I can only speak for my own (extremely limited) experience with the machine, and I'd really like to hear from the Openbuilds team and/or anyone with more than my tenuous fingernail grasp on the wet & slippery edges of engineering.

Personally I only ordered the high-torque stepper for my Z axis because of the incautiously heavy spindle I was planning to mount on it - I didn't (and still don't) know if it was strictly necessary, but I had to buy another motor anyhow, and figured "better safe", and all that. For X and Y I used my old leftover KL23H256-21-8B steppers, which have only slightly (and certainly not significantly) higher torque than the Openbuilds standard models.

So far I haven't done anything very heavy (I've only tried 1/4" endmills on MDF and 1/8" on aluminum), but even when pushing the feeds higher than the spindle wanted to cut, I didn't see any indication that I was short on power - it felt like structural flex would become an issue before the torque ran out.

It's also important to keep in mind that high torque steppers aren't just "the same motor with more torque" - in addition to requiring an upgrade to both drivers & power supply, they also tend to suffer in the speed department compared to the standard models (and/or the speed where they stop providing useful torque at a realistic voltage is significantly lower - it's been a few years since my reading on stepper motor theory, the details are a little hazy, and I've never seen a torque rating curve for any of the Openbuilds models). I wasn't worried about speed on the Z axis (and already had beefier electronics), but would've been afraid of killing my rapids on the X & Y.

Don't take any of this to mean "don't go for high-torque steppers" - there are definitely situations where they're appropriate. Just be careful to make sure your planned application is one of those. It also wouldn't hurt to wait on advice from someone who actually knows what they're talking about and/or doesn't have bats in their belfry.

I've never used the Dewalt or Makita trim routers, so I can't comment on those - my previous machine used a string of Harbor Freight trim routers that probably had about 1/3 the power (at $25 a pop they're an attractive looking option for wood & plastic on a DIY machine, but the runout means tool changes can take 10-20 minutes of fiddling with an indicator if you're trying to do any remotely detailed work with small cutters - recommended only for the most hopelessly impoverished of masochists) - but a lot of people seem to find them (the Dewalt/Makita, not the HF) perfectly acceptable, even for cutting aluminum.

I've got a similar power problem to you over here - a full breaker box means running a 220V line to my work area isn't a cheap proposition, so I'm pretty much stuck on 110V too, which limited my spindle options and meant that I was out of luck on anything with an ER20 collet (to fit 1/2" shank router bits - something I really wanted for larger surfacing bits) without going for an absolutely enormous 2.2kW spindle (which, aside from weighing far too much, would've required a spare 20A circuit - something I, naturally, also don't have in the shop). After much searching and a number of awkward google-translated conversations, though, I did finally manage to find someone who could give me an ER16 (which can take collets up to .4375") on a 1.5kW 110V water cooled spindle ($307.84 shipped, with VFD, pump, tubing, and collets from Lisa Zhu at the awkwardly capitalized "HLTNC zhejiang bearing factory Store"; ordered 12/11, arrived 12/26) - the drawback being that it's one of the big chunky 80mm diameter models, not the more reasonable 65mm sort which tend to weigh about half as much.

To give some indication of just how big this thing is, this is it next to the old HF trim router (with a 1/4" endmill in it), on a length of 80x40 C-beam, and, finally, mounted on the Lead Machine itself:


You may notice that the machine looks a little small in that last shot - it's not just a clever use of perspective. I had to build mine narrower than spec to fit the space available, but the short gantry is also the only reason I felt comfortable getting such a heavy spindle to hang on it. I may still end up widening the machine a bit more (using half-length beams ended up even skinnier than planned because I can't math), but the longer that gantry beam gets, the more it's going to flex with that weight in the middle. If you're planning to build a full sized machine, keep that in mind - you'll likely either want to use a smaller spindle or start making plans to rigidify your gantry.

If you're content with an ER11 collet (which tops out around .3125" - still big enough for 1/4" and 7mm tools) you can find smaller 800W 65mm versions all over Ebay. Technically it's only half as powerful, but my suspicion is that the weakness of the machine - of any aluminum extrusion-based machine - will keep you from ever tapping nearly the full potential of the larger model (using the G-Wizard speed & feed calculator to derate spindle power based on machine weight makes it look like even an 800W spindle will go largely untapped).

The spindle, though, is fantastic, and I'd be in love just on volume alone - or the complete lack thereof. Where the trim routers scream, this purrs so quietly it gets drowned out by the cooling fans on the PSU, the VFD, the PC, and any other capitalized letters that happen to be in the area, and where trim routers fill the air with clouds of dust, this leaves polite little piles around the workpiece (this isn't always an advantage, as chip buildup can dull a tool and/or start a fire, but generally spindle fans make a mess and still fail to clear chips & sawdust from the deep slots where they cause the most trouble - as can be witnessed by the MDF fire my spiteful previous "spindle" tried to start when I made it cut a mount for the new one).

The speed control from the VFD is incredible - I've used it as low as 5000rpm & could probably go below half that without any perceptible loss of torque, and as high as 24k, and it can be dialed in precisely with the pot and (more-or-less accurate) RPM display on the control panel, programmed as presets, or wired up to be controlled by G-code (I'm waiting on the latter until I reroute all my wiring & build a new box for the electronics). Being able to run at low speeds is great since it means I can match it to whatever feed rate the calculations come up with, the machine feels stable at, and I feel comfortable with, instead of struggling to get the machine racing fast enough to keep up with a high speed tool.

Power-wise, it breezes through anything I've pointed it at (including the edge of a steel washer, which the rest of the machine wasn't very happy about), although with the 1/8" endmills I've used so far, that's not especially demanding.

You can see how it did with a smaller piece a few posts back, but this spindle mount is the biggest thing I've given it to tackle so far:

It's a piece of .5" (12.7mm) thick 6061 aluminum scrap that I was cutting (or was trying to cut, before the XP box running the machine decided it was time to crap out on me) last night, using a 1/8" 2-flute carbide bit running 18000rpm at 40ipm (16mm/min) taking .125" (3.175mm) DOC passes - no coolant this time either, just occasional compressed air to keep the slot clear of chips. [note that it was using a trochoidal ("adaptive" in Fusion) toolpath on that curve and straight cuts on the three outside edges, with only about .045" engagement - trying to plow straight into a standard slotting cut that deep & that fast would probably have made the machine extremely unhappy]. I obviously didn't bother surfacing it first.

I'm not sure how much of the chattery finish on the outside has to do with tool deflection on the skinny endmill, how much has to do with the temporary MDF spindle mounts or a general lack of rigidity in the machine, and how much has to do with the fact that I later discovered all eight of the screws in the gantry's endcaps were rattling around loose (that conversation about blue threadlocker? yeah. probably a good idea), but I feel pretty safe in saying that it wasn't caused by any inadequacy of the spindle or a lack of stepper torque. It's also missing a finishing pass, since that's around the point the PC bluescreened on me - naturally losing all sense of position in the process. [insert 13 pages of profanity here]

Also, did you notice all that talk about spindle mounts (of both MDF and aluminum flavors)? The other big thing you need to plan for when using a non-stock spindle is how you're going to attach it to the machine. If you order a mount with your spindle, that may be as simple as a plate with one set of holes to mount to the Z-axis and another set (tapped) to mount the mount (be aware, the mounts for 80mm spindles are much wider than the Z-axis C-beam and will limit your X travel - I haven't looked at the dimensions of the 65mm ones). If you get a 65mm spindle, you can also shim it into the stock Openbuilds mount (either with a 3D printed shim, or lots of sliced beer cans - you should have plenty of empties if you follow my assembly notes for the machine) until you decide whether you want to make something better.

If you get the water cooled sort (which I absolutely recommend - otherwise ignore all my glowing comments on the silence, lack of blowing debris, and super low speeds, because none of those will apply to your air-cooled model), you'll also have to plan out your cooling loop. That's fancy talk for "ya need somewheres ta stick a five gallon bucket with a couple o' tubes & a power cord". Eventually I'm planning to put together a much smaller PC-style closed loop system using a radiator & fans with a much smaller reservoir, but the pump-in-a-bucket model seems to be industry standard for home systems.

So, as a reward for reading this far, here's the short version: Yes. Absolutely. Chinese water-cooled spindles are awesome and wonderful... but do as much research and planning as you can before picking one out, because they're definitely more work to get up and running than going all-stock with your system.

I went a long time without one, figuring it really wasn't all that important. I was an idiot.

If all you're doing are two dimensional cutouts, you can get by without one - just run the tool down until it starts digging into the spoilboard, hit zero, and it's probably close enough... but as soon as thickness becomes important, that stops working. You can get by for a little longer with tricks like sticking a piece of paper between the tool & spoilboard and adjusting until it stops moving when you pull, but it's slow, it's awkward, and you'll still never get much precision or repeatability out of it. Using a touch plate is far more repeatable, but (and this is really the important bit) it's a whole lot quicker and easier - especially if your software lets you set up a simple one-button zeroing script.

As far as which one, that's a much less important question - the difference between none and the cheap little $3 Chinese one is vastly wider than the difference between the $3 one and a $50 one. It's basically just two wires that connect to your controller, one with a flat chunk of conductive metal on the end, and the other with an clip on it. You clip the clip to your bit/endmill/cutter (or anything with a conductive path to the cutter, which is great with all-metal spindles, because you don't have to worry about forgetting to unclip before turning it on), then the script runs the Z down (usually using a G31 probe) until it reads continuity between the two wires, and sets that as your new zero (my script for Mach3 does a first quick pass to figure out where the plate is, then retracts very slightly and does a second slow pass to get a more accurate zero). The most important part about the hardware is that you reliably know the precise distance between the top of the plate and your workpiece - so if the top isn't flat and rigid or the top & bottom aren't parallel, it's not going to work very well (and the larger the surface area, the more you have to worry about this) - that's really the only reason not to just skimp out & wrap the foil from last night's baked potato around a hunk of cardboard.

The biggest problem I've found with the little $3 ones is that they don't weigh much, so you have to look first and make sure they're sitting flat & not being held at an angle by the wire. No, they're not high precision, but if you're using an MDF spoilboard under your work, the flex is going to introduce more uncertainly in Z height than an only-almost-parallel 1" surface that mis-sets your zero by a thou or so (and I haven't had one even that far of).

If you really want to get fancy, you can inlay a plate into the spoilboard by your machine's home position (maybe even surface it at the same time as the rest of the board) so it can zero all three axes at the same time - but it's good to have a mobile plate for those times when you want to zero on top of a piece of stock instead of on the spoilboard. In the other direction, if you mostly work with aluminum & want to get really simple, you can just use two alligator clips - one on the tool & one on your conveniently conductive stock.

The one feature I do like the look of on some of the more expensive designs is the corner finder... but once you've got a something that can machine aluminum, why buy something that's just a chunk of aluminum plate machined to uncertain tolerances?

Sorry about all the answers


-Bats
( what're you looking down here for? didn't I type enough up top? )

 

Thank you very much for all the information.
This is a spindle kit I was looking at, any input?
https://www.amazon.com/dp/B01N1EJEQ1/ref=cm_sw_r_sms_c_api_i_OKttCbMC96XHD

 

The main pieces in the kit looks fine (you'll probably want to replace the tubing if they send that orange crap, you'll need to get your own wiring, and at least one reviewer mentions the mount cracking - which unfortunately seems to be a rather common complaint for whatever garbage they cast those things out of. it also looks like the mount is half again the width of the Z-axis c-beam, so you'd still need to make an adapter of some sort to hang it, and drill your own holes in it), and it's the 65mm format so it shouldn't be as heavy as mine. I'd still recommend caution with regard to the weight, though, since the seller indicates it's about 7lbs. Of course, in another question the same seller says the spindle weighs twelve pounds and elsewhere flatly says it can't be controlled by Mach3 when there's no reason it shouldn't work, and at least one reviewer says they're doing exactly that (it's also my plan for the same VFD)... So you might want to hunt around and see if you can find a more reliable source of information, too.
If it really is 7lbs, that's close to twice the weight of a Dewalt (or so I claimed in this thread, where I dumped a lot of the spindle data I was able to dig up while making my decision). It's probably lighter than mine (at least going by the numbers in that same thread - I'll have to actually weigh it when I change mounts), but that's still a whole lotta spindle pulling down on the middle of a meter long rail. It might be perfectly fine, or at least fine for what whatever you happen to be doing, but with that much of an increase in weight over the stock version you're taking a gamble if you don't test the machine first. Setting up a dial indicator to measure deflection and putting a load equal to the spindle weight (pulling down with a fishing scale, using sandbags, water bottles, or a pile of cast lead Chinese baby toys) on the Z axis in the middle of the machine will tell you exactly how much it deflects, and you can decide whether that's acceptable before spending the money. Or I suppose you could probably just chance it, try it out, and send it back to Amazon if you don't like it (I would've had to ship mine back to Guangzhou, which was a little less convenient).

If anyone else happens to get their machine built & feels like doing a few tests (comparing deflection with the Dewalt, vs something half again as heavy, vs something twice as heavy) before putting it into service, I think a lot of people would be grateful to have the data (including future people, who might have exciting futuristic ways to express gratitude!). It sucks to have the machine together and then get stuck waiting for someone to ship a spindle - or, conversely, to take a chance and find that you have something too big to use, or smaller than you could've gotten.

(not that it really matters, but that power supply is also an odd choice of freebie for them to toss in. aside from the fact that the spindle doesn't even use it, it's a 12V 30A unit, while any CNC rig using that large a spindle would want a whole lot more voltage & could get by with a lot less current - I'm using a 36V 10A on mine. not that I'd turn up my nose at any free PSU - it just seems a bit random)


-Bats
( just think about it... someday I'd be a futuristic grateful person! I hope that doesn't mean I have to run around in a white jumpsuit like Buck Rodgers. I can't do white. )

 

Has anyone figured out dust collection for the Lead 1010?

 

Yep! I've got a whole room in my basement dedicated to collecting dust, and now that the Lead's running in there, I can collect it faster than ever!


-Bats
( if I stick to cutting MDF, I figure I should have this one full and be able to move on to a second room before the end of summer! )

 

I’ve been following mine with the shop vac going to eventually build something though. There are a bunch of plans out there for boots that fit the 611.

 

So, I've been using/building/tearing apart/fixing/breaking/rewiring/generally mucking around computers in one capacity or another for a good few years decades now, and I've seen/fixed/caused a wide, wide variety of strange and interesting problems. That's not to say I always know what's causing an issue or how to fix it, but it's not often that I run into something truly novel.

I think I mentioned a little while back about some trouble with the CNC machine bluescreening. While I have yet to actually solve the problem - or even figure out exactly why it's happening - I think I finally stumbled across the cause... and, as bugs go, it's a neat one.

After running (and attempting to run) a wide variety of system tests - complicated by the fact that's it's Windows XP, the constantly-requested CD suffered a fatal scratch some ten years ago, and I'm not sure I even have a functional optical drive left in the house - I was able to find nothing at all wrong with the machine. Or, well, least nothing other than the fact that it's a decade-old install of XP (possibly upgraded from 98 and/or 95 at some point the foggy fog of history) and suffering all accumulated cruft and bitrot that comes with extreme age... but nothing that looked likely to be responsible for this problem. I also hadn't been able to reproduce the issue in a few days, so I eventually gave up on it, fired up Mach3, and got ready to make some chips...

...and bluescreened right as I started to run the first program.

So....... powered down the spindle and back to troubleshooting. Looked at the hardware this time - pulled everything apart & put it back together, re-seated all the connections inside and out (none of which showed any sign of being loose), marveled at the pristine cleanliness of the interior of the PC and congratulated myself on my fantastic dustproofing and clever use of the pantyhose that I seem to have inexplicably accumulated (don't ask - I said it was inexplicable!), just before reproaching myself for getting all smug without actually solving anything. So I put it all back together and cautiously started jogging & then increasingly incautiously racing the machine around to see if it might be vibration-related, and...

...and nothing, actually. (I know, you were waiting for a spectacular failure to laugh at here. So was I, honestly - this was a pretty anticlimactic)

After another day or three without a solution, cause, or recurrence, I got brave and/or stupid again & decided to cut another limit switch plate, which also worked perfectly. In fact, it worked so well I thought I'd tweak the feeds & speeds a little, and see if I could push the machine a bit harder on the next one. So, with feeds calculated to match, I cranked the spindle to full power & got ready to let 'er rip...

...but thought I heard a clunking noise. It didn't seem to be coming from the spindle itself, but when I turned it off to hear better, it went away. Dialing the speed back up, the sound came back - rather incongruously seeming to originate in the vicinity of the hard drive... but before I could be certain, the machine bluescreened again.

So from there I [ page removed, because no one wants to read the details of your last ten hours of trial and error, Bats. Ever heard of getting to the point? --Ed ] and finally realized that with the VFD running at anywhere between 385-400Hz communication between the PC and hard drive seemed to stall out completely - the drive would keep seeking (and without that sound I never would've gotten this far), but watching the graph for the disk queue in the performance monitor, it would climb but never empty until I turned the VFD down below 380Hz. Presumably that's not the only problem being caused, as stalled I/O by itself doesn't seem likely to cause that sort of BSOD (0xF4, critical object termination), but aside from the crash itself, it's the only visible symptom in the system.

For the short term it's easy enough to program the VFD to just avoid the problem band (and it looks like I may be able to do it without even losing any speed, with some tweaking of the max RPM setting) and in the long term I was already planning to stick the electronics in a shielded case, but as I said up top, I still haven't worked out exactly why it's happening yet - and that bit bugs me. It's obviously being caused by some sort of RF/EM noise from the VFD, but the 3-phase VFD-to-spindle line is shielded & grounded, and the VFD (and wiring to/from it) is a couple feet from the PC and plugged into an entirely separate circuit - if it's radiating that much noise, my next project will have to be cutting some "NO PACEMAKERS" signs to hang nearby. It could also explain why I never get any phone calls while I'm in the shop. Of course, it's somewhat less likely to explain everywhere else that I never get any phone calls.


Moral of the story: RFI stands for "Rarely F....un or Intuitive"


-Bats
( this was going to be a short post, but I've been feeling guilty about my last short post ruining Gray's tea break. it has since been lengthened as a of courtesy to tea-drinkers everywhere )

 

I generally end up doing the same - although having a spindle without a fan blowing crap everywhere means that now I can usually get away with just letting said crap pile up (with while chasing it out of slots with a little compressed air) and suck it all up afterwards. My shopvac appreciates the change, as it already bears the scars of far too many altercations with spinning tools.

That said (and despite my earlier flippancy), I would like some sort of less-manual chip control system - I just haven't quite figured out what form it should take. My problem with most dust shoes is that even the clear acrylic sort make it difficult or impossible to see what the tool's actually doing, which, being a chronic fidgety knob-tweaker, leaves me in a constant state of paranoid panic, having to wait a whole inch or three to see what sort of damage the tool has done. Not that having a clear view saved me from snapping off two end mills in a row at exactly the same point on the stock yesterday. In two entirely different toolpaths, even.

Right now I'm leaning towards getting a cheap mister (running dry, for wood/MDF) to keeps chips clear of the tool, and continuing to clean up afterwards with the shopvac, but, while that's great for small aluminum pieces, it obviously doesn't scale so well to large pieces or long runtimes.

I've also seen people rig up lowlight cameras inside their dustshoes - giving a much clearer view of what's happening at the bit - but I suspect the idea of a live video feed may meet some resistance from the PC's antiquated onboard video.


-Bats
( sorry, only one leisurely tea break today. now get back to work, you slackers! )

 

Just to chime in here, I've seen something on MSC's website called a "chilly bits air chiller", that could be used for cooling aluminum during milling as well as used to blow wood chips clear of wood routing.
It's on Amazon and MSC, and only uses dry air pressure to operate.
We've got one at work, but it's never used because all our machines use flood coolant or misters (cursed mister units..)

 

Hrm... maybe I'll try pairing some headphones to the VFD.

Not entirely inapropos - I spent a while waving a between-stations AM radio around the system today. There's noise all over the place (there's noise on most of the stations, too... but I'm not quite ready to blame talk radio and religious music for my technical difficulties) - even the idling steppers generate their fair share, once you get close - but the VFD is broadcasting so loudly you'd think it was run by Clearchannel. At certain speeds I could pick it up roaring like a revving engine (a proper engine, too - not one of those whimpery little hybrid things) from several feet away - completely unlike the near silence radiating from the unit on my lathe (which, of course, is a quarter the size, a quarter the power, and probably less than half as Chinese). The noise tended to peak at lower speeds, though - not surprisingly, it tracked pretty closely with current draw - and by the time it got around half speed it'd calmed down to fairly steady background static, while the PC was running into problems at the very top of its range.

So, based on the evidence, I think it's safe to say that the problem stems from a lack of quality broadcasting on the AM spectrum. And that I have absolutely no idea how to apply any lessons I may or may not have learned today, beyond "stick everything in a big metal box".

Also, the PC completely neglected to misbehave throughout the testing. This somehow fails to reassure me.

-Bats
( it may be plotting something )

 

Interesting... I can't find much (any) information on how it actually works, though. Given the pricetag, I have to imagine there's more inside than just a tiny pressure tank with a tinier nozzle to cool the expanding air.

I probably won't be picking up the real thing any time soon, though - considering it costs more than my spindle & VFD combined, it'll probably have to wait until I have a substantially more expensive machine for it to blow on. The $10 Chinese misters are a little better aligned with my budget right now.


-Bats
( Do you think your boss would object to you cutting it in half so I could see how it's made? )

 

True enough... but if it's as simple as a nozzle on a high pressure line, you'd think it would've been discovered long ago by dozens or hundreds of different shops, and either widely adopted as a cheap cooling solution, or widely written off as ineffective - either way, not something that would be an easy sell at $3-500 a pop (in addition to the Abanaki version, MSC also sells several suspiciously similar "Vortek" branded models selling for as much as $483 - or $677 with a filter & magnetic base - in addition to a couple other Vortek designs for $519 and $1,875).

Seriously! The pressure looked had looked manageable, but I didn't even notice the throughput. I think my frankenpresseor might have been rated for a third of that (meaning it was probably built to deliver less than a quarter), and it already makes more noise than I'm happy sharing a room with. Or a floor with - it's right on the other side of the bedroom wall, and if I forget to kill the power it usually waits until about 3am to remind me. Of course, I also bought the frankenpressor for $25 - possibly due to the breeze coming through the rust holes in the tank, which no doubt resulted in still lower throughput.

Still, who wouldn't be willing to drop a couple grand for something as useful and versatile as a nice new compressor? Just think about it - not only can you run your $400 ChillyBits on your $1500 machine, at the very same time you can run a pneumatic impact wrench to drive in those $#!%*@& "self-tapping" screws, an angle grinder to cut down the gantry C-beam to reduce flex, some sprayers for the kids to repaint the house and/or cars and/or dog, and get an airbrush makeover from your live-in cosmetologist!

Hell, while you're at it, you can pick up one of these, and you won't even have to worry about sharing a room with the noise and or running a new 30A 220V line!

Or I suppose I could just go back to chasing the tool around with little puffs from a blow gun.

...which is a shame, because I'm really overdue for a makeover.


-Bats
( just make sure the cosmetologist doesn't pick up the wrong tool )

 

Two words: Vortex Tube


-Bats
( Just like Arthur Dent, You have: No tea. )

 

Annnnd.... having identified the name of the underlying technology, suddenly the prices come crashing down. It looks like unbranded vortex tubes are, unsurprisingly, coming out of China in a variety of shapes and sizes at a tenth of the price. One thing that seems consistent, though, is the airflow requirement (assuming, that is, that "0.42 m⊃3;/min" translates to something like "0.42m^3/min").

I also spotted one of the $500 Vortex units being offloaded at $35, which was awfully tempting, even knowing I didn't have the air to use it.


-Bats
( making up for post length - and quality - with sheer quantity )

 

I'm the proud new owner of a pile of parts that are on the way to becoming a LEAD machine. I'm a noob, so of course the first thing I did was screw up. I assembled all my Extreme V Wheels without putting the shim in between the two bearings. It isn't clear to me how to pop the bearings back out to fix it, or why the shim was necessary in the first place. Is that something I need to fix? If so, how do I pop the bearings out without breaking the wheels they are snapped into? Thanks.

 

Use a small flat screwdriver to pop the bearings out of the wheels. The shims are needed between bearings so that the wheels roll smoothly without binding.

 

Sorry, but pop them out how? There appears to be zero give between the edge of the bearing and the wheel. Do you mean wedge the screwdriver between the wheel and bearing and pry or pry from the center or push through the center or what, exactly? After a lot of cussing and manhandling, I finaly got one out by prying from the center, but not without damaging the bearing and putting a small gouge in the wheel.

 

Put the screwdriver between the bearings in the hole where bolts go through.

 

I haven't had to take any of mine back apart yet (possibly the only piece of the machine that's true for), but I believe the idea is to push the bearings out from the center, rather than prying them from the outside.

Something like this:


[ edit: it might also make things easier if you can set it on top of something that'll support the wheel while leaving room for the bearing to pop out. finding a hollow cylinder of exactly the right size is probably too much to hope for, but setting it on top of the jaws of a partially-open vice should work, or even a couple blocks of wood set the right distance apart ]

-Bats
(assuming you're building the limited edition transparent easter-colored wheels, and using a slurpee straw instead of a screwdriver)

 

Thanks, Bats!

(To any future dummies that google "how to get bearings out of wheels", I basically did what Bats showed above, but braced the bearing over the end of an adjustable wrench such that the lower bearing had a place to pop out while the wheel was supported. I then had to tap the screwdriver with a rubber mallet while the tip pressed against the lower inner bearing ring until it came out enough that I could pop it out with my hand. It was really annoying to do twenty times. I highly suggest not doing this as your first move in building a new machine. Or at all really...

 

Tramming a LEAD CNC spindle. Has anyone cross this bridge yet? If so what worked, didn't work to dial in alignment.

FYI
I have the Kress 800 and the spindle mount designed by Martin Barfoed.

43mm Spindle Mount for 80mm Z-axis

 

I'm sort of in the process of tramming mine right now (I've got a two-piece top/bottom mount arrangement on my 80mm spindle, which makes things a little more awkward), but got sidetracked when I made the mistake of putting an indicator on other parts of the assembly first. It is important to establish a baseline to measure from, and to make sure the spindle is not only square to the table, but also square to the axis... but it showed me things I didn't really want to see. Between the top & bottom of travel, it looked like my Z axis was off ~30 thou from the gantry's upright on one side and ~6 thou (in the other direction) on the opposite side, which suggests the uprights aren't quite parallel to each other (or precisely "upright") either. That shouldn't matter too much, but I'm also seeing about .001" between the Z C-beam and the edge of the gantry plate, and .01" between the Z C-beam and gantry C-beam, which worries me a bit more.

I think some of that can be accounted for by flex in the frame - leaning on the edge of the machine can make the needle do more than just wiggle - so I think my first step is going to be throwing down a slightly flatter & more rigid surface underneath, in hopes of stabilizing the measurements and maybe making a few of the scary numbers slightly less scary (of course, if I went back and used a dial indicator with a .1" dial instead of a dial test indicator with a .8mm dial, it would probably also make the numbers look less scary). In the long run I'm planning on a torsion box - probably with a new table to go with it - but I need to get the Lead to a passably functional level and then assemble the mill that's in pieces all over the shop before I have room to comfortably work on something that big, so I'm probably looking at a few sheets of ply & MDF.

Once things are a little more stable, I think the rough plan is to surface the spoilboard, unscrew the arm of the dial test indicator stand from the magnetic base & stick it in a collet, spin it around on the board, and then swear profusely at the spindle mount until the measurements (on the almost-certain-to-exist peaks in the surface) come out the same all around. If cursing at the mounts doesn't work, I'll probably see if tweaking the screws & shimming will work. After that's as close as the uneven surface allows for, resurface the spoilboard (which this time will hopefully be something resembling flat, instead of just level), stick the indicator arm back in the collet, and then swear & shim out any remaining unevenness. Replacing the indicator+arm with a length of rod (or even a coathanger - whatever fits in your spindle and is rigid enough not to flex under its own weight) with a 90° bend at each end & checking visually will also do the job (instead of watching the dial, adjust until it just barely brushes or just barely clears the spoilboard all the way around), with the advantage of increased precision from a potentially much larger diameter, but the disadvantage of feeling less like a reel musheenist™ using fancy metrology tools.



-Bats
( Fun fact: Metrology is the study of metroids, a field developed by researcher Samus Aran. Don't poke Mother Brain with an indicator, though - she hates that. )