Getting really irritated with the quality of wheel/bearing assemblies and how they behave over time.
I’ve had bearings fail. That’s ridiculous. There just isn’t much load. My challenge in sourcing better bearings has not been particularly fruitful. It’s difficult to choose a better bearing without totally breaking the bank. A $10 bearing would be acceptable, a $30 bearing isn’t. A $4 bearing just doesn’t seem to hold up. I’m surprised, because the loads just aren’t that high. I’ve had about 6 bearings so far out of a set of 32 (16 wheels, 2 per wheel) start clicking or get rough. The 625 bearing size isn’t uncommon, but there seems to be a big jump in price with nothing in between ($4->$30+ per bearing)
I’ve worn out (apparently) wheels. I am actually using some heavy duty wheels and bearings I sourced from Luke a while back, and while an improvement over the stock ones at the time, I’m still having trouble. I’ve got several pairs (top and bottom) that can no longer be tensioned properly - I’m assuming that’s what “worn out wheel” does. I don’t think this is just flat-spotting, as rotating the wheel doesn’t allow any better adjustment at different parts of the surface - the whole thing is undersized at that point.
I believe I’m installing and tensioning properly (tighten the adjuster until the wheel just barely can’t be spun by hand), but within 6 months or so and a few adjustments along the way I’m back where I started, and can’t adjust the wheel far enough to be tight enough against the rail. As I’ve reached this point, I’ll replace a top and bottom wheel assemblies in the same position. I’m NOT putting lots of hours onto the machine (a little less than 15 cutting hours in the past 6 months according to the Hobbs)
Anyone else having issues? Am I doing something wrong?
Might I ask whose bearings you are using? 625 written on a metal thing is unfortunately meaningless now since some manufacturers flooded the market with low cost clones of parts.
Have you tried sourcing NSK or SKF from a reputable dealer who is actually selling you the real thing? I have found these for around £5 each from resellers I think are selling the real thing over here. If I don’t see the laser engraving of a bearing manufacturer I know, preferably Japanese, and the right printing on the box I assume it is made of cheese.
Are these a ‘regular’ 625 as supplied by Carbide? (I can’t see any manufacturer owning up to them on the bearings in my V Wheels here). These are just regular deep groove which don’t have any real axial load strength and are known to wear very quickly under axial load. Unfortunately the SKF angular contact equivalent is not cheap.
There seem to be some sort-of-angular-contact bearings out there, some described as magneto bearings at much better prices. These folks have both SKF and NSK branded magneto bearings at prices that won’t make you cry.
Next question; Which bearings, where on the machine? Are they the bearings above the rail or the tensioning bearings below the rail on the eccentric or both? As we know that axial load on a deep-groove bearing is a way to punish it, I’d be very interested if you’ve got a particularly high failure rate on, say, the lower bearings on the Z axis mount to the X rail.
And… Have you checked the width of the precision shim between the bearings and the fit in the V-Wheel? If the bearings are being compressed by the mounting bolt they will wear out in no time at all.
I believe the intent of these plastic V wheels is to be softer than the aluminium rail and to wear in place of wear on the relatively soft aluminium which would be much harder to replace. There are also plenty of application notes warning about accellerated wear in dusty environments where the rolling surface is allowed to become dirty.
Have you been able to measure the worn out units and compare them with the dimensions of the new in-spec wheels to see where they are wearing and whether it’s laterally or radially symmetrical?
I have the opposite problem at present, my HDZ backplate lacks a little in the precision department and consequently I cannot release the lower left eccentric far enough to un-jam the V wheels on that side without attacking one of the V wheels with some wet and dry on a shaped block to gain sufficient clearance in the ‘loosest’ position of the eccentric, they’re quite malleable.
We should, perhaps, consider these to just be a wear-out consumable like brake pads? However, not at 15 hours, that just sounds bad, I have many more hours than that on mine, I hardly ever clean the rails, the wheels are covered in lumps of dust and crud, they get brass brushed when I remember and they’re all still sort of OK.
I’ve been looking for a verifiably high spec NSK/SKF/FAG bearing, and that’s where I keep ending up with $30/ea bearings (shielded deep roller). Although an angular contact bearing would probably handle the load a lot better (by spec anyway), the lack of shields is problematic in this sort of environment.
My primary problem is the X to Y plates, somewhat less the bottom two on an HDZ. When I take this particular worn set off, I’ll measure things up and see what there is to be seen. I’ll check the shims at the same time.
The pretty standard “cheapo” 625 bearings i can find all have about the same ratings as more expensive ones - about 150lb static load, 375-400lb dynamic. Both of which seem well in excess of what kinds of loads I would expect. Now, that’s claimed, and they’re all pretty no-name, so yes, cheese would be an appropriate description.
I’ve looked at using a polycarbonate wheel, still a lot softer than aluminum. I do certainly expect wheels to be consumable, just like belts, I’m just having a little trouble wrapping my head around why they seem to be wearing so much more rapidly than I would expect. I’d be fine with 50-75 hour life.
As I look deeper into the SKF data sheets, they make reference to insufficient radial load causing “smearing” and ball damage. I did change the way I tensioned the wheels from “really firm” to “barely tighter than I can make a wheel slip” - I wonder if I’ve wandered into a range where there isn’t enough load instead of too much load causing bearing failure?
A ‘good’ 625 from SKF or NSK appears to be in the £5 range from reputable suppliers (so $5)
Now that’s interesting, those wheels and bearings should be under a lot less load. There are 8 of them not 4, they’re spread out over a longer lever baseline and their axial loads are shared rather than cantilevered across pairs.
Is it possible that your X axis rail is slightly too short or too long and there’s a permanent preload as you tension the V wheels down onto the Y rails?
Ok, had a lot of things apart today, took some measurements. Surprised at some of the variation.
top front. bot front. Top rear. bot rear
vwheels. 20.02-20.1 19.99-20.02. 19.99-20.13. 20.02-20.18. (mm)
shim edge. .986mm. .998mm. .988mm. .989mm
shim center. .973mm. 1.009mm. .970mm. .971mm
vwheels were measured for high and low spots by rotating and taking multiple measurements at the deepest part of the v. Tried the measure over a couple pins trick to measure further up the sides, but just didn’t have enough hands, and ⅛" pins are small enough to make things pretty awkward while holding the wheel
All the plate-side washers were accounted for, and installed with the sharp punched side toward the plate, the round side toward the wheel.
The “precision shims” are likely a punched product or made on a terrible lathe. Tooling marks look like a lathe. The center area is considerably thinner than the edge, which likely means there’s a small burr on the edge or the whole shim is cupped. The anvils on my micrometer are smaller than the whole shim. I’m not sure what real impact the thickness of this shim has on the overall assembly.
I replaced all 4 wheels on the right side x-y plate. Coming off the machine, the top ones felt a little gritty, not a lot, and the bottom ones had very smooth bearings. After taking the wheel assemblies apart, all the bearings, including the top wheel bearings felt just fine. I can’t really explain that. I did not measure the wheel bores for roundness, but as soft as the wheels actually are, I’d be surprised if they had enough gumption to distort the bearing and cause that. I had expected the rear wheels to be in worse shape than the fronts, as these were the ones that I was having adjustment issues with - but the data says they should have been all about the same.
The x rail could possibly be too short, but I don’t have a good way to tell by how much that I’ve thought of yet…will need to noodle on how to figure out if it is, and by how much.
I have carefully avoided measuring any of the shims, I don’t want to worry about those too
So the V wheels are quite out of round in the central slot with a size range of 0.2mm That’s pretty consistent for a deformable plastic, but they’re so worn now you can’t get them to tension up any more?
As for the X axis alignment, when I was taking some measurements with the dial gauge, I accidentally noticed that the Y rails were actually rotating slightly as the X axis traversed in Y, this was due to the V wheels pulling the axis into their alignment. If you have a dial gauge and mag base perhaps clamp the base to one of the Y plates, put the dial gauge against the vertical face of the Y rail it’s on and then loosen / tighten the eccentric nuts, if the distance or angle between the Y rail and Y plate changes then the alignment of the Y rails and Y plates is out somewhere, does that make any sense?
I’m getting 19.56 to 19.60 on one and 19.58 to 19.62 on the other measuring the bottom of the V, however, the caliper (Mitutoyo) jaws are not sharp enough to go all the way to the bottom of the V.
For the OD I am getting 23.54 to 23.57 on the smaller and 23.58 to 23.61 on the larger.
Not sure how to get a cleaner measurement to the bottom of the V, or whether that is desirable at all, the notch at the bottom looks like it’s deliberate and not part of the running surface.
I also had a quick go at measuring the distance between the wheel bolts on the Y plate, I measured from the edge of the welded nut to the edge of the eccentric as that was a point to point distance I could get without trying to find centre on the holes, it’s not that accurate but might be a worthwhile first ‘ball park’ distance check. You can see I’m about halfway on the eccentrics to get them to “can’t quite turn them with a finger” tightness.
