Let me respond to your points as well as I can.

Jars. I've gotten some cheap jars with a sealed lid. Tried leaving 2 jars, one with water and one with IPA, for about a week, and didn't notice any evaporation. So the silicone gaskets seems to do the job. I'll give it a try with some naptha, since that thing is relatively stinky, and see if there's any fumes escaping.

Lid. My problem with the lid was that I have a coupler from the motor axis (5mm) into another 5mm shaft, and finally into the basket holder coupler. A stepped approach that doesn't leave space for a lid. I've order a 8x5mm (ODxID) seamless capillary tube, which I plan to cut to length and drill a couple of holes. Then 8mm shaft adapters, one on the motor axis and another on the basket holder (like the current design). That way I have a seamless run from basket adapter to the motor, which will let me do something like the following (hacked it together in Fusion).

Baskets. The basket holder I designed from measurements and references of an Elma basket that @JimInOz provided. Problem is that I haven't validated it against actual baskets, so I may take your advice and order a couple. That's a very good suggestion, since it allows to either buy or 3D print them, depending on budget.

Basket space. I've been struggling with that. Initially I had 25cm shafts which were tight even for a 2-basket holder. Then I upgraded to 30cm shafts on standoffs, like it is now, and that left more breathing room. My jars are 10x12cm, and the 3-basket holder (which I've already printed) has about 42mm in height that needs to be submerged. So it may actually fit. I'll try that out with a longer shaft on the motor and see how that works out.

Drying. Right now it's a "turbine", a 40mm high fan that could be used as a leaf blower. My initial idea was to use a 80w PTC heater with an integrated fan. That thing takes ~10 seconds to come to temperature and [when warm] only pulls about 3 amps. But when cold, it may try and pull 10 amps, which simply tripped my power adapter. Plan is to have a MOSFET drive the heater at 20-30khz pwm (on/off) for the first 10-20 seconds, so it can't pull that many amps, but pulls enough to get warm. If that works out, it'll be great really.

Ultrasonic. I only know that those transducers are huge and they need a big controller on mains power. Any way you see that working here, maybe as a separate module? Ignore the z height.

Rubber feet. You mentioned this on our private conversation, but let me post the photo here. I designed the base lid with some cutouts for 2mm thick (or thicker, doesn't matter) silicone or rubber sheets. I didn't do that great of a job cutting them though. They do help a lot in keeping the machine steady. Weight also helps. The thing is 2.1 kg without jar.

Jars: yeah, i think that seal would be fine. They are way better looking than the 3d printed L&R ones I have, which do a decent job. I'm a touch concerned about the lack of screw-top (as I suspect these will be handled a lot), but it is perhaps my clumsiness speaking more than anything 😀

Lid:
One option instead is to use a thin piece of flexible plastic with a rubber/silicone gasket under it. Let the sheet 'bend' upwards as you descend. That way it would cause its own seal. You could probably then just mount it to the head.

Drying: Yeah, the heating elements are all pretty high draw. I had to order slow-blow fuses for my L&R else it would pop them in the first few seconds.

Ultrasonic: Yeah, you need a transducer. AND you need a way to transfer the waves to the liquid in a way that causes 'resistance' to the transducer.

As far as the controller, there ARE smaller ones, but they are definitely sizable, though as you said, you could do it as a 'power brick' type deal. If you see the one on my L&R, it is mounted to the head and gets put into the jar (and has to touch the liquid to work!). The other way is to make the base under the jar of SOMETHING that will transfer the energy, I didn't do any real work on that. PLUS you'd have to find a way to build the transducer in...

A bit of a 'start over', but if you built a lot of your current mechanism INTO a cheap pre-purchased ultrasonic (such that the plastic mounts to the 'top' of the ultrasonic), you could just use the factory bits to do that (and it would be your base). Perhaps remove hte controller panel on the ultrasonic and replace it with the ras-pi/display.

Again, perhaps a total redesign and letting feature creep kill it 😀 I suspect you'd get most of the benefit by doing an ultrasonic step before the initial clean.

GuiltyGear · May 7, 2026

@ErichKeane I'll do some research and see if there's any angle to approach this. Appreciate the input.

sheepdoll · May 7, 2026

I needed to replace the jars on my L&R what got broken in a move. Used some jars from Hobby lobby and made swirl breakers for them. These have the silicon gaskets. I also got a small one for the ultrasonic, as I did not like filling the whole tank with cleaning solution.

One project I was working on, was an automatic film processor for the obsolete APS film. Where the developing fluids need to be kept at temperature.

If I was designing a cleaner, I would use a modern micocontroller and have multiple jars with a system to automatically move the basket (or jars.)

When I was researching the film processor, someone made one from LEGO. Was a linear track system that lowered the film reels into the solutions. Something like that could work for watch cleaning and would not need to be in a light proof box.

Many film processors use pumps to move the chemicals into the main tank from the supply tanks. There is also a large rinse tank since the chemicals can not be cross contaminated or else they get neutralized.

There are also spin agitators what can be used on standard tanks.

As for baskets. My old one uses a really simple bayonet connector. The frame is stamped sheet metal. The baskets formed from woven stainless mesh. Not sure I would want to use a plastic 3D printed one.

GuiltyGear · May 7, 2026

sheepdoll
If I was designing a cleaner, I would use a modern micocontroller and have multiple jars with a system to automatically move the basket (or jars.)

That's fair, but all solutions out there do exactly that. Including DIY ones like the Uberkiwi, which costs around $1k to build.

My approach was to try and solve the single jar machine problem, for people that need this occasionally (or as a backup/specialized use) and can't justify the cost or space. Right now you can either buy one of those Chinese battery powered motors on a lid, or hack up a laboratory agitator. None a good middle ground imho.

ErichKeane · May 7, 2026

GuiltyGear
That's fair, but all solutions out there do exactly that. Including DIY ones like the Uberkiwi, which costs around $1k to build.

My approach was to try and solve the single jar machine problem, for people that need this occasionally (or as a backup/specialized use) and can't justify the cost or space. Right now you can either buy one of those Chinese battery powered motors on a lid, or hack up a laboratory agitator. None a good middle ground imho.

Yeah, agreed. Right now, other than some cat-in-the-bag used machine, you're stuck at about a $800 cheap indian-made machine (I think? The Miracle/Supreme brand). The only thing cheaper than that are those horrible battery powered lids (absolutely awful, don't buy them!), or the $150 hand-cranked one.

Something that came in ~$300 like the above + some mild improvements would be a sweet spot that would make most hobby watchmakers happy.

Heck, I probably wouldn't have the machine I did now (thanks to footprint!) if there had been something like the above available.

sheepdoll · May 7, 2026

Yeah, they do take up a bit of space.

Turbulator - DIY Watch Parts Cleaning Machine

GuiltyGear

JimInOz

Deafboy

Archer

Endshake

GuiltyGear

ken_hodson

GuiltyGear

HIXTAR

HadgeBadge

rendo

ErichKeane

GuiltyGear

GuiltyGear

ErichKeane

GuiltyGear

sheepdoll

GuiltyGear

ErichKeane

sheepdoll

Similar threads

Adding Elma Style baskets to a L&R Cleaning Machine-

New(old)Tool day! L&R Console Watch Cleaning Machine

New addition, a Cheap Cleaning Machine

Vintage Watch Cleaning Machine

Wanted: Automatic Watch Cleaning Machine... and advice