"Grinding" a small spindle for ABEC 7 bearings

[Richard King]

Rich,I haven't seen the ceramic coating before. Having seen millwrights at work, I'd be more than a little worried about fragility. For a precision spindle rebuild however, I expect millwrights shouldn't be within a country mile. I looked at your link but there's nothing about the ceramic coating process. I will do some googling.

Ceramic coating projects Spindle Coating

Torrington Bearing -Grinder Spindles

www.saturnceramic.com

Spindle Shafts, Spindle Units | NB Spindle Products | Nippon Bearing

With flexible manufacturing, precision machining capabilities, and innovating surface treatments among the many key benefits, it's clear why customers choose NB for their spindle shaft manufacturing needs. No matter the industry, the NB Spindle is backed by expertise & suited for a variety of...

www.nbcorporation.com

 

[MCritchley]

Why not make a new spindle from prehard material? Your risk will be zero to the existing machine and then you can machine the nose and center taper in situ and have great concentricity.
I would use viscount 44, it’s very tough as a shaft material and machines nice. We used it at a previous job for small jet engine shafts. It’s 44 Rockwell and if your careful you won’t notice unless you have collets going in and out of the spindle on a daily basis.

 

[rhb]

Well, that didn't work worth spit. The Arkansas stone removed an unmeasurable amount of metal. Perhaps, maybe a half tenth.

Despite attempting to use the tailstock to align the test bar in the chuck, the copper crush resulted in excessive run out at the spindle journals. So I didn't try the corundum stone.

While evaluating how much metal I removed I discovered I have a tenth out of round at 90 degree orientations. I got the same values of 0.9844" and 0.9845" at the two orientations alternating orientations by turning the chuck. Much more careful measurement than my initial evaluation.

So it's time to make an inside lap. I don't actually need to have it meet toolroom specs. That "would be nice" and is worth some time and money. But the education has been very sueful.

After clocking the MT 4 test bar between centers I no longer see any point in a tool post grinder on my Clausing. A real cylindrical grinder would be a much better investment of time than attempting a single job with a TPG.

 

[rhb]

The OEM bearings are 30205 taper roller bearings. The replacements are 7205C-P4-DGA angular contact bearings. There is some ambiguity about whether that will work without a bunch of fiddle besides the fits.

There are doubtless *many* things I don't know yet, but I like project oriented education. I have zero interest in taking courses on anything.

So I simply start a project with the expectation that various things will go wrong and have to be redone. Possibly more than once. Though I do try to avoid that.

The bearing alignment issue has me considering lapping both journals and housings in a single pass using tools made to do both at once. Time to take a break and think things over.

 

[michiganbuck]

Angular contact bearings. usually, go into a counterbore and have some sort of take-up feature, it's pretty simple, the shaft or tube OD journals should be run between centers so as to be true with no wobble.
Counterbore ends should be square to the bearing OD. Bearings thrust facing the wide shoulder, and acentric marks facing the same way.

 

[rhb]

I've been looking more closely at the mounting. If I want ABEC 7 machine tool spindle bearings, I bought the wrong bearings as these are open and there is no room to add seals. I'll discuss getting the sealed versions with Boca Bearings on Monday. I opened one of the bearings to look at the markings, so I assume that's now mine.

Before I do any more work on the spindle and housing I'm considering buying a set of 30205 bearings from a top line OEM distributor and testing them.

I've been having a lot of fun on this project as it's very much outside my normal envelope. Unlike what I *have* to do which is cobble together an underground line detector so I can locate a sewer cleanout for a commercial property my family owns. I have to design some search coils for different frequencies as it's not clear what will work best. I've got a handheld portable oscilloscope/multimeter/function generator which should take care of the electronics and spare me designing and building custom electronics for a one off job.

 

[triumph406]

The OEM bearings are 30205 taper roller bearings. The replacements are 7205C-P4-DGA angular contact bearings. There is some ambiguity about whether that will work without a bunch of fiddle besides the fits.

What is wrong with taper roller bearings? My previous lathe, a TSL-800 had taper roller bearings for the spidle. Perfectly adequate I thought

 

[rhb]

There is *nothing* wrong with roller taper bearings. But ABEC 7 grade are ~$500 each in the same size from the tier 1 OEMs. Angular contact bearings are slightly cheaper, but not by much.

I bought my "ABEC 7" bearings from Boca Bearings which is a legit bearing maker and distributor for ~$60 each. For the small loads on a 6" x 10" lathe spindle either are completely satisfactory. For a larger machine, taper roller bearings are more appropriate, at least from my understanding from lots of reading. The mill head spindle uses a different angular contact bearing.

Unfortunately, I ordered the bearings before I had completely stripped the headstock and the open angular contact bearings are not usable, so I'll be getting sealed units.

I plan to price and buy a pair of 30205 bearings from Timken et al Monday solely to see what the difference is relative to the Chinese OEM bearings. From what I was able to find searching that's a $60 "folly" which I can easily afford. If they get me to less than 4 tenths over 10" I'll skip the AEBEC 7 bearings.

I also am going to be getting quotes on a full rebuild of my 10" x 20" Clausing 4902.

I go through cycles (binges??) of electronics, metalwork, woodwork/lutherie and playing guitar. I focus on one of these for 1-2 years and then shift randomly to another.

My motivation is simply learning something I don't know. A $150 book or $150 worth of bearings are just the price of the education. The only vacation that I'm actually interested in is to go to the Kensington Science Museum to see Maudslay's screw lathe and similar artifacts. I spent a lot of time in the machinery hall of the Smithsonian. Robert's planer was the most memorable.

Have Fun!
Reg

 

[rhb]

I trepanned an inside & outside lap blank from ductile iron bar.

I'm a stingy SOB so I spent $32 on hole saws to avoid turning that nice iron into chips. Go figure. Probably would not have done that if I had suitable size round bar to work with, but what I had was a 4.25' x 1.25" by 2 ' piece of rectangular bar. Making that round would not have been much fun.

So now I wait for dial bore gauges to arrive so I can finish them accurately.

 

[CarbideBob]

.... But ABEC 7 grade are ~$500 each in the same size from the tier 1 OEMs. Angular contact bearings are slightly cheaper, but not by much.

That seems cheap. I'm more used to $1500-$3200 each and long leads.
Will the new be doubles on both ends, double/single or single on both?
If single both ends decided on angle direction yet and contact angle to use?

 

[rhb]

These are 25 x 52 x 15 mm. *Very* small.Most work will be 3/8-1/2" stock in a collet for precision RF connector gauges and other similar size stuff. Loads will be very light so 15 degree should be fine. Minimal TIR is the most important requirement.

The spindle setup is single bearings 3" apart outside to outside. The bearings will be opposed as per standard practice.

I bought a pair of bearings from Boca Bearings before I took the headstock apart. When I did I discovered I needed sealed units but had bought open. I meant to call them today, but spent too much time in the shop roughing out the lap.

I'll have a crazy investment in time and money when this is done. But small precision lathes with the ability to thread are few and far between. I don't actually know of *any* current or recent production machines other than a Myford for ~$12k plus shipping. Used machines are a huge gamble without first hand inspection.

 

[triumph406]

These are 25 x 52 x 15 mm. *Very* small.

Yes very 'cute' little bearings, hence the $60 price tag. Would have expected something a little more 'beefy'

The spindle setup is single bearings 3" apart outside to outside. The bearings will be opposed as per standard practice.

3" really? No matter what you do, no matter what bearings you choose, the 3" dimension is going to be a limitation in respect to shaft and bearing stiffness.

I'll have a crazy investment in time and money when this is done.

Why? Your putting lipstick on a pig.

I see on another thread you want to get the bed ground. Again why?

It maybe a learning experience for you, but your better off getting a lathe with real potential of being a decent machine, rather the the Clausing. You might learn a lot, but your going to be dissapointed with the end result

Bit like a 60 year old Hollywood actress getting surgery and expecting to look 20 and fuckable again.

But small precision lathes with the ability to thread are few and far between. I don't actually know of *any* current or recent production machines other than a Myford for ~$12k plus shipping. Used machines are a huge gamble without first hand inspection.

Look for a Hardinge HLV, many out there that are affordable, and the easiest lathe to thread on that exists (imho)

I have a very early HLV, with a worn bed, and original spindle bearings that performs better then your Clausing ever will. and was $1k. I'm not going to, but if I wanted to get the bed ground, it's no more then $1k, and cheaper then your going to pay for the Clausing. Cut your loses while you can.

Agreed on the inspection, I would never buy ANY machine without inspecting it first.

 

[rhb]

Yeah, you're going to get an HLV reground for $1000. What then? Grinding the bed is just the start. I'm sure that's why you haven't done it.

An HLV in anything better than klunker condition is a *lot* more than $1000. And by the time someone like Schiemede rebuilt yours it would be a $25k+ piece of machinery. Very nice, but not cheap. And serious overkill for making 3/8" parts.

 

[triumph406]

Yeah, you're going to get an HLV reground for $1000. What then? Grinding the bed is just the start. I'm sure that's why you haven't done it.

An HLV in anything better than klunker condition is a *lot* more than $1000. And by the time someone like Schiemede rebuilt yours it would be a $25k+ piece of machinery. Very nice, but not cheap. And serious overkill for making 3/8" parts.

I haven't done it because despite the bed wear it still puts out very accurate parts, better then a Clausing rebuilt by the best builders in the world ever could.

Don't fall into the trap that the research you have done makes you an expert on lathes. Get a decent lathe, make a lot of parts, then you can decide whether it needs upgrading or not. And what features are important, and which aren't

If I was making firearm fasteners, the only manual lathe I'd want to use is an HLV, or CNC lathe. I have a SAG14, making small threaded parts on that would be a nightmare.

 

[rhb]

I'd like to suggest you look up the reason for the C in 5C. Come back when you know a bit of history.

I have *no* interest in persuading anyone to do the stuff I'm doing. I take the time to post it for the simple reason that *someone* might find it useful.

I'm a retired research scientist. After 10 repetitions I get bored.

 

[triumph406]

I'd like to suggest you look up the reason for the C in 5C. Come back when you know a bit of history.

Enlighten me.

I have *no* interest in persuading anyone to do the stuff I'm doing. I take the time to post it for the simple reason that *someone* might find it useful.

Your going down a rabbit hole of your own making. About the only lesson anybody is going to learn is, Don't start with a totally unsuitable lathe, and don't throw valuable time and money down a hole with zero ROI.

I know this is not about ROI for you, but as soon as you spend hundreds of hours and thousands of dollars at it, and find it doesn't achieve what you want, it's going to get pushed over into a corner. You'll have learnt a valuable lesson maybe, and not made many firearm screws or parts.

I'm a retired research scientist. After 10 repetitions I get bored.

I kinda guessed you were

 

[triumph406]

I've been the rabbit hole of converting a non running Hardinge CHNC with LinuxCNC and then a Chinese controller.

Gave up because it was going to be a big drain on valuable time, which I don't have while trying to run a business. And the end result would have been a hodge podge of Hardinge and aftermarket parts. Despite the fact that there's lots of documentation to convert CHNC's to Linux, I decided life's to short to f'k with the likelyhood of poor outcomes.

 

[triumph406]

I'm a retired research scientist. After 10 repetitions I get bored.

You would be better off getting a Monarch 10EE with a blown drive, your into electronics it seems. that would be a challenge to get that running again. Then you would have a lathe second only to an HLV.

10EE's go quite cheap at auction sometimes as well

 

[rhb]

Ever hear of Hardinge of Elmira, NY? They produced a famous series of machines for toolrooms and the watchmaking industry 100+ years ago. That series of lathes was named Cataract. The largest was the Cataract 5. All of them were small and would fit your "too flexible" sobriquet. But they were very popular. So popular the whole world still uses the patented collet system designed for them. That's the C in 5C.

You might learn something reading this:

Cataract Lathes

Cataract Lathes - high-precision American-made for toolroom and experimental shop work

www.lathes.co.uk

 

[CarbideBob]

Since you are single on each end I'd go the cup and cone route for best results.
Also if using angular ball you may consider DB mount This puts the bearings "X" outboard to reduce your runout at 12 inches.
Normally here you want a preloaded pair up front. That a big modification.
I would not be worrying tenths on bearing fit but keep in mind shoulders count also here.
Square and true within millionths on the inner?

This mount arrangement cares a lot about temperature. Under speed expect 120-140 degrees which will be different than a cold test.
The shaft will grow faster than the housing in most cases. (smaller thermal mass and less exposed to world)
You seem to be concentrating on bearing tolerances and ignoring all the other stuff.
It is the details that count.

Once again I ask "CONTACT ANGLE". Given all your research I'll assume you know why they come in flavors.

You address many as if we are not very well mentally equipped.
I think most know a Catarct lathe despite their limitations.
Perhaps you would be better at the "zone" rather than here.
I try to say this nice but "Me thinks you read too much and not enough chips in the shoes."

Given your asked for specs, size, layout and bearings it will be pure shit ass luck to nail it but go ahead as each step is something learned and for sure I did it the hard way.
You have come here wanting to be aggressive. Do not be upset if it comes back.

My world is microns and sub in tracking so maybe past what you are trying to do.
Bob