[wizard]

shrink fitting doesnt achieve preload. the bearing is preloaded from the factory as ive said twice now.

shrink fitting is viable, we do it on ball screws. most commercial spindles WONT have this because it makes things unservicable, but it WILL most definitely work, and is a good option for an inexpensive high speed setup.

This item I have to disagree on, many a lathe I've worked had the inner race for the nose end shrunk in place. Yes it is a bit of a problem removing said bearings, but it isn't something you do everyday. Experience with mills is more limited at the very least though some of those have fits requiring a stiff press fit.

as for deep groove radial tail... kavo, jaeger, colombo, tekno, hsd...

the sieg x2 comes with a deep groove raidial bearing in the tail AND nose.

you forget hes talking about a cheap hobby "make it yourself" spindle. one that should cost less than a single skf p4 bearing you might find in a VMC spindle. its not about being the best ever, its about being cheap and functional.

Yeah this point was obviously missed by many here.

[Chrisjh]

Hi All,

I have been off line for a while but based on an inspiration from Mactec's link, I have come up with another design based on a readily available low cost ER32 straight shank Collet Holder.

Details for the collet holder including price can be found here: ER32 20MM x 150MM LONG NOSE STRAIGHT SHANK COLLET CHUCK #F71

I have plenty of ER32 Collets from Ø2 to Ø20 in my tool kit (sourced from CTC Tools in the link above) as well as the R8 Tooling.

I guess there is a need for 2 design approaches, one for R8 and one for a lower cost ER32 approach so I now have 2 designs available.

Thanks to the feedback and advice I have had from this forum, I have tried to incorporate the best of all suggestions into the simple ER32 design. The most advantageous part of this design is the use of a very low cost straight shank collet holder, which removes the need for complex taper machining and hardening process.

The lower bearings are grease loaded Angular Contact, mounted back to back and preloaded via a wave washer and tube arrangement as suggested by Mactec. The upper bearing is a ”slip fit” in the housing which allows it to move longitudinally with thermal expansion, compressing the wave washer and maintaining preload tension.

Speeds of up to 10,000rpm could be possible.

I considered starting a new topic but am not sure of forum protocol, so let me know what you think.

Regards

Chrisjh

[handlewanker]

Hi all, just to add a bit more info to the ER collet Versus the R8 system, you only need 19 collets to go from 1mm up to 20mm with ER32, but it would mean a thousand or more R8's ('cos they only fit 1 size and 1 size only) if'n you wanted to cover all odd sizes in .1mm increments etc.

One of the main reasons for shrink fitting bearings to shafts is that when a bearing is fitted and if it's a slide fit, it has to be EXACT to make it slide....a couple of tenths either way and you have a sloppy fit, whereas a shrink fit is always exactly tight.

When I was fitting gearboxes on the mines we always fitted bearings using hot oil for the shrink fit, and to remove a worn bearing we just hit the one side of the inner race (after the outer race and balls/cage were removed), with an Oxy/acc torch flame for a few secs and it fell off.....bearings were about 100mm diam angular contacts mostly.

I just had a look at the design of the ER32 spindle you posted in post #102.......I may have missed something in previous posts, but in the design you have two A/C's at the bottom and one deep groove, whatever, at the top, held rigidly in the housing by the outer race, applying pressure on its inner race with a wave washer and so on to the lower AC bearing inner race via the tube........what holds the spindle in the lower A/C bearings......the top bearing floats on the spindle.....you would have to have the bottom one of the A/C bearing pair shrink fitted to the spindle as the main locating point of the bearing assembly, and it's back to back mate above it a sliding fit to allow it to apply pressure to the pair.....this doesn't make sense as the bottom A/C bearing is being pushed out of it's shell.

What you have then is the bearing at the top of the spindle applying preload to the upper of the bottom A/C bearings and the bottom A/C bearing just doing nothing.

As I see it, the only way you can utilise this hardened and ground spindle with the ER chuck attached, but without having any method of screw cutting the spindle to apply the conventional bottom bearing locating screwed collar to lock the A/C bearings back to back, is to Loktite a screwed collar assembly to the spindle to apply pressure to the inner races of the 2 A/C bearings and so fix them to the spindle.....the very top bearing just floats as normal......gonna have another look at the bearing set-up to see if'n I've misinterpreted it.

Hi Chris, ignore all this, as Mactec54 has shed the light, (post #107) and all is now clear, no problem with design.
Ian.

[mactec54]

Chrisjh

I have done a ER 20 / 25 & a 32 they all work great, watch buying from CTC as there runout is anywere from .0006 to .003 I have not had one from them that has been in spec usally around .002+ runout

I regrind the Taper once they have been assembled, so the runout is not a problem for me

Your new design, use steel or 303 ss for your spacers, Brass/aluminum parts are not a good choice of material for spindle parts

I like the ER25 & the ER32 both of these sizes are better than a R8

[mactec54]

handlewanker

The spindle is retained at the top when the pulley is in place, with a end cap & shcs this locks the spindle in the front bearings & the rear bearing altogether

I'm now using Poly-v belts for the pulleys

[handlewanker]

Thanks Mac, I've also bought an ER32 spindle 100mm long for a project and wondered how you'd retain the spindle in the bearings.....clear now.

Dahhhh, I'm getting dimmer as I get older......did you drill and tap the end of the spindle....I thought it was hardened, or is it just a case you can break through to a soft core?
Ian.

[mactec54]

handlewanker

Most already have a hole that is already tapped in the end, but in most cases it does not run true, The solid one's with no hole, Machine some off the end & then you can drill & tap the end, Most are case hardened, the core is still tough, but you can drill & tap them ok

On the threaded one's, I put a put a threaded (Bolt/Capscrew Etc) piece into the end of the ER spindle locktite it in, & retap a new thread in the end, so it run's true to the spindle shaft

[handlewanker]

Thanks Mac, I think I'll go and give the spindle a test spin to see if'n I can see any runout in the taper.......the worrying part is the collets....ain't no way they can be "trued up".....just bought a 19 piece set (2mm-20mm) from China, worked out to $5 each with postage!!!!!....just one of them bought at out local supplier (also Chinese import, probably same factory....LOL) cost $29 each.....I'm in the wrong job, metaphorically speaking.
Ian.

[wizard]

One thing that hasn't been pointed out in this thread is that if you want a slip fit due expansion or other issues it is best to have the slip happen on the outer race. **A slip fit on an inner race is a very bad thing and leads to premature bearing of shaft failures. *The inner races should either be a press fit or locked in place with a lock nut. *The only common exception is if the bearing away from the nose is used to adjust preload. *

AC induction motors are assembled this way, that is press fits to the motor shafts with the outer races a slip fit in the bells. *While thermal expansion is an issue this also allows the rotor to center in the stators magnetic field. *Of course on machine spindles you don't want the nose moving around thus that is restrained and the opposite end allowed to float. *

Hi all, just to add a bit more info to the ER collet Versus the R8 system, you only need 19 collets to go from 1mm up to 20mm with ER32, but it would mean a thousand or more R8's ('cos they only fit 1 size and 1 size only) if'n you wanted to cover all odd sizes in .1mm increments etc.

One of the main reasons for shrink fitting bearings to shafts is that when a bearing is fitted and if it's a slide fit, it has to be EXACT to make it slide....a couple of tenths either way and you have a sloppy fit, whereas a shrink fit is always exactly tight.

When I was fitting gearboxes on the mines we always fitted bearings using hot oil for the shrink fit, and to remove a worn bearing we just hit the one side of the inner race (after the outer race and balls/cage were removed), with an Oxy/acc torch flame for a few secs and it fell off.....bearings were about 100mm diam angular contacts mostly.

I just had a look at the design of the ER32 spindle you posted in post #102.......I may have missed something in previous posts, but in the design you have two A/C's at the bottom and one deep groove, whatever, at the top, held rigidly in the housing by the outer race, applying pressure on its inner race with a wave washer and so on to the lower AC bearing inner race via the tube........what holds the spindle in the lower A/C bearings......the top bearing floats on the spindle.....you would have to have the bottom one of the A/C bearing pair shrink fitted to the spindle as the main locating point of the bearing assembly, and it's back to back mate above it a sliding fit to allow it to apply pressure to the pair.....this doesn't make sense as the bottom A/C bearing is being pushed out of it's shell.

What you have then is the bearing at the top of the spindle applying preload to the upper of the bottom A/C bearings and the bottom A/C bearing just doing nothing.

As I see it, the only way you can utilise this hardened and ground spindle with the ER chuck attached, but without having any method of screw cutting the spindle to apply the conventional bottom bearing locating screwed collar to lock the A/C bearings back to back, is to Loktite a screwed collar assembly to the spindle to apply pressure to the inner races of the 2 A/C bearings and so fix them to the spindle.....the very top bearing just floats as normal......gonna have another look at the bearing set-up to see if'n I've misinterpreted it.

Hi Chris, ignore all this, as Mactec54 has shed the light, (post #107) and all is now clear, no problem with design.
Ian.

[wizard]

My question is this do you have a favored brand or series of poly-v belts? I'm looking into a few things to build around the shop and for the most part poly-v seems to be the way to go.

handlewanker

The spindle is retained at the top when the pulley is in place, with a end cap & shcs this locks the spindle in the front bearings & the rear bearing altogether

I'm now using Poly-v belts for the pulleys

[mactec54]

wizard

I have always used the Poly-V Belts for many years, The Orings are still good for Engraving spindles, But these bigger spindles need the Poly-V Belts to drive them so you don't get any slip

I like the Gates belts as my choice, But other manufacturers work fine as well

Here is one I use, they will cut the belts to any number of grooves you want

Contact Us - Belt Store

[handlewanker]

Hi Wizard, thanks for the enlightenment on the preferred "loose" fit for a bearing in the slip mode.

I would definately not make a bearing a press fit to the shaft as in the upper bearing in the spindle design, better to have it attached between a locking nut and a spacer on the spindle and as you state let the outer race move in the housing.

A bit of a worry if'n the bearing seized up and wore the housing out.

Some time back I had to bore out and sleeve the housing on a 1 HP motor that had a seized bearing in it.

The previous owner said "It was a bit noisy for a while, but now it's gone quiet".

I didn't twig on to the inference of the "gone quiet" bit till some time later when I fitted it to my lathe and after a while it started to get hot and also made funny noises when the rotor rubbed on the stator.

I think the bearing was "hammered" to death by the centreing of the rotor in the stator field, as when the motor started it always gave a "thump" and a wirring noise....played havoc with the centrifugal switch too...long story......cured it with a thrust race and spacers in each housing end.....that was 25 years ago....going to go 3 phase with VFD now.
Ian.

[mactec54]

handlewanker

The top spindle bearing in my spindle builds have there own housing, the housing slides with any thermal expansion, which is a very small amount of movement, The bearing is a normal fit in it's own housing

[wizard]

Hi Wizard, thanks for the enlightenment on the preferred "loose" fit for a bearing in the slip mode.

I would definately not make a bearing a press fit to the shaft as in the upper bearing in the spindle design, better to have it attached between a locking nut and a spacer on the spindle and as you state let the outer race move in the housing.

A bit of a worry if'n the bearing seized up and wore the housing out.

This can happen but more often than not a seized bearing will tear up the shaft.

Some time back I had to bore out and sleeve the housing on a 1 HP motor that had a seized bearing in it.

The previous owner said "It was a bit noisy for a while, but now it's gone quiet".

I didn't twig on to the inference of the "gone quiet" bit till some time later when I fitted it to my lathe and after a while it started to get hot and also made funny noises when the rotor rubbed on the stator.

I think the bearing was "hammered" to death by the centreing of the rotor in the stator field, as when the motor started it always gave a "thump" and a wirring noise....played havoc with the centrifugal switch too...long story......cured it with a thrust race and spacers in each housing end.....that was 25 years ago....going to go 3 phase with VFD now.
Ian.

Nice to highlight the use of 3Phase motors and VFD, I highly recommend this approach as it isn't anywhere near as expensive as it was in the past. I use to work on old PTI drives, right down to the ring counters. A half horse power drive occupied a lot of panel space, now ten times the functionality fits in the palm of your hand.

[handlewanker]

Hi Wiz, I fully suscribe to the bearing in housing approach to the spindle top bearing design, but only if'n it has room for the sleeve and is prevented from turning......slot and pin etc.

It means more machining, but is the only safe way to go.

I am considering another approach for the spindle of a Beradi (Italian make) jig borer I bought many years ago and want to re-make the spindle from 2 Morse up to 30 int taper.

The top bearing in my case is probably (jury still out) going to be a straight roller bearing, fixed in the housing, and the fact that roller bearing inner races with rollers are able to move sideways in the bearing without restraint will accomplish the movement that will occur from any thermal expansion.......I toyed with the idea of a needle bearing but that was a bit too radical and problematical design wise, but nevertheless an exciting prospect, as the bearing length would offset any tendency for sideways deflection.

Previously the re-design was for two plain sealed ball races at the top, loose in the housing, but with the possibility of housing wear, if ever, I decided to re-think the layout.

The previous re-design had a sealed ball race so lube was built in, whereas the roller bearing design calls for housing seals top and bottom to keep the grease in, dirt out.

It has been pointed out that there is some side deflectiion in ball races....practically non existant for all intents and purposes, and definately not measureable to any degree worth worrying about, and the roller bearings, being non taper, will also have this "problem" but not a problem as I see it for practical uses.

As the quill has a thin wall at the top, I will make a screwed retainer with a seal to retain the bearing outer race against a lower shoulder, and a screwed collar on the spindle to retain the inner race against a spacer on the spindle that retains the two angular contact bearings at the bottom.

The 3 phase motor interest came about when I bought a small (3/4 HP) VFD on Ebay for $250, and the fact that 3 phase motors are practically given away and can be speed controlled, which single phase can't, was a deciding factor.....the housing package is also a lot smaller for 3 phase.

It all started out when I found out about how to make a rotary phase converter for my Bridgeport mill, but the non ability to vary the capacitors to match the fluctuating load made a "buy in" approach necessary instead of making one.

The capacitor pack is "adjusted" electronically by a sensor in the circuit board and maintains the voltage balance on the three phases, which a home made rotary converter can't do.
Ian.

[handlewanker]

BTW, the design of a spindle is decided by the manufacturing process, and not always by the best way to achieve a bearing layout.

When you produce a 1000 items for sale and can makle it $1 cheaper by "adjusting" the bearing design, you make $1000 without having to increase the selling price.

This means that whatever design the Taiwanese, Chinese or any import or home produced piece of machinery is made to, it will always default to an economic aspect as the design feature.

Design is driven by economics, so when we look at a home made, or better said, custom made design, the best bearing layout, material and feature content we want to incorporate, the better will be the outcome.....even if'n it would be an economic disaster and not competitive.

So, for all things that go round and round, most of the materials today will work in one configuration or another and depending on the machine shop availability.

Two designs may start out with the same characteristics, but in the very fullness of time the better design will will not necessarilly be the longest lasting......it depends mainly on the useage.
Ian.

[wizard]

Hi Wiz, I fully suscribe to the bearing in housing approach to the spindle top bearing design, but only if'n it has room for the sleeve and is prevented from turning......slot and pin etc.

I think you are excessively concerned about the outer race causing trouble. I'd just follow the example on your jig bore.

It means more machining, but is the only safe way to go.

The other way to look at it is that you add more variables to your machine. Remember kiss.

I am considering another approach for the spindle of a Beradi (Italian make) jig borer I bought many years ago and want to re-make the spindle from 2 Morse up to 30 int taper.

The top bearing in my case is probably (jury still out) going to be a straight roller bearing, fixed in the housing, and the fact that roller bearing inner races with rollers are able to move sideways in the bearing without restraint will accomplish the movement that will occur from any thermal expansion.......I toyed with the idea of a needle bearing but that was a bit too radical and problematical design wise, but nevertheless an exciting prospect, as the bearing length would offset any tendency for sideways deflection.

Previously the re-design was for two plain sealed ball races at the top, loose in the housing, but with the possibility of housing wear, if ever, I decided to re-think the layout.

I'd follow the original design as close as I could.

The previous re-design had a sealed ball race so lube was built in, whereas the roller bearing design calls for housing seals top and bottom to keep the grease in, dirt out.

It has been pointed out that there is some side deflectiion in ball races....practically non existant for all intents and purposes, and definately not measureable to any degree worth worrying about, and the roller bearings, being non taper, will also have this "problem" but not a problem as I see it for practical uses.

As the quill has a thin wall at the top, I will make a screwed retainer with a seal to retain the bearing outer race against a lower shoulder, and a screwed collar on the spindle to retain the inner race against a spacer on the spindle that retains the two angular contact bearings at the bottom.

The 3 phase motor interest came about when I bought a small (3/4 HP) VFD on Ebay for $250, and the fact that 3 phase motors are practically given away and can be speed controlled, which single phase can't, was a deciding factor.....the housing package is also a lot smaller for 3 phase.

Eventually the message will get out there and three phase motors won't be cheap anymore. At work we are getting really good results from small three phase motors running off inverters.

It all started out when I found out about how to make a rotary phase converter for my Bridgeport mill, but the non ability to vary the capacitors to match the fluctuating load made a "buy in" approach necessary instead of making one.

The capacitor pack is "adjusted" electronically by a sensor in the circuit board and maintains the voltage balance on the three phases, which a home made rotary converter can't do.
Ian.

[handlewanker]

Hi Wizz, verry true....KISS in all cases, complication achieves not much except instability in one way or another.

I find the ability to ramp up and down the speed is a biggy in the feature parade, but the biggest biggy is the variability of the drive.....no belt changing.

At the same time the design using VFD must cater for the motor cooling which suffers if'n you direct drive and expect to maintain torque at low speed without overheating.

One reason I would always have a belt drive primary input to keep the motor turning at a reasonable speed for cooling, also I like the 3,000 rpm motors for speed control applications where you have the ability to go down to about 50 revs on the machine and the motor still running at about 200 revs......sacrificing if necessary the very top speed range.

If'n the top speed is needed and the bottom too, I'd go to a 2 step pulley to either go in high or low range, maybe 3 step too if'n the bearings could stand 5,000 rpm.

One of my projects is to re-motor a small drill press to a 3/4HP 3 PH motor with VFD, one of those Chinese imports that has a 1/10 HP motor and a 4 step pulley.

But I want to have a single step pulley, Poly Vee belt drive with 1:1 ratio, the VFD can do a 200% increase in the speed......I always wanted to drill 1mm at 5,000rpm...LOL.....hope the bearings are better than just deep groove ball radials.
Ian.

[mactec54]

handlewanker

The top Bearing can be a Roller Bearing or a Needle Roller Bearing, I have used both in spindles & motors

These both work better than a Ball type Bearing, But cost quite a bit more for a precision Roller Bearing to do the job ABEC5 minimum

If you don't have a Precision Bearing, there is to much clearance in a standard Roller bearing for a spindle, & you must have a grease retainer for them, at high speed they throw the lube out & then run dry

A caged Needle Roller works best for high speed, & holds the lube in better

As for running VFD & having cooling problems,There is no need to have a cooling problem, I place a small fan above the motor that turns on from the VFD, it keeps the motor cool at high speed & low,& is wisper quiet, I remove the motor fan as well, the photos are of a Bridgeport I did, the motor is a standard Bridgeport 3450RPM motor with all the steel casings removed,This one run's up to 6000RPM With poly-V belt at 1/1 Enjoy

[handlewanker]

Hi Mac, that's a pretty fancy Bridgeport rebuild......with that smart powdercoat finish (?) I bet you don't dare use it in case a chip makes a mark on the paintwork....LOL....only kidding.

I like the idea of the bearing at the top of the spindle being a caged needle type.....pretty compact, and will probably be the design I'd like to use.

The body of the quill at the top originally had a regular ball bearing, as the spindle ended in a splined shaft that went through the top pulley with it's own bearings in the top of the head, so the top spindle bearing only had little work to do.

I'll most likely have to bore the quill out a bit deeper to take a longer needle bearing and make a sleeve for the spindle to take the inner race if one's available as I don't anticipate making a hardened sleeve for the inner race.....If'n it gets too complicated I'll go to the precision roller bearing with seals and be done with it......the original ball bearing would probably be somewhere near the dimensions for OD and ID of the roller bearing anyway....early days yet.

BTW, on the subject of the ER collet spindle nose, and the problem most would encounter in hardening and subsequent grinding of the taper, I suppose the Hard turning approach could be used......it only needs one taper to be cut......but grinding would be nice and relatively easy by comparison if'n you have a tool post grinder for the lathe.

I've never done any hard turning, and only read about the method, so another trick to maybe try when the time comes.

Actually, I will be using the ISO 30 taper with ER32 chuck in my spindle end, and various other ISO 30 tools too.

I have a problem with the draw bar.....the spindle ends in a spline shaft that goes up through the drive pulleys, like as in the Bridgeport drive, and that being only 12mm diam at the core of the spline rules out a hole for the draw bar, but the ER 32 collet chuck in ISO 30 taper has a through hole, so I'll use a draw bolt up through the collet chuck and into the end of the spindle inside the ISO taper.
Ian.