Minamil: a minimal CNC mill. And friends.

Each axis: ̶$̶5̶ ̶$̶8̶ $10 motor+lead screw, 3x LM6UU, 3x 6mm x 100mm rod, 1/8in hardboard, PC case screws

Public Chat
Similar projects worth following
Developing a cheap small 3-axis CNC mill for $̶1̶0̶$15/axis.

Powered by:
$̶5̶̶ $̶8̶ $10 motor+lead screw CH-SM1545 (price rising - hello supply risk)
Linear bearing cost minimum at 6mm (smaller costs more)
1/8in hardboard is practically free per small area
#6-32 x 1/4in computer case screws are practically free
3x 6mm x 100mm rods per axis can be $1 each or 6mm rod is common in printers etc.

It's working. Selected log entries linked from "Details" below.

(update: killed the counterweight; CNC Kitchen video)

CNC Kitchen video - thanks Stefan!

Even though the Minamil is minimalistic, the part quality Paul can achieve with it blew me away! writeup - thanks Bryan Cockfield!

Hackaday Podcast 136 - thank you @Mike Szczys & @Elliot Williams for the kind comments!

"I'm frankly shocked by the quality of the parts coming out of this" 11:54

Hackaday Prize 2021 finalist! - thank you judges!

(mocking the style of over-hyped package labeling, in case that got lost in translation)

Many of the images and videos here show a big counterweight rigged off the side of the side of the little machine. But nevermind. That's just (7 July '22) become history. Fresh rev of the Z axis does better with less. Yay for maximizing minimality!

This little CNC mill works well enough to produce an eye-candy demo video:

(that's a "new" (June 2022) demo vid; here's the first eye candy demo

...and mill fine-pitch circuit board traces:

What little CNC mill?

Minamil: a minimal CNC mill

First there was #CDCNC, a highly-constrained just-barely-functional one-off toy built by improvisation with found junk and simple tools.

Here I'm developing an idea that came from thinking about whether or not there is any space between a dumb stunt like CDCNC and a commercially (i.e. efficiently) produced entry level CNC mill/router. Early results seem encouraging.

In contrast to CDCNC, this is about reproducibility from a simple BoM, economical access to laser cutting, and instructions for building your own sub-mini CNC mill.

Next there may be #"Desk Accessory" CNC Milling Machine. For progress in that direction, with not much more effort in the frame+enclosure department:

(different Z axis in those two pics - part of the not-quite-so-minimal "and friends" part of the project title that I haven't written much about yet) 

Minamil to go
click image above for video of setup/packup

Telescoping axes allow the enclosure to fold up smaller when parked and expand for operation:

Selected Log Entries

  • 1 × XYZ mechanics follow "Parts source / cost notes" link under "Selected Log Entries" in the Details section
  • 1 × Electronics
  • 1 × Laser-cut parts from 1/8" (3mm) hardboard
  • 1 × frame & counterbalance
    • crude or fancy - first was very rough scrap
    • follow "Build!" link in Details & scroll for simple frame drawing
    • follow "packaging" link for slightly more elaborate example
    • counterbalance: 500ml bottle, beam, pivot, string (flat parts include a hook)
  • 1 × rotary tool
    • included clamp fits Dremel®-like tools with 1-7/8" cylinder bodies like canonical model 395 type {n<=5} or current models 100 & 200, and many clones
    • or adapt removable clamp for something else

View all 6 components

  • ... still sorting Z axis repeatability ...

    Paul McClay6 days ago 2 comments

    tl;dr: thermal expansion ... and skipping steps upwards?

    Wringing repeatability from this counterweight-less cheap-motored Z axis redesign has stalled me for a while.

    Here's why I'm still trying:

    Microns. That shows <10 μm variation for 720 Z probes over three hours. Each ~1 hour block of 240 touches starts with a long up/down excursion. Indicated steps alternate between 1 and 2 μm because the actual microstep size is ~1.5 μm. The three downward spikes are the first touches after the long excursions and could be avoided by adding a small up/down hop after long up/down moves.

    Other than those avoidable spikes, and a ~5 μm drift over ~20min (~80 probes) in the second hour, that looks like micron scale repeatability over usefully long time spans and cycle counts. Relative to a target of maybe 25 μm variation (match X&Y full step), that looks very encouraging and worth pursuing.


    That 3hr stretch of wunnerful repeatability is cherry picked from data like this:

    first big graph
    first big graph
    second big graph
    second big graph

    ...preceded by a bunch of manual data collection, preceded by a bunch of random thrashing trying to figure out what's the deal.

    • The two graphs show about 30k points ((9k + 6k) x 2).
    • Both are discontinuous;
      • the first more so than the second.
      • Only the one large-ish gap in the first graph represents elapsed time between sampling runs.
    • Red/lower points: touch down/make contact
    • Blue/upper points: lift off/break contact
    • I can guess at causes for some but not all of the variation of difference between the two

    Generally, I've been seeing that short term repeatability can be, and often is, very good -- and long term repeatability can be, and sometimes is, acceptable (relative to a target of 25 μm which is somewhat arbitrary but comparable with the full step stride of the X&Y axes and aimed at cutting just through ~35µm nominal thickness of 1oz copper cladding that isn't flat).

    Some of the runs in the first graph show strong similarity in fine structure:

    I've pretty well convinced myself that the wiggles in the initial downward slope reflect non-uniform microsteps. The distance/step cycle runs through two full steps. At least two of the 28-BYJ48 motors I've look at -- closely enough to notice -- show significant asymmetry between even & odd full steps. Haven't dug into that much yet, but for now I'm counting two full steps as the practical resolution of these motors. In the latest rev. I've shrunk the pulleys again to make two full steps move about 25µm for parity with the X & Y axes. (Finer string -> more turns fit -> less motivation for big pully diameter; more about string below.) These data were measured with 8x µstepping which is where the 1.5µm steps come from. While that's impractical "resolution", I'm much surprised that this marginally constructed device actually distinguishes single µsteps at ~16k steps/rev. While kinda neat, that's not really useful and as of right now I've dialed that back to half-stepping (~6µm/step).

    Thermal expansion(s)

    Apart from the finer wiggles, I think the general trend looks like different parts heating up differently. I'm guessing that the fast initial slope reflects the pulley heating quickly. Then the slower up-slope reflects the XY table growing taller as it heats up less quickly. That's a little less speculative because I can test that by unplugging the X&Y motors. The second-to-last spike-drop-slow-rise in the first big graph dropped like others then turned back up less quickly with the XY motors unplugged, which seems to corroborate that idea. The fact that there's still some slower exponential-like change happening suggests something else warming up. I'm guessing that it's the bulk of the Z axis heating up. Maybe changing the relationship between the stepper axle(s) and the tool clamp. ?.

    Anyhow, this kinda sorta validates the idea that the Z probe data...

    Read more »

  • Byproduct of debugging Z repeatability

    Paul McClay6 days ago 0 comments

    More ornamental than functional, but skimming ~50µm from copper-clad board still tests control of cutting depth.

    Copper figures "over" (but not really over) blue acrylic paint. Made by first cutting the blue parts, painting, cutting the background, then sanding the paint off the remaining copper. Sanding carefully -- the first try ended with entirely removing the copper from one spot while trying to clear a minor blemish from another spot.

  • Z v3 -- counterweight delete -- progress

    Paul McClay08/17/2022 at 22:06 0 comments

    From last week's episode...

    The result is surplus torque and steps/mm, so the next rev can use a larger diameter to get more (all?) of the string wound onto the drum in a single helix of more nearly uniform diameter. Currently the steps/mm changes when the string fills the (thin) drum and winds back over itself [...]

    So... yes.

    Z v3.01

    In this iteration:

    • larger diameter lift winch drum per last log -- this combination of drum diameter, material thickness and lift cord gets the full range of motion in a single layer of wraps (about two) for more uniform steps/distance
    • space for a redundant hoist -- one motor still works but the chain of lightly constructed single points of failure holding the heavy/spinny/cutty part up from freefall might be too much minimality even for this project
    • various other fixes/tweaks

    (...and that little screw on the right frame side -- for stowing the counterbalance rig -- can go away now)

    While assembling this rev I made a short list of minor tweaks for one more iteration that should be ready to share (because optimism). 

  • Z v3: counterweight delete (& other wins)

    Paul McClay07/08/2022 at 00:07 3 comments

    tl;dr: it works -- scroll for gear example

    The proto-concept for this project had a weak Z axis moving the heavy tool horizontally. For reasons, the idea got switched to conventional vertical Z with a counterweight helping the weakly driven axis. And so it was, from then until a post-MRRF epiphany about how my whole concept of Minamil's Z axis has been broken since that early reorientation.


    Here's a cheaper, smaller (vs v2.x), stronger Z axis that doesn't need any counterweight:

    Because gravity is free anti-backlash (for a << g), and there's room up there to bodge on a bigger motor.

    The last log entry showed a lashed-up proof of concept but not yet any attempt to see if it would actually work in anger.

    Since then: to CAD and laser cutter... Largely because I'd already planned to try a different arrangement of the slidy parts.

    Behold: Z v3

    Trading a ~$10 stepper+leadscrew unit for a ~$2 geared stepper, the ubiquitous cheapous 28BYJ-48:

    Z slide v3.beta
    longer throw, too :)

    The motor turns a laminated (big disk/little disk/big disk) winch drum to haul the carriage up a bit of string that seems to be sufficiently unstretchy.

    The result is surplus torque and steps/mm, so the next rev can use a larger diameter to get more (all?) of the string wound onto the drum in a single helix of more nearly uniform diameter. Currently the steps/mm changes when the string fills the (thin) drum and winds back over itself -- but that won't really matter until someone tries cutting much thicker stuff than I have so far.

    Result: yes

    For a first "real" test, I tried cutting one of the little bevel gears from the recently shrunk demonstration project:

    even step-down
    little bevel gear -- 7.6 mm diameter

    Those look like encouragingly uniform 50 μm step-downs. That's really the main point of this log entry.  ☺

    Besides losing all the counterweight rigging, this version gains some other wins:

    • reduced cost
    • top of the machine is solid when not holding the rotary tool  (vs. moving/fragile parts)
    • longer range of motion (>70mm -- obstructed to ~66mm by limit switch in this rev)
    • smaller/shorter than V2.x slides
    • stiffer than V1.x slides, and I think not much less stiff than V2.x
    • operates without tool mounted and counterweight rigged (or a finger under the empty carriage)
    • option of using two longer rods vs 3x 100mm rods, or one long+one short; four bearings on two long rods should better resist twisting of the tool carriage
    • long rods, if used, supported in middle to maintain stiffness -- otherwise prolly too long for 6mm

    For reference/comparison/contrast:


    (lame photo)

    Uses same slide mechanics as X & Y. Most compact.


    (also lame photo)

    Mechanically based on two copies of the X/Y slide in tandem. Because that seemed like a thing to try. Much larger than v1.x. Used relaxed size constraint (XY axes limit minimum footprint area) to spread out the bearings for greater stiffness - probably well into diminishing returns. Lower two bearings spread 1/3 wider. Upper bearing much higher above lower bearings relative to v1.x, especially in upper range of motion. Room for two motors+screw units, but didn't try/need that in Z with counterbalance and mild acceleration. Added a lot of height overall. The long tail of the moving tool carriage stuck up vulnerably from the top, while the "telescoping" configuration didn't add any benefit for this axis -- reasons why the next rev was going to drop the telescopiness for a less unconventional arrangement of bearings in a shorter tool carriage running on axially offset rods in a similar tandem base.

  • #MRRFX and whY rethink Z

    Paul McClay06/27/2022 at 23:48 4 comments


    Last weekend I parked the current iteration of this project on a table at #MRRFX for a day. MRRFX being the tenth (Xth) gathering of the Midwest RepRap Festival (MRRF) "the largest gathering of 3D printer enthusiasts in the world!". There I met some really neat people, learned some stuff I didn't know I didn't know (busy day for Dunning & Kruger), and realized that I'm doing Z wronger than I thought.

    I have only casual acquaintance with 3D printing and nothing of the sort to show, but teh Internetz said a few mill/router projects had crashed the show in the past without scandal. This year Joe Spanier brought his aluminum-eating Milk Cr8, and Duet3D had a colorful router running their controller(s) -- all news to me.

    The trip was kind of an impulse idea decided only a few days before and prepped in haste. I wasn't at all prepared to be socially competent or retain names and contact info for many of the people who stopped for substantial conversations -- about this and about their own projects and experience too.

    soft threads

    Prompted by meeting Stefan Hermann of CNC Kitchen, I checked out some of his work including his 2019 review of threaded inserts for 3d prints where he measured torque-out and pull-out strength of several different types of heat-set brass insert -- including none. For "none", he drove screws into bare holes in relatively soft printed material. He confirmed expected low resistance to torque-out -- a reason why inserts exist. But, contrary to expectation, he measured pull-out strength nearly equal to the better inserts and several times better than the simplest/cheapest insert.

    Most of the minimized design of this little CNC project is held together with short machine screws run edgewise into soft "hard"board -- which seems like it shouldn't work at all. The very low resistance to torque-out i.e. very great vulnerability to stripping threads out of hardboard viscerally reinforces the expectation of weakness. Stefan's finding that un-reinforced threads in relatively soft material can usefully resist pull-out even when vulnerable to easy torque-out gives somewhat better rationale than simply shrugging at the unexpected result that the method of assembling these "Minamil" designs apparently works.

    A consequence of low resistance to torque-out -- i.e. ease of stripping threads -- is that while screws can be used to hold parts together, they can't be used to used to draw parts together and generate clamping force in the usual way.  Most of the screws in one of these linear slides hold the clamps that hold the rods & bearings. A very early lesson in working with this design was to not use the screws to close the clamps but instead to mash the clamps directly then carefully turn the screws down to hold them. Also it becomes Very Important to take care to re-engage existing threads when removing/replacing screws in previously formed threads.

    Z realization


    We don't need no stinking counterweight!

    but first, some history...

    (or skip to last heading)

    I first used one of the prototypes of the slide used for X & Y because it was handy to try. Then I used the same slide as X & Y and added a tool clamp to it because it was already done in CAD. That's the currently published design, it works, and later developments have not compellingly surpassed it, so that's ok. But it's not really a satisfactory endpoint because there's no reason for the Z axis to mimic the X+Y axes when it does a different job under different constraints. (for completeness: the zeroth Z axis was a fixed stand over the XY table)

    Gravity, obviously. The counterweight arrangement got me over that hurdle and I mostly stopped thinking about it. Until recently when... I'll come back to that down the page.

    Cutting forces at the tool tip act on the Z slide through a pretty long lever, which ups the ante for stiffness. The size and clearance requirements of the XY stage set a minimum footprint. There's...

    Read more »

  • Runout reduction revisited

    Paul McClay05/18/2022 at 02:15 0 comments

    Poorly controlled runout has haunted this project since first musings about its silly predecessor. Pulling back from an earlier somewhat helpful idea seems to have helped to escape a local maximum. At least for parts that can be finished with a single tool.

    To prove the pudding, an even smaller differential gear set   ̶i̶n̶ ̶p̶r̶o̶g̶r̶e̶s̶s̶.̶.̶.̶   done:

    I prolly should add something here about reducing runout with cheap dremeloid clones...

    Already kinda glossed the method in this log. The main thing since then is just getting better at getting more predictable results with less effort. Also I partially undid an earlier attempt to constrain the poorly constrained collet, which made it easier to just push it around until it points in the right direction. 

    The "in progress" pic:

    Done with a cheap generic DremelⓇ-like rotary tool.

  • not yet understanding how to drive steppers

    Paul McClay01/31/2022 at 05:30 6 comments

    (...and: are A4988s clocked?...)

    aspirational statement: This project includes clear guidance for configuring hardware and software to drive these leetle steppers as hard as they'll go without cooking them.

    current status: ooh... pretty pictures...

    This lack of actual useful progress reflects   ̶s̶o̶m̶e̶  much digression into scoping small differential signals riding large common mode steps with fast (20nsec) edges — in order to measure voltage across a low value resistor as proxy for current (i.e. "current shunt"). Because no current probe.

    New skill unlocked: "gimmick". Because big fast common-mode edges. 

    constant off time PWM current traces
    just a pretty picture

    Yes, I'm still fussing with A4988s

    For the last several years and currently (early 2022), generic Pololu-style A4988 modules are the cheap current-limiting dir+step-signaled stepper driver, and this is about what can be done with those rather than what to get instead.

    The generic modules I'm using are wired for full-time mixed decay mode (Rosc tied to GND). I've had them configured for 8x microstepping for most of the time that I've been using them because that seemed to work best in naive initial trials and I have not yet revisited the question. 

    Some of this may hold for some other somewhat similar drivers.

    Atypical motors. Typical results?


    For this CNC project I'm using much smaller motors than pretty nearly any other CNC project ever:

    ChiHai CH-SM1545
    same 15mm motor as many CD/DVD/optical drives but M3 lead screw

    I expect that load characteristics will affect driver behavior, and that this is an atypical load. Anyone who feels like checking any of this against a more commonly used sort of motor, please comment!

    A naive idea: measure Δt by ΔR by ΔV for constant current

    And thereby work out how much current won't cook the motor.

    tl;dr: so far it looks like that may be a fair way to measure stator (coil) temperature but neglects eddy current heating internal to the rotor (magnet).

    I tried something like this while playing with 15mm motors out of CD/DVD drives and made some progress by measuring coil R promptly after shutting off power. Resistance was already falling by the time my DMM stabilized on the falling value, so ? what it was while powered. I also tried measuring RC-filtered voltages across a coil or current shunt, which seemed plausible but uncertain considering I-don't-know-what pitfalls of RC vs RMS and floating a meter at ~30kHz. Shrug. I got my A4988-based drivers dialed to a current & μstep count that worked and left a lot of questions for later. 

    Some days ago I came back for another swing at questions left for later.

    I'd like to come up with a recommendation for Minamil-like applications and to see what can be gained from "better" tuning of current, supply voltage, microsteps, acceleration, max speed, and possibly some heat sinking — and how to get there using simple/cheap/common instruments in line with the theme of this project. To sort this out, and validate any "simple" measuring scheme, I'd like a way to see some "ground truth" of what's really happening. 

    Oscilloscopes are great for watching voltage signals, but watching current is a different problem. One way to watch current is to pass it through a low-value resistor ("shunt") to create a small voltage drop proportional to the current and small enough to not significantly disrupt whatever you're measuring. That turns current into a simple "single-ended" voltage signal if one end of the shunt is tied to a stable reference, like ground, as are neither end of a full-bridge–driven motor coil. The "correct" solution is a current probe. But the probes used in the linked example cost $5k each and belong to the author's employer.

    Measuring voltage across a shunt when both ends are moving requires measuring the difference between two similar signals. Most scopes can invert one of...

    Read more »

  • Hackaday Prize 2021 Challenge 5 contest entry notes

    Paul McClay11/04/2021 at 03:43 0 comments

    Hackaday Prize 2021 - Challenge 5 - Reactivate Wildcard

    Have an idea that fits the theme, but not a specific challenge?

    Well, actually...

    (this log entry copied here from the contest entry)

    Wildcard: to refresh or redefine the technologies we know and love in order to create a brighter future for all

    A technology problem facing people today...

    Although we may be getting over it, COVID-19 closures and remote learning have been pretty hard on anything that might have smelled like shop class or hands-on practical education, including opportunities to practice CNC programming and machining.

    Personal 3D printing has become common. While laser cutters remain costly to own, their straightforward operation can work well as a commercial service or community resource.  Meanwhile, "Subtractive" CNC milling/routing, the CAM part of CAD/CAM for 70 years, remains relatively less accessible. Obstacles include cost, space, mess, and complexity. Potential benefits of access to CNC milling include working more kinds of material with greater precision, and, for tackling the complexity, entree to a valuable career.

    ...and an idea of what a solution might be

    This project aims to significantly reduce the entry cost of personal CNC milling by way of an apparently unprecedented combination of:

    • assembly from ready parts vs. fabrication
    • lowest cost
    • good precision
    • table-top use and shelf storage in home vs. fixed installation in garage or shop

    And also some more apparently novel (?) spin-offs with wider utility:

    • cheap, strong & precise linear slide mechanism
    • compact screw-in-tab joint for laser cut flat parts
    • cheap effective anti-backlash from laser cut flat parts and a paperclip
    • cheap effective clamp from laser cut flat parts and string
    • an apparently little-known but important limitation of shallow pitch lead screws

    a couple of assumptions

    Let's take care of this up front. The "lowest cost" premise includes a couple of assumptions -- or selection criteria for identifying circumstances where this makes sense.

    • "Free” or low cost access to a laser cutter
      • “Free” may mean zero or small incremental cost above a fixed cost hac/makerspace membership, school tuition, or public library millage already committed
      • Commercial laser service fees may be impractical for the current design of flat parts, but there is a lot of room to optimize for quicker cutting time (a primary cost driver) and even a small quantity order can significantly reduce unit cost
      • As a personal conjecture: I think it is likely that laser cutting is more accessible than shop tools like a drill press or table saw for a significant and increasing number of people
      • Where different fabrication options make more sense, differently made structural parts could carry the same low cost mechanical concept of this design 
    • Use of a classic DremelⓇ-like handheld rotary tool.
      • I propose to leave that out of the project cost because it’s a general-purpose tool that readers likely to pursue this project either already have, or can use this as the excuse they’ve been waiting for to get one
      • The tool clamp may be adapted for tool shapes that differ from the classic DremelⓇ cylinder profile with 1⅞" & 2" diameters and trunnion-like brush bosses
      • I have parts to try an idea for making a low cost high speed spindle of similar laser cut construction, but haven't actually tried it yet

    Say what‽

    "apparently unprecedented"

    I’ve been logging work on this project on for about a year and posted an Instructable about five months ago. Both sites have drawn 10k± views and so far no one has mentioned a similar precedent. Please comment if you know of one, or eleven.

    "assembly from ready parts vs. fabrication"

    Certainly many people have made very low cost CNC machines from scavenged parts, scrap materials, an equipped shop, and a fair idea of what they were aiming for.


    Read more »

  • PROXXON vexxation; generic adaptation

    Paul McClay10/17/2021 at 02:19 0 comments

    Over the last several weeks I've burned a lot of project time chasing runout, including a vexing flirtation with PROXXON, because this "Minamil" design manages linear motion more precisely than a lot of rotary tools can manage the sharp end of a cutting bit. Actually, I was burning time chasing runout before that, which was why I layed down the (virtual) cash for a PROXXON "Professional Rotary Tool IBS/E" and its promise of <1mil runout (at the collet).

    Runout being the "wobble" of a thing that's supposed to be round and revolve around its own center but in reality never is or does. The bits I'm using are close enough to round but the rotary tools I've used for milling spindles (generously interpreted) to spin the bits do not reliably hold bits close enough to the spinning axis of the spindle. That means the pointy ends of cutting bits that are supposed to spin on a point instead get swung around in little circles that are often not little enough.

    The really horrible part about lesser (i.e. le$$er) tools is that because they hold bits so randomly, they're occasionally dreadfully near to perfect, which terribly provokes belief that it can happen again, and maybe will on the next try… tempus ardet.

    So... the PROXXON. Runout inside 1mil promised at the collet, with steel 3-jaw collets providing the implied reduction of randomness. And other features for more plausible use as a light duty machine spindle. And quiet.

    The first one was great...


    As for runout, it repeatably held short v-bits close enough to true to make ~5mil cuts through copper plate with a 4mil (0.1mm) v-bit. Close enough for this application. Repeatably - as in chuck up a bit and go. Delightfully tweakless. And quiet.

    For example:

    Dremel v. PROXXON isolation line width
    click for log entry


    ...axial free play, about 0.2mm (8mil), exceeded lesser tools on hand. Not a big obstacle to milling circuit boards where the two cut depths that matter are just through the copper plate and entirely through the board. But not so great for shaping more subtle surfaces. Especially in comparison with otherwise inferior tools noisily swinging gobs of runout but holding close enough to zero axial play.

    Proxxon (enough shoutcase) doesn't actually promise any particular limit to axial free play for that tool. I asked. The only data point I found for comparison was one report of 0.1mm axial play for an IB/E after long use. So I decided to call it a defect and exchange.

    The next unit did indeed have less axial play, tho still more than lesser tools. But fail for runout -- about 3mil at the collet and several times more at the end of a short bit. Maybe better than some options, but 3x worse than advertised and the larger swing at the end of even a short bit suggested poor parallelism too. Second exchange.

    Third unit no joy neither. Axial play similar to second, which I maybe could live with in exchange for repeatable low runout, and less runout at the collet. But apparently similarly poor parallelism so again swinging big circles (10s of mils/10nths of mm) at the tip of a short bit.

    I had both #2 & #3 in hand for a while before shipping #2. (yay Amazon for cross-shipping exchanges) So I burned yet more time swapping & testing collets (several sizes included) & collet nuts between units. Runout seemed to follow collets rather than spindles, so I wondered if it was simply a bad run of collets. Proxxon agreed to send another collet, maybe from a different production run assuming it came from somewhere other than an Amazon warehouse, which seemed about the same as what came with the third unit.

    What gives?

    Proxxon has generally positive reputation. The first unit was tantalizingly near superb. Are they really this inconsistent? Will they get back to being better after a crap batch flushes through some warehouse? Are they entirely random and I should just keep exchanging? Has the brand simply been gutted and I should forget it? Am I looking at good units with bogus expectations?...

    Read more »

  • Minimally milling fine-pitch circuit board features

    Paul McClay10/12/2021 at 03:19 0 comments

    update: text style edits; comment re tool change/through-holes

    Responding to a comment on an earlier log, I wrote:

    I haven't written much yet about the blue Wen dremeloid used for most of this project so far to keep it cheap. The main trick [...blah blah...] acceptable runout. I don't think it was ever going to cut the SSOP footprint tho.

    Since then, I got some renewed motivation to try again...

    ...and made it happen. A perfect spindle would be nice, but in the spirit of pushing down the hardware requirement and frustration with the "better" tool, I went back for another round of tweaking away runout. I'm finding more tricks but don't really have that down to a method for a good how-to yet.

    The overall process for milling this example went something like:

    • use stable magnification
      • I've used a phone (camera) held in a phone claw on a tripod...
      • ...until gaining unfair advantage coincident with this iteration
    • tweak the v-bit to small runout (...see? just type the words and its done...)
    • adjust spindle speed to minimize resonant vibration at v-bit tip
      • I've read of doing this by sound and/or feel, but I had to actually watch the tip of the bit under magnification while adjusting speed
      • assuming speed control either onboard or external to the rotary tool 
    • peck a hole just through the copperplate
      • 1 oz. copper is nominally 0.035 mm or 1.4 mil thick
    • measure the breadth of the hole - I did it like this:
      • using magnification...
      • turn the v-bit sideways to view the tip edgewise as a fine point
      • set jog step to 0.025 mm = one whole step (20 steps/rev, 2 revs/mm)
      • position the bit over the hole just a step above the surface
      • jog a few steps to one side
      • step the bit point back toward the hole and stop at the first edge
      • continuing in the same direction, count steps across to the other edge
      • hole diameter = steps * 0.025 mm
        • ideally 4 steps for a narrow-angle 0.1 mm v-vbit
        • more likely 5 for a "perfect" setup, and 6 or 7 will be pretty good
    • use measured peck diameter as tool diameter for generating gcode

    Tool changes for drilling and milling won't be so carefully precise. Notice in the video below:

    • pilot drilled the through-holes with the v-bit in an attempt to help the center the less carefully chucked drill bit
    • maybe might have helped with precision of through-hole locations maybe
    • didn't prevent the swinging drill from carving an oversize hole in the top surface
    • drill did self-center and exit holes are not oversize
    • narrowing of through-holes with depth somewhat visible in vid thumbnail below

    Sub-2-minute video for eye-candy:

View all 43 project logs

Enjoy this project?



Clayton wrote 07/24/2022 at 12:45 point

Ordered the parts from Aliexpress yesterday to give this a try.  Is there a Discord or other community where people talk about this project, or is the Hackaday page the best place?

  Are you sure? yes | no

Elliott Dyson wrote 07/17/2022 at 06:48 point

Amazing work! Do you think you'll ever work on a larger version for a little more money (but still a lot less than the machines you can just buy)? I'll be Keeping a close eye on this project!

  Are you sure? yes | no

AJonkhart wrote 06/02/2021 at 19:37 point

This is going to be my first aluminium project for my WorkBee cnc

  Are you sure? yes | no

Paul McClay wrote 07/27/2021 at 04:34 point

(two months later...)

I'm sorry I didn't see this before - I didn't get any notification when you posted.

Did you mean that you were thinking of cutting the flat parts from aluminium plate? If so, I would very much like to hear more about that!  

  Are you sure? yes | no

Ahron Wayne wrote 01/08/2021 at 03:35 point

The closeup revealed when you zoomed out on the gear and it was on your fingernail was fantastic and the video kept getting more and more impressive as it went on. Not just impressive but just full of eye candy and I'm absolutely stunned that the CNC wasn't even the focus of it but obviously wow, look at what it can do! Seriously, you should win all sorts of prizes for this if you haven't yet.

  Are you sure? yes | no

Paul McClay wrote 01/08/2021 at 22:47 point

Glad you liked the vid! - that was a bump for #the Metaproject of describing projects.  This project hasn't drawn much attention beyond spillover from #CDCNC 's 15 min of fame -- but I'm kinda sandbagging until simplification and documentation meet in the middle somewhere. I guess I'm trying to take most of the "hack" out of building your own. :)

  Are you sure? yes | no

Dan Maloney wrote 09/14/2020 at 20:18 point

I like the idea that you're building on the ideas tested in the CDCNC. Looking forward to seeing what you come up with.

  Are you sure? yes | no

Paul McClay wrote 09/15/2020 at 04:43 point

Hi Dan - thanks for skulls! Sorry to keep you waiting for results. While slow shipping and other life block progress for a while, I think early results give [me] good cause for confidence.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates