[handlewanker]

That Starret type of indicator design is very old fashioned now that the Mitsutoyo and other dial indicator makes with the rectangular long body shape having 2 dovetails for mounting and the longer slimmer indicator tip, are on the market.

The Starret type came onto the market many years ago, but due to the short stubby indicator tip is not so highly desirable now, it fulfilled a need to replace the large bulky older type of dial indicator for clocking on the mills etc in awkward small bores.

I had the John Bull boxed indicator set and it came with a right angle attachment with a long leg to get into bores and recesses, but was very bulky.

I now have a Mercer dial indicator, same type of tool with the square body, slightly different end, better clocking ability with it's longer sensitive tip, but I would not consider this to be the way to do the job by inserting it into the bore for 2"......BTDT......the bung is as accurate as the hole in the 4 jaw, and once set in will not require resetting.

The bore is 6" long.........the Starret dial indicator max reach is ? long......the best you can get into the bore and still see the clock dial is approx 2".....this may be "good enough" for most people, but I went to a different school.

The bung method requires only a few taps with a soft hammer to get the 1.75" diam bore true.

I could name possibly 1/2 a dozen methods that would work too, some more accurate but more time consuming than the rest, but essay writing is a bore.

The very very simplest way is to use two faces at 90 deg on the outside of the block as your datum faces and set the block in two planes true to the long axis.

You won't have to worry if the 1.75" hole is out once it is running true at the end......the other end is automatically true if you've set the faces of the block square in two planes to the longtitudenal axis in the first place.

The point is, if'n you start off right you end up right without a lot of drama.

As Christopher balks at the bung....LOL, I would suggest then that in the first instance the block be set central in the 4 jaw, but with two faces (90 deg to each other) set true to the long axis of the bed.......this is as simple as you can get and is a simple clock/Saddle wracking exercise, running the clock with the saddle along the block faces untill both are true.......then adjusting the jaws slightly to get the centre location right....it doesn't even have to be in the centre of the block and could be to one side of centre......the result is still the same.....the bore will be true to the block.

When you come to do the second bore, provided you clock the faces true by the same saddle wracking/clock method the bore will be once again in line, and an adjustment for the 1.75" hole to centralise it is kiddies stuff.

I think we ought to see a sketch of the design for the block, some alarming design aspects are coming to light.
Ian.

[christofer]

Here is a drawing of the spindle I intend to build.

[handlewanker]

Hi Christopher, this is starting to look good, very doable.

I assume the spindle, being a bought in part, already has the step behind the A/C bearings which means the spacer has to be bored with 2 diams in the bore, unless you "modify" the spindle slightly and make the spindle parallel all the way up to the back of the A/C bearing seat.

The spacer bore needs to be a very close fit to the spindle diam, .04mm (.002") max clearance, so that it will automatically run true and concentric with the spindle, otherwise, with too much clearance in the spacer bore you could have the spacer running out of centrality and so an impossible balancing situation could occur with the spacer acting like a bob weight.

The design with the nut at the top of the spindle securing the whole lot together does make assembly much easier but belt and braces as the cover on the A/C bearings does this too, so if the nut securing the A/C bearings was directly behind the A/C bearings, so removing the spacer from the design, then the ball race could butt against a shoulder on the spindle and a second nut on top of the pulley would secure the ballrace in position whilst allowing the ballrace outer race to float in the housing.

If no shoulder is available on the bought in S3 spindle, a circlip groove could be machined on the spindle behind the ballrace and with the second nut used to secure the race and drive pulley in position as no thrust is anticipated in this area.

A spline or twin keyways are needed for the pulley drive, but you might get away with one keyway if some balancing is attempted.

You will also need to finalise the seals in the A/C bearing cover, and the ballrace would be a double sealed variety.

It is also possible that two sealed ballraces at the top would increase bearing life and spindle accuracy for very small increased cost, as the belt drive does place a load on one side of the bearing, and any looseness that develops here would show up at the cutter tip, but if the motor is a direct drive to the spindle, only one bearing is needed.
Ian.

[wizard]

Here is a drawing of the spindle I intend to build.

I would watch out with the end cap, you don't want it to rub on the inner races. Obviously there is no detail in the drawing so maybe you already have made provision. Also a seal of some sort might be in order for that end of the spindle.

Other than that it looks like a pretty normal spindle. Post pics after the build.

[RICHARD ZASTROW]

You might consider a Spieth clamping sleeve instead of keyways to secure the pulley to the spindle shaft. We've been using them for years in our gear hobber designs without any failures.

They actually hold better than key/keyways. The shaft doesn't lose strength as with keyways. No balance problems either plus easy install and removal.

Dick Z

[mactec54]

christofer

Your Bearing placement, Option (1) is incorrect for spindle use, Option (2) is the correct way
for your spindle bearings to be installed, Its called Back to Back when installed this way

[RICHARD ZASTROW]

christofer, Mactech54 is correct on the back to back arrangement for spindle bearings.

Also, the suggestion for a seal is a good idea. I would use ABEC 5 A/C sealed bearings, especially in a vertical orientation.

Another consideration is axial thrust such as plunge milling or drilling. You may want more angle in your angular contact. I like to use 25 deg. with medium preload. The axial load increases when drilling etc. and the bearings can take the thrust (up to their design limits). A smaller angle and/or lighter preload can "unload" the second bearing of a pair when pushed hard as in drilling.

Just a thought

Dick Z

[christofer]

Here are pictures of what I have for this build so far. A drawing of the spindle is here.

The bearings I have are nachi brand and are all C3 rated. The ACs are 40 degree contact angle. For preload I intend to experiment and possibly remove a bit of material from one of the bearings with some fine grit sand paper to get some preload.

I know these bearings are cheap and may not work well but for $20 each it was hard to pass up. I am unsure if these bearings will bind when put back to back due to their runout. Perhaps I can selectively fit them so that they do not bind.

The spindle I bought already has a spline so I intend to cut a matching spline in the drive pulley. The spindle is long so I will cut off some of the spline once everything is set up.

[JohnZ]

That is the SX3 spindle from little machine shop? I bought the same one but have not gotten to building it yet. I am eager to see how yours turns out.

[christofer]

JohnZ: Yes this is the SX3 spindle. I hope it doesn't have much runout and is not bent!

I have some new findings to share.

I went to the school last night with the intent to start machining the spindle spacer and spindle housing. What I ended up doing was cleaning the lathe... I put a test bar in the three jaw chuck that was mounted on the lathe and found it had about 15 mils of runout! I removed the jaws, cleaned and oiled them (it looked like it had been years since this had been done). This did not fix the problem so I removed the chuck and cleaned the area where it seats with the spindle of the lathe. While I had the chuck removed I indicated the spindle of the lathe and found it to only have a few tenths of runout. I put the chuck back on in all three orientations and found one that gave about 7 mils of runout.

Needless to say this chuck is worn out! Anyway, the school happens to have a four jaw chuck that doesn't seem to have been used much (probably because no one knows what it is) that I will end up using for everything. I measured its inside diameter for the bung and found it to be 2.210". I suppose I better indicate this inside diameter too so that I can be sure it is concentric to the spindle. I assume it will be as this chuck has not seen much use.

I suppose the runout in the three jaw chuck is not a huge deal as long as you can do all of the machining you need in one operation. If the piece needs to be removed or flipped there would be no way of getting it back in the correct position.

Question: I see some people using timing belts to connect their servos/steppers to their ballscrews. This would be advantageous for me to use because I can get some reduction and end up using a higher lead screw (which is easier to find). How much play do these belts have in them? Will using a belt drive introduce much backlash?

[wizard]

JohnZ: Yes this is the SX3 spindle. I hope it doesn't have much runout and is not bent!

I have some new findings to share.

I went to the school last night with the intent to start machining the spindle spacer and spindle housing. What I ended up doing was cleaning the lathe... I put a test bar in the three jaw chuck that was mounted on the lathe and found it had about 15 mils of runout!

Are we sure about that? 0.000015 would be extraordinary accuracy for a three jaw chuck. Even the test bar would need to be exceptionally accurate to get that sort of run out.

I removed the jaws, cleaned and oiled them (it looked like it had been years since this had been done). This did not fix the problem so I removed the chuck and cleaned the area where it seats with the spindle of the lathe. While I had the chuck removed I indicated the spindle of the lathe and found it to only have a few tenths of runout. I put the chuck back on in all three orientations and found one that gave about 7 mils of runout.

Needless to say this chuck is worn out! Anyway, the school happens to have a four jaw chuck that doesn't seem to have been used much (probably because no one knows what it is) that I will end up using for everything. I measured its inside diameter for the bung and found it to be 2.210". I suppose I better indicate this inside diameter too so that I can be sure it is concentric to the spindle. I assume it will be as this chuck has not seen much use.

I suppose the runout in the three jaw chuck is not a huge deal as long as you can do all of the machining you need in one operation. If the piece needs to be removed or flipped there would be no way of getting it back in the correct position.

Generally you don't use three jaw chucks if accuracy is your goal.

Question: I see some people using timing belts to connect their servos/steppers to their ballscrews. This would be advantageous for me to use because I can get some reduction and end up using a higher lead screw (which is easier to find). How much play do these belts have in them? Will using a belt drive introduce much backlash?

Actually a belt drive might reduce backlash a bit. For the most part you need some sort of coupling between the motor and lead screw, these can introduce backlash, resonance and other problems. Speaking of resonance timing belt tension can be critical to good operation of a servo system. It is the old not too tight and not too loose game.

Generally I consider belt driven servo systems far easier to work on. At least it seems that way as direct drive (though a coupling of some sort) always seem to be inaccessible or have poor mechanical connections to leadscrew and motor. Plus with belt drives you are always free to change ratios.

[christofer]

Are we sure about that? 0.000015 would be extraordinary accuracy for a three jaw chuck. Even the test bar would need to be exceptionally accurate to get that sort of run out.

I think maybe I used the wrong word? I suppose I should have said thou. I mean to say that the chuck had 0.015" of runout.

Is mil the correct term for 0.001"?

[wizard]

I think maybe I used the wrong word? I suppose I should have said thou. I mean to say that the chuck had 0.015" of runout.

Is mil the correct term for 0.001"?

In other words 1 mils would be 0.000001".

As to chuck run out even 15 thousands isn't surprising for a 3 jaw chuck. Yes it could be better, but I wouldn't be surprised to find far worst being used in industry daily.

[handlewanker]

Hi Chris, you won't be wanting to hold the square material in the 3 jaw anyway.

The biggest problem with 3 jaws is they go bell mouthed which means they grip the material at the back and not at the front of the jaws......grinding out is an option but needs an essay to describe the method.

If you intend to use the bung method, the 4 jaw with the 2.210" hole being smaller than the A/C bearing hole in the block means you can turn the bung from a piece of round bar in the 3 jaw in one operation.

Make the bung with the 2.210" diam first, 3/8" long, then the A/C bearing bore diam size also 3/8" long.

If you are using a longish piece of round bar for the bung, I hope your parting off skills are OK as you'll be parting material off at least 3" diam...(100rpm for HSS parting tools) or whatever the A/C bearing diam is....drilling a 1" diam hole up the middle helps when the parting tool breaks through, and you can also use a revolving centre lightly in the I"diam bore to support the job against parting pressure.

If you only have a short piece of material to make the bung with, then just allow 3/4" longer to hold in the 3 jaw, make the diams to size and length, turn it around in the 3 jaw and holding it on the 2.210" diam carefully face off the excess piece, and forget about parting off.

The bung only needs to be a light push fit in the chuck body, and .001" clearance on the 2.210" diam and the A/C diam will locate it OK.

Don't grip the block in the 4 jaw with huge force as you might distort the A/C bore.

It will pay you to make sure the block starts off square and parallel by lightly skimming all faces and set in line before initially boring the A/C bores and 1.75" hole.

It is almost a certainty that the 1.75" diam hole will be tapered....all lathes bar none cut tapered, but this is of no concern as it's just a clearance hole and will only be used to dial up to when you do the second end.
Ian.

[handlewanker]

Oh dear, a splined shaft........that in itself is a major problem.

The spline is a multi key sliding fit method of ensuring a drive while the pulley is stationary and the shaft moves, as in a drill press.

Attempting to fit a pulley to a splined shaft is masochistic without a dividing head and slotting head.

You can plunge a single keyway in the lathe with a boring bar held on it's side and racking the saddle back and forth.

You can also plunge a spline in the lathe PROVIDED you have some means of indexing the pulley to plunge each spline groove exactly.....a loosely fitting spline is totally useless.

In this case the spline, being already on the spindle, is a pain.

You could bore the pulley a sliding fit to the OD of the spline, and fit 2 keys opposite one another to fit into the splines in the spindle, locking each key with a grub screw.

There is another way that does not require keying, and that is before boring the hole, cut the pulley boss in half through the middle up to the belt diam, and cut down one side to the middle to give you a loose half moon shaped segment and then drill and tap the side of the boss for 2 cap screws.

The half moon shaped block is then fitted to the pulley with 2 cap screws with a shim between the faces, and the pulley and half moon shaped block are bored out to the OD of the spindle spline.

You now have a pulley with a loose side piece on the boss that becomes the clamp to attach the pulley to the splined shaft diam.....no keys needed.....fits any plain shaft diam.

I have several Vee pulleys that use this method and can post a photo if more detail is needed.

The best method would have to be the taper lock, but in this size it would be quite fiddly to make.

The alternative to the taper lock could be a split tapered sleeve that has a screwed collar on the top to draw the sleeve out and lock it to the spindle.

A simple balancing method will be required for the pully after "adjustment".

This area unless resolved will be a problem.

If this design using the splined spindle gains acceptance, then a split taperd sleeve, with or without a splined bore, and a screwed collar on the end fitted to a pulley with a tapered bore could be marketed as a package to allow pulleys to be firmly attached to spindles.
Ian.

[christofer]

Ok, I think I prefer the idea of leaving the spline that is already on the spindle in place and clamping a pulley to it. I Believe I will be using a poly V belt to drive the spindle. I plan to cut my own pulley out of aluminum and use a clamping mechanism to keep it in place.

[RICHARD ZASTROW]

christofer, You could install a splined bushing in your pulley.
Sleeve Bushings - Grob, Inc.

Dick Z

[handlewanker]

Hi Rich, that's a fantastic idea, instant splined fitting.

I think the mild steel flanged bushing would be the one to go for, Loctiting it to the splined shaft (flange at the bottom)and making the pulley a sliding fit to the bushing with screws to hold it on.

The flange would need to have the holes tapped so that the pulley could be slipped on from the top and secured with cap screws countersunk in the top face.

Balance would not be a problem with that method as all diams and fixings are concentric.

The other alternative is to cut the splined shaft down to length and turn the splines off to give a parallel shaft, but it might leave too small a shaft diam which is not too desirable, and then fixing the pulley to the spindle is also difficult.

I think I would prefer the splined bushing as it's straight forward and neat.

Loctiting the bushing to the spindle (a press fit is DEFINATELY NOT recomended here) does make it more difficult to dissasemble the head should maintenance be required, but a small amount of heat (150 deg C) to the bushing will make the Loctite go soft for removal, so that might not be too much of a problem after all......bearings don't require removal unless knackered, and that is many hours of useage.

The next design detail is the retainer cover for the A/C bearings and needs to have the seal properly implimentad.

Rich Zastrow mentioned a labryinth Z type seal by SKF (?) some time back so that could be looked at for ease of a sticky problem.

I would not "adjust" or increase the preload to the A/C bearings initially before they are run and bearing temps measured.

BTW, looking at the photo of the spindle I see there is a threaded section at the end of the splines.......this means you will have to use the long spacer method between the bearing sets to clamp the inner race of all the bearings together.

Personally, and it does require more work, I would turn the long diam behind the A/C section down and thread it to allow a screwed collar directly against the A/C bearings to secure them in place, so doing away with the long spacer.....just my preference.
Ian.

[RICHARD ZASTROW]

christofer, I'd use A/C bearings with integral seals. They're available from most bearing manufacturers. I like to use GMN because they seem to have the ones I want in stock or short delivery.

Don't confuse SEALS with SHIELDS. Shields don't seal very well, especially in a vertical spindle attitude.

Dick Z

[wizard]

christofer, You could install a splined bushing in your pulley.
Sleeve Bushings - Grob, Inc.

Dick Z

Or try something like this: Ball Spline [ THK || USA ].