[Guldberg]

Here is my idea of a spindle including ATC, trying too keep it simple. Anybody have something to add or comments, the design is mostly stolen from around the net, especially buildyouridea.com
What material to make the spindle and the toolholder of?, i have some stainless lying around, would that work?
Anybody have a good european source for high speed bearing?

[NC Cams]

The use of a single bearing at the cutter end as opposed to a preloaded duplexed angular contact bearing will result in poor spindle rigidity and a potential for cutter chatter and other machining problems.

You might want to look more closely for machine tool spindles to borrow design ideas from. This design might barely work for a tool post grinder but it is hardly suited for an ATC based cutting tool spindle.

Moreover, the use of position preloaded bearings causes problems as the parts heat up.

You WILL ultimately sacrifice performance potential for the sake of low cost and ease of fabrication with the proposed design. Not what one would recommend, ESPECIALLY for a "high speed" spindle.

Then again, you don't give a lot of specifics as to what you expect in the way of performance, however, I would not expect "machine tool grade" performance from a design like this.

Might want to do a "spindle" search on this website. There have been a number of ATC style spindles designed and discussed over the past 4-6 months. Why reinvent the wheel?????

[Guldberg]

Good point, i did search the forum, but the designs that people comes up with seems a bit complicated and most of them only ends up as cad drawings. I can see what you mean about the bearings, but what about the ATC part?, cant see any real problems there

About specifications, right now its going to replace a 20$ router, so better than that is progress:-) eventually i would like to cut some alu, but for now hardword and polymeres would be the goal

[fkaCarel]

But yours could as well end up as a CAD drawing. NC is right about the double bearing fixed at the front and sliding on the other end to avoid thermal difference problems. Personally I think the taper is on the wide side. Side forces will pull on the tool. The reducing of diameter for the pulley is a bit too compromising. The spacer under the pulley enlarges this problem. See these comments just as positive.

[Kipper]

Strange you should mention duplex angular contact bearings NCC I was looking in a trade catalogue at bearings and came across the angular contact ones. They appear to be used as the bearings for the very expensive ball screw support blocks.. Also available in high speed variety (50.000 RPM with oil and as I remember 30.000 with grease) rs components was the catalogue and I believe it's available on the net. Nice drawing Guldberg :thumbs: I understand how the ATC thingamabobs work now (never seen one in the flesh)

[Guldberg]

You mean putting something like this at the cutter end?

[fkaCarel]

These are angle contact bearings. There are also spindle bearings, look almost the same, lower loads, higher rev's. Be forewarned about questions about bearing fits and look what they will cost you. The price may make it another CAD drawing.

[Mcgyver]

These are angle contact bearings. There are also spindle bearings, look almost the same, lower loads, higher rev's. .

spindle bearings are whatever combo you use to make a spindle; angular contact bearings, tapered roller bearings plain and thrust bearing, or even babbit. There are different classes of each and the designer mixes and matches to his/her objective. certain combo's like angular contact and plain being more appropriate than others for certain types of spindles, but there is no separate type of bearing called 'spindle bearings', they are just angular contact bearings used in a particular spindle design afaik

[tivoidethuong]

you can use bearing no 72xx

[handlewanker]

Hi all, I'm rebuilding a small vertical mill spindle and will be using this two ang. contact back to back and floating radial bearing layout. There will be a labyrinth seal on the bottom. This was featured in an SKF bearing catalogue, 1985, and was recommended for grinders and mill spindles.
The speeds will be as per conventional mill layout (no CNC). The spindle is being turned from Nickel chrome steel and the taper, R8, and spindle nose will be case hardened and ground.
The problem with any high speed home brewed spindle project is the seals.
Also you can't just use any old angular contact bearing. High speed requires special cages or they'll just disintergrate.
Previous to this project a toolpost grinder was made, for 100mm diam. alluminium oxide stones, using the same bearing layout and running at 6,000 rpm. and lubed with high melting point axle grease. To date it has performed well.
Ian.

[Guldberg]

Most of the bearings i found, including skf 72xx has been rated at 15000 or lower. I'll like to go all the way up to 20000, maybe thats a bit optimistic. Does anybody know a good online source for these kind?

Kipper, do you have an rs number of those bearings?

[fkaCarel]

but there is no separate type of bearing called 'spindle bearings', they are just angular contact bearings used in a particular spindle design afaik

Angular contact bearings:
http://www.fag.com/content.fag.de/en...ow_angular.jsp
Spindle bearings:http://www.fag.com/content.fag.de/en...e_bearings.jsp
If the manufacturers split them in different category's, I will also do it. Note the differences in design.

[NC Cams]

Bearings simply don't "disintegrate" at ultra high speeds (as in the limit speed of the design) but they do have issues that you need to protect for.

Most commonly, ball skidding and/or the centrifugal loading of the balls on the raceway becomes a problem. "Spindle bearings" are not appreciably different from regular A/C's - asside from running accuracies and tight tolerances and preloads and how they're matched/selected. This is why most bearing suppliers recommend that you seek the aid of their engineering depts if you're going to build a true "HIGH SPEED" spindle

Moreover, most machine tool/spindle grade bearings are ABEC 5 or 7 in accuaracy and anymore have plastic cages. Plastic cages tend to have less problems and stabilize the balls better at high speeds than steel or brass caged bearings or machined bronze ones for that matter. Thus, the prime difference of concern in this case between a regular A/C and a machine tool grade one involves ABEC toleraces (as noted above) and preload.

Preload (either Low, Med or High or some variation of that "quantification") foces the balls into the raceway to make running accuracy much higher. It also provides more consistant ball loading so that balls don't roll into and out of the load zone - such uneven loading will cause skidding and drastically shorter lift.

The HIGHER the speed of the spindle, the LOWER the preload and vice versa.

The HIGHER the axial loading potential, the more contact angle you run and vice versa.

The preload and contact angles DO vary according to whether you're making a grinding spindle, radial cutting spindle, lathe spindle or combo radial/drilling spindle.

Grinding spindles have lighter preloads due to higher relative spindle speeds and lower relative loads.

Lathes tend to have higher preloads due to lower relative speeds and higher radial and axial thrust potentials.

Mills tend to run somewheres in the middle as they are more universal in the loading that they might see.

Although the bearings PER SE are essentially non-application specific, they tend to be selectively fit/matched (contact angle and preload wise) so as to form a "lath spindle bearing" or a "grinding spindle bearing" or a "milling spindle bearing".

Sort of like making pizza - same ingredients but different permutations to satisfy a particular "taste" or useage requirement in this case.

When it comes to spindle speeds, you really need to look at what you're going to machine and the speeds and feeds that the material wants. What 2024 aluminum wants is NOT what M2 or M4 tool steel wants.

Thus, to make a 20,000rpm spindle and then run it at a small fraction of that speed 90% of the time due to the predominance of material that you'll be cutting is not the way to achieve optimum performance from your spindle and or your cutters or bearings.

You really need to develop very realistic expectations with regard to your useage and the resultant design requirements of/for the spindle.

[handlewanker]

Hi NC, what you say is so very true. We had a Beaver CNC mill, in 1975, and when it was used for those high speed drilling and milling ops, a supplementary spindle was used. I'm not sure if the spindle was mounted to the side or just mounted in position as needed, but when it was working we all had to wear ear protectors, the type that cover your ears as used on airport runways, and that was 40 metres away. They used to run this for days at a time.
Ian.

[Guldberg]

I never intend to cut any other metal than soft alu. My thought is that if my 20$ router can run 30k rpm, i can make a spindle that run, lets say 15k rpm cutting wood for maybe 100$. This is for purely amateur work, maybe thats the different. I am fully aware that there is no such thing as a cheap highspeed spindle for doing "serious" work:-)

[Kipper]

online 23.500 RPM (on oil) 12mm ID I'm not at work bud but heres one I found...hth

[NC Cams]

"...My thought is that if my 20$ router can run 30k rpm, i can make a spindle that run, lets say 15k rpm cutting wood for maybe 100$..."

Do you have any idea of the engineering, development and testing that went into designing and developing a $20 router motor that would work/live at 30K in a retail/DIY market???

Having worked with in an engineering department with the engineers who did just that, I can assure you that there is MUCH more to it than "paraprhasing and plagiarizing" spindle design ideas from fotos on the internet. The recreation/restatement of the engineering involved to do this type of analysis/development is beyond that which can be conveyed on a message board.

The bearings in 30K spindles (or any high speed spindle) are the result of a very careful optimization process that can take a year or more. This includes cage development, grease testing (very much testing to find what and how much works), preloads (absolutely critical as speeds and housing/shaft temps increase), support housing issues like you can't believe and hour upon hour of testing.

This does not count the load projections and analysis work done to assure that the bearings have the load capacities for the intended cutter forces which most DIY'ers never even look at/for. Most people SWAG it by looks and/or availability and then wonder why the bearings won't live.

Even with all "professional effort" that can be spent, first guess bearings almost NEVER sailed thru testing and made it thru into production. Essentially that's what you'd be hoping for when/if you bought off the shelf parts as were originally illustrated in your design.

Would they fit? Yes. Work? for a while. Live? probably not.

Quite frankly, the cheapest way I know of to come up with a 15K spindle to cut wood would be to find a $100 commercial router motor (IE: Craftsman, Dewalt, Bosch, etc) and then modify/adapt it to your application. Another albeit more costly option would be to look into buying a "cartridge spindle" which is essentially what you're trying to recreate.

Yes, this is DIY and I'm often accused of turning a bottle rocket into the space shuttle. BUT, one really needs to recognize that there is some serious engineering and technical know how involved in the creation and development of spindles. This work is more detailed and difficult if you want them to function properly, with an ATC no less, and cut metal or wood smoothly and accurately for an appreciable period of time.

To try to oversimplify the effort needed to make it properly for the sake of giving someone false hopes of being able duplicate it on the cheap is simply wrong in my opinion - especially in light of my prior professional experiences in the bearing industry.

If you use a pre-made router motor, you wouldn't have the "satisfaction" of having made it yourself. BUT you'd be able to take FULL advantage of the high speed development that the folks who made your motor have already done - work that is well beyond the time and budget of even the most avid/intense DIY'er. Besides, why reinvent the wheel???

I wish you well with your endeavors and my comments are purely of good intent. I'd simply rather you be peeved at me for providing an "overcomplicated reply" as opposed to giving you an "over-simplified-feel good reply" that ultimately doesn't work to your satisfaction. You really need to know your challenges before you conciously decide to get involve with some projects.

[NC Cams]

BTW, the idea of creating a new, DIY router spindle has been addressed previously. You might want to check this post for alternated ideas:

http://www.cnczone.com/forums/showthread.php?t=10695

[Guldberg]

I just read read the entire post refered to in #18. I kind of come to the conclusion based on the post and your statement that maybe i could go for, lets say maybe max 10000rpm, but with some torque opposed to most routers at this speeds. I fully understand your point NC cams, i am (almost) an engineer myself, so i know what kind of work that goes into making a product:-) But still im convinced that its possible to make some sort of spindle that can suit my need, maybe it wont live more than 100 hours before i have to change the bearings, but then again if i only paid 30$ for the bearings i had fun building it and 100hours of work will proberly last many month for me. My point being, im not trying to make a professionel product, mainly just something less noisy than my cheep router and with ATC capability.

Actually im planning to making this a special course at my university, this way i have litterally all kinds of machinery and knowledge available. Doing some modelling, force and FEM calculations etc. That should be an easy 5 ects credits:-) Lets see where it goes...

[handlewanker]

Hi Guldburg, I read your endeavours to get some ideas together for a DIY spindle. How about a spindle running at 10K and using a 3 tooth cutter for roughing out the profile of the job with feed rate to suit?
Then finishing the job with a bur type cutter (10 tooth?) still running at 10K which will effectively put more cuts on the part without screaming around with a few teeth.
Never done any routing , but the process is self explanatory. I don't suppose your going into any serious production so the method used will be a compromise in that you just want to shift a bit of wood, alluminium, plastic or whatever takes your fancy, and just get there, even if it takes a bit longer.
As far as changing the bearings if it conks out, well if you're into a job and the damm thing seizes up on you, it'll mean a total strip down and rebuild, usually at the most inconveniant time.
I would give the design a bit of carefull thought, and don't go beyond your capability.
Ian.