• Scam Alert. Members are reminded to NOT send money to buy anything. Don't buy things remote and have it shipped - go get it yourself, pay in person, and take your equipment with you. Scammers have burned people on this forum. Urgency, secrecy, excuses, selling for friend, newish members, FUD, are RED FLAGS. A video conference call is not adequate assurance. Face to face interactions are required. Please report suspicions to the forum admins. Stay Safe - anyone can get scammed.

Cheap angular contact bearings

DPittman

Ultra Member
Anyone have experience with cheap angular contact bearings?
I sort of gathering bits to someday make another tool post grinder and am thinking it would be based on Clough42's youtube build he did. Although he didn't use angular contact bearings.
I have been browsing these bearings on line and am astonished at the price range, from cheap to OMG. However there seems to be choices in between also and I'm wondering if there is any chance that cheap ones are any earthly good?

I see Aliexpress has NSK bearings labeled as made in Japan in the mid price range but I'm going to guess those would be similar to my "Mitutoyo made in Japan" dial test indicator that I ordered from them.
 
If you believe the skate bearing vendors, what the average skate boarder needs is ABEC 9 bearings and they'll sell them to you for $3.99 each. Point is, there's not a roll-your-eye's smilie big enough to describe the BS regarding some of this bearing stuff.

Here's a few things I would say are current truths - and its changed a lot over the years, and some ideas what you might use for this application. A bit verbose, but lots to say :)

- The world's low cost providers, and their propensity to arbitrarily claim whatever class or certs they think the customer wants to hear (something that is truism of that supply chain across product lines, read "Poorly made in China" for example) means unless you are buying from someone who's a part of the big manufactures supply chain, you have no idea what you getting. Buy from industrial houses that sell SKF, NSK etc. You buy from Aliexpress, Amazon or Ebay, you have no idea what you are getting....but they might be great for skateboarding! :)

- The cost varies with size obviously, but also with class. Regular SFK AC's are P5 (an ISO class) and SKF Super Precision AC's are P4. Magnitudes of price difference between the two. P4 are roughly the equivalent of ABEC 7, but as bearing sizes are 99% metric as are the makers, giving the ISO class makes sense. With ISO, a smaller number is a higher class (overall greater precision, less runout) and with ABEC a higher number is a better class

- Bearings are better than ever and often better than their stated class. Super accurate modern CNC's make things easier to keep to tighter tolerances. I've spoken with bearing mfg engineers from two different mfg's who state that there is no difference between the manufacturing of a say a P5 bearing and a p4 bearing, the price difference is from the additional time to quantify and certify that it meets P4 standards. Maybe there is the odd reject from the P5 pile, so there is value in buying the P4's, but in lots of cases the P5's would be just as good.

- The final jump in price comes when you get matched sets of P4's I've bought a lot these over the years, this is where you spend $800 for pair sort of thing, and it hurts. You have to understand a little bit about how AC's work to appreciate it. There has to a be a preload with AC's and for most machines, that is determined by either the difference in thickness between the inner and outer race (in the case of matched) or the thickness of the spacers (you're grind them to create preload). You can, and I've done so, grind the spacers to create the preload you want and then P4 universals become the same as P4 matched, but the situation doesn't always allow for it - e.g. replacing existing back to back arrangements with room for spacers. Even with this, non-matched are now called universal and come with a preset preload...another example of overall manufacturing quality improving imo. The advantages of matched P4's imo is literally, getting very small.

- All the tool post grinders I've had apart use a spring to create preload - there is hardly axial force so it works. In such a case, there is zero need for spacers or matched pairs. You do not need the 800 bearings, nor would give any advantage.

- Several TPG use magneto bearings, not a particularly high end solution, but it just says they can still work when things are less than perfect

- Deep groove ball bearings (even the good quality ones are cheap and P5's are likely as good as P4's) with either a spring or belleville wash arrangement giving a bit of preload is an option. This is how a unimat spindle for example is set up. Or how a Boley 82 or 83 is done (one of the highest end watchmakers....although I confess I converted mine to AC's :) ). There is so little axial force in grinding it might be good way to go.

Hope that helps. Unless you are into matched super precision sets, good quality bearings from reputable makers are reasonably enough priced that imo its make no sense to use mystery bearings. After how much time goes into a project it seems (to me anyway) false economy to cheap out on the the mission critical part of a spindle. I'd go SKF P5's AC's, (which are not matched and I believe have a universal preload) and set the preload up with a spring. Probably <$80 depending size, but you'd putting top quality bearings into it, worthy of the considerable amount of time you'll be putting into the project
 
Last edited:
Good info thanks. I'm still trying to get my head around all this bearing stuff and I know surprisingly little. Figuring out what bearing is suitable for my application is one challenge but then I also have to figure out how to use/install the bearing correctly also.
 
Depending on the size you need, you may look for an insert style automotive wheel bearing. Double row angular contact. On some front wheel drive cars, the rear wheel bearing inserts are 20ish mm id, 35-40mm od. Automotive timing belt, serpentine belt, idler and tensioner pulleys are similar, but smaller. I'm an automotive tech so i see this stuff all the time and i'm always wondering what else i could use them for. Due to competition some of it is really cheap in the aftermarket.
 
...All the tool post grinders I've had apart use a spring to create preload - there is hardly axial force so it works. In such a case, there is zero need for spacers or matched pairs. You do not need the 800 bearings, nor would give any advantage.
Good info. This is where I get confused. Maybe you can correct my logic

Some of the cross sections of TPG like spindles seems they have 2 bearing systems. On the tool end, some combination of precision bearing setup; ACB sets, or other kinds of bearings with spacers, shims, cap screw adjustment... some means of setting preload so the runout is low. From what I can determine the output shaft is thus radially & axially confined to that preload setting. Then on the other (drive) end of spindle is another bearing system. My understanding is to stabilize the shaft axially obviously, but to accommodate assembly growth under temperature rise. So that end has to allow float otherwise the close fit of the ACB bearings gets squished & would become compromised. So that's where springs or spring washers & slide/float allowance would reside. If this whole assembly was a low speed (= low temperature difference) or had coolant flowing through to control temperature, the spring system wouldn't be required or not to the same degree. Am I wrong about this logic? I have some screen grabs at home I can dig up.

The problem with this stuff in the hobby world (or offshore supply world) is someone comes up with a spindle & calls it good, but you never really know if it is or isn't or what are the limitations.
 
What I have handy. I think its the basis of Quorn but maybe I'm wrong
 

Attachments

  • SNAG-4-5-2022 000.jpg
    SNAG-4-5-2022 000.jpg
    35.9 KB · Views: 13
The problem with this stuff in the hobby world (or offshore supply world) is someone comes up with a spindle & calls it good, but you never really know if it is or isn't or what are the limitations.
The price of quality and certainty is high......so I roll my own mostly :D

There are lots different spindle bearing arrangements. The Quorn one you showed works with pairs A/C's magneto and even deep groove (as I mentioned, the Boley 83 and unimat are two examples). Preload, which limits axial pay is established by springs. This is effective for grinders and light duty machines were axial loading is not that great.

When you have the bearings so far apart, you kind of need a spring as you point out, as it also provides some allowance for change via thermal expansion.

The other to way is the put the AC's together (or at least closer together), either race to race or with spacers. With matched pairs the race thickness is specially ground to create the preload - they either go together directly or with the spacers are the same length.

When using spacers to create the preload you grind them different lengths....we're talking tenths and is gleaned from those delightful tomes of prose, the bearing catalogue. The other end of the shaft can be supported many ways, commonly one or two deep groove bearings that float in the house (to allow for thermal expansion)

Here's some images of a spindle I designed and made for my BCA jig borer that shows this arrangement...it goes in the housing below it just as soon as I get the soft bearing dynamic balancer done! The sketch at the bottom, if not entirely clear, does not capture outer race of the bearing to right by the pulley....it can within the house. Axial movement is arrested the AC arrangement at the other ed. The drawing also shows spacers, but I ended up going with a matched pair. To make different length spacers, you just grind two rings, pull one off the grinder and grind the second however much more the preload calls for.

"The last diagram is another one I designed - a replacement spindle for an Aciera mill. It has two AC's at the business end and two deep grooves float at the pulley end (floats on the shaft this time). The shaft of this spindle is what is being drilled above.


DSC_7534-1300x870.JPG





DSC_8795-1300x870.JPG



bca spindle.jpg


Grinding equal spacers for an XLO spindle using matched bearing pairs

20200614_161400-1300x1733.jpg



1649191541372.png
 
Last edited:
@Mcgyver So I wonder if your Aciera mill example corresponds to attached layout I kept as reference. Can one just select a back-to-back ACB set from a catalog with defined preload deemed appropriate for the application? Seems to me the spindle housing for this would be simpler because all you need is a stop & means to clamp them together flush. That achieves the preload, end of story? As opposed to a somewhere in between preload setting which (I assume) means knowing what you are doing - trial & error, measuring play, running them in, observing temperature etc.

The last diagram is another one I designed - a replacement spindle for an Aciera mill. It has two AC's at the business end and two deep grooves float at the pulley end (floats on the shaft this time). The shaft of this spindle is what is being drilled above.
 

Attachments

  • SNAG-2022-04-05 5.20.44 PM.jpg
    SNAG-2022-04-05 5.20.44 PM.jpg
    100.4 KB · Views: 6
Can one just select a back-to-back ACB set from a catalog with defined preload deemed appropriate for the application?
Yes, thats what the matched sets are and what is shown in my spindle with the black and white cross section. Its in the first post I think, either bearing to bearing if matched (or you're ok with the 'universal' preload) or you grind the spacers to different lengths (the spacers could 1/8 long, that's not important that they are long, its like one is say .1250 and other is .1248 (if you're working with matched or going universal. As I tried to cover, universal is good for lots of applications, which lessens the need for matched pairs or stepped spacers lots of the time imo. The designer has to decide:

1) is the preload from universal AC's ok for this application
2) if not, do I want to buy matched pairs (expensive, but no futzing grinding different length spacers)
3) if not, whats the difference in the inner and out race spacer length I need to create to get the preload I want

None of it is chance or trial and error, its all to numbers you look them up in the catalogue/tech reference (most tedius and worst part of the job.).
 
Last edited:
Back
Top