[Matt Ribble]

First off - Apologies if this is the wrong forum for this.


I've been entertaining the idea of building my own spindle for some time now and I feel overwhelmed with all the potential data.

Criteria:
-Using a BT30 Taper/Holder (50mm seems about right)
-Spinning at approximately 6000 Rpm (Aluminum cutting with 150ipm at max). Any higher and I feel that balance would be a huge issue.
-Drilling/Reaming operations on alu as well as milling.
-Over all combined accuracy across the machine (neglecting the spindle) is ideally .003+/-
-I am in Canada

Other data/Thought process

I (believe) I need 3 bearings in total. Two angular contacts and one "regular"


Questions
-What am I looking at for these bearings? Can I get away with These? rather than going for an actual spindle set or am I just cheaping out and setting myself up for failure later on down the line?

-Where should I be looking for these/Project materials in general? Ebay seems the obvious choice but in particular are there sites with decent customer service/etc I can frequent? I'd rather give them my money if I have the choice.

-I've clearly never taken on a project like this before. I can machine the parts no problem but I'm really ignorant on things like cooling the bearings, over all loads I will encounter and things of that nature. Any input would be great.

[rowbare]

Is the goal to build a spindle or to simply to have a spindle?

If the building process is important to you, there are several threads to be found here. For cutting metal, you should probably look at using two pairs of bearings. I am not sure that the double row bearings will be suitable. Certainly the ones you are linking to would not be, they are only rated to 5000 rpm with grease lubrication.

If the goal is to have a spindle, Tormach sells a BT30 spindle cartridge for a very reasonable price 30505 - BT30 Spindle Cartridge Assembly. If you look in their replacement parts section, you can see which bearings they use for their R8 spindle (not sure if they are the same for R8 and BT30).

Also I saw this post recently: http://www.cnczone.com/forums/1210182-post68.html
For what it costs it could be a good starting point.

good luck
bob

[Matt Ribble]

I lost this thread for a little bit but here goes.


I have decided I want a small prototyping machine to work in aluminum. The requirements mean that I have to drill, ream, mill, etc which excludes me from using the router heads/2.2kw spindles that are common (if these can ream/drill/function at low speeds then I am completely in the wrong)

I wanted to take a er32 collet chuck and build a spindle with it to be powered by a decent motor capable of doing all the necessary operations.


Parameters:
-Cutting aluminum only. Some wood is probably going to be cut as well. 150 IPM at .02 cut depths is probably my desired working parameter.

-Small scale, parts less than 10x10. Mostly platework.

-Speed of cut is less important than over all precision and repeatability

-Looking to have approximately 10k spindle speed. I'm currently looking at
These bearings.

-My understanding is that a 15 degree contact angle would be far superior to a 40 degree (thus the price difference) BUT with the innate innaccuracy of a home made machine I don't know if it will be enough to really break the camels back so to speak.


I appreciate all the input I can get. This whole project has been eyeopening about what it is I don't know that I thought I did.

[handlewanker]

Hi, It's a pretty standard bearing design using two back to back ang contacts and a deep row sealed radial ball race at the top.....been deep discussed on the forum in other threads.

You have to decide if you want to direct drive, have a belt drive to the motor or go to a motorised spindle.

You can also explore the possibility of buying a ready made and water cooled high speed motorized spindle on Ebay in the router categories....about 1.5 hp, 6,000 rpm $500 etc, whatever.

Making one will need expertise in the machining dept.
Ian.

[mactec54]

Matt Ribble

You can do it almost the same as this one, you will need a different drive pulley, poly-v belt is best, for the size you want to build ER32

The bearing you have the link to is not suitable for a good spindle, they have some backlash & you can not ajust them

http://www.cnczone.com/forums/cnc_wo...dle_build.html

[RICHARD ZASTROW]

You may want bone up on bearing specs. As mactech54 mentioned above, those bearings have radial clearance, C3 has more than normal clearance allowing "play" in the bearings.

After you're more familiar with the spec's, I think you will find you need some amount of preload.

How do you intend to lubricate them? Lotsa things to think about before you "freeze" your spindle design.

Dick Z

[RICHARD ZASTROW]

Matt, Try this for reference; Timkin Engineering Manual pdf

I typed in the www stuff but somehow it goes to a different address than I want to go to.

There are others as well.

Dick Z

[handlewanker]

Hmmm.....I think I'm losing the plot somewhere....we are talking about angular contact bearings, and as such angular contact bearings can be adjustable if the initial "fit" does not give you a running condition that makes them tight for exact spindle location and freedom from backlash.

By this I mean the bearings indicated are designed to be run back to back (both outer race cups facing inwards) and give a running fit for a certain condition......possibly not tight enough for a high speed spindle that requires a closer fit.

The simple solution in all cases of angular contact bearings is to add spacers of specific thicknesses to both the inner and outer races.....BTDT.

"Normally" you would fit these indicated bearings into a housing with the races making face contact as the manufacturers made them and the running conditions would be set as designed and having the races pressed together by either screwed collars on the shaft or caps on the housing.

It does not stretch the imagination too much to suggest that if you fit a spacer between the inner race faces and a spacer between the outer race faces, and the two spacers are the SAME thickness.....then the running condition as the manufacturers show will be met.....but if one spacer is .02mm thicker than the other, you will be adding a preload to the bearing set, and so this is how "adjustable" is achieved......a surface grinder is needed to ensure the spacers are parallel on the faces.

Preloading is a matter of running condition, and too much preload at high speed is a recipe for overheating and bearing failure.....it just depends on how much preload you determine either by calculation or trial and error to arrive at the ideal condition for the task.

Just as a matter of interest, some years ago I built a toolpost grinder for my ex employer and the angular contact bearings were "off the shelf", IE, not a matched pair, and were in the bearing configuration of two at the front and a pair of radial ball floating at the drive pulley end....standard configuration.

To anticipate a preload condition two spacers were fitted between the angular contacts, and to start off they were of equal thickness etc.

After a test run one spacer was reduced in thickness by .02mm to give a preloading and after five minutes of running the bearins were quite hot.

The conditions were...... grease lubed, 3,000 rpm, grind stone size 100mm diam.

Making the spacers back to equal thickness the bearings ran cool and the ground finish satisfactory with no indication of a loose bearing pack.

Angular contact bearings if bought as a matched pair with a coded preference will only work for the condition the manufacturer indicated.

It is a simple process to "adjust" the preload of a random angular contact pair if you make spacers that can be tailored for the conditions required.......the bearing pack temperature will give you that information in the first 5 minutes......the spacers MUST be made parallel.

You could of course make the front angular contact spring loaded to apply a constant but soft pressure to the bearing pack....(the back angular contact must be fixed and against the housing shoulder to maintain spindle position) and so reduce the risk of too much preload, but this is a bit more complicated and needs carefull making as there is really no indication of the amount of preloading you'll get or an easy way to measure it.

Lastly, running in a grease packed bearing with no room for circulation of the grease will lead to overheating.....note the requirements for car front wheel bearing lubrication as a guideline.

For those people who need to get the utmost life from angular contact bearings once they have been run for a considerable period and are still in good condition but have gone slightly "slack", or not quite as tight as they once were, then the differential spacer method will allow more life to be obtained without the need for a replacement set.

Differential spacers are much easier to apply than fitting a shim between the bearing faces, and you can get as little as .005mm differnce quite easily with a surface grinder.

IT IS HIGHLY INADVISABLE TO ADJUST THE ANGULAR CONTACT BEARING PRELOAD BY LAPPING THE FACES OF THE RACES......there is no going back.
Ian.

[mactec54]

handlewanker
Hmmm.....I think I'm losing the plot somewhere....we are talking about angular contact bearings, and as such angular contact bearings can be adjustable if the initial "fit" does not give you a running condition that makes them tight for exact spindle location and freedom from backlash.


Double Row Angular Contact Bearing, these you can not ajust, they have a fixed clearence, depending on the bearing grade, with the only possible way to ajust these bearings is by how tight you fit it to the housing or shaft, which only helps with radial clearance, there is no way to preload them, they usually have between .0005 & .001 backlash

[RICHARD ZASTROW]

IAN is correct on matching bearings which is why they make universal sets. they can be used in any position or orientation within the set.

Also, spacers increase rigidity and by separating the bearings, the heat generated by the bearings is spread apart. The spacers should be at least as wide as the bearings themselves. As IAN also said, each pair of spacers must be EXACTLY parallel or the bearings will bind and generate heat and wear quickly. We hold the spacer pairs parallel to within .002mm (.000079") At .00254mm (.0001") the bearings tightened up noticeably. These were a set of (4) 17mm x 28mm P4 A/C sealed bearings.

5200 series double row A/C bearings have clearance built in. There are preloadable double row A/C bearings with a split inner ring but why not use a pair of bearings instead?

Bearings should have only 15 to 20% of available volume filled with grease. Which is why we use greased at the manufacturer sealed precision bearings. Lubed for life and all that.

Dick Z

[mactec54]

RICHARD ZASTROW
Read the post ( 1 ) by Matt Ribble

Matt Ribble asked this question I gave him the answer, the double row a/c bearing he was looking at is no good for a spindle bearing

Questions
-What am I looking at for these bearings? Can I get away with These? rather than going for an actual spindle set or am I just cheaping out and setting myself up for failure later on down the line?

[RICHARD ZASTROW]

"or am I just cheaping out and setting myself up for failure on down the line?" to which I say yes.

As far as the double row 5200 series A/C bearings I agreed with mactech54, they are not really suitable for spindles.

Dick Z

[handlewanker]

Yes, post #1 did state double row ang contact......but further on in post #3 he asked about the single row ang contacts .......price was also mentioned somewhere.

There was some confusion in that the double row priced at $61 was 50X90X30.2, whereas the single row priced at $27 was 20X47X14.....I fail to see the comparison on price as both are significently different in size.

Anyway, the pair of singles is the way to go....sealing MUST be a prime consideration.
Ian.