Adding a high speed milling sub spindle

[precisionworks]

In order to get decent surface speed from a 1/16" end mill in aluminum the rpm needs to be in the 40k-60k range. This mod will use a 50k Air Turbine Tools 202SV: http://www.airturbinetools.com/ht/ss/202SV.html

When done the mount will resemble this photo from CNCZone:

First job is to cut a blank from a piece of aluminum plate that's 2.250" thick & 6" wide:

The Delta band saw does a nice job cutting an arc to match the radius of the faceplate.

Positioning blank on the faceplate. Note that the center line has been marked. The point of a live center will be pressed somewhere along that line & the live center will hold the blank to the faceplate while the clamps are set up:

Making sure that the edge hits nothing when it spins. The point of the live center is just visible in the upper right corner of the photo. The blank has about 1mm clearance which is plenty.

Using a spade drill to make a 2" hole from the solid. Even at only 115 rpm the spade cut fast and punched through the back side quickly.

Clearance hole for spindle is done except for a light boring pass to make it look nicer. Will make the counterbore tomorrow.

 

[gadget_lover]

That's a 1/2 HP tool to spin a 1/16 inch drill. Do you really need that much? Why?

Daniel

 

[Th232]

Hmm... am also interested in knowing why, but my money is on one of two options:

For a given power, high speed -> low torque, so at 40-60 thousand RPM there's just not that much torque available if a smaller tool were used...

...or it's just what he has on hand/found for a good price.

Cue Barry giving a third option which I haven't considered at all.

 

[precisionworks]

For a given power, high speed -> low torque, so at 40-60 thousand RPM there's just not that much torque available if a smaller tool were used...

+1

There are quite a few dedicated sub spindles available & some are even set up for CNC machines with an automatic tool changer. The dedicated spindles cost from $2k-$5k so I wanted to find a hand tool with a straight forward section to make mounting easier. In the ATT line that means either the 201SV (.2hp for $370) or the 202SV (.5hp for $375) so the higher hp unit was selected. Early on I'd considered using my 200SV (.3hp @ 50k rpm) but the shape does not lend itself to sub spindle use.

The other option that initially looked promising was to use my Bosch Colt trim router but there were issues with that approach, primarily the runout of the factory supplied collet. That can be addressed with an aftermarket collet & nut for $75 & is not a bad choice. The size is huge compared to the ATT spindle & it looked like there would be no way to get a 1/16" - 1/8" tool anywhere close to a flashlight body fixtured in the Super Spacer chuck. For flat machining like O-ring grooves this would work well.

1/2 HP tool to spin a 1/16 inch drill. Do you really need that much?

LOL ... that always reminds me of Tim "The Tool Man" Taylor & his desire for "more power" :devil:

The 1/8" collet allows using any 1/8" shank tool so end mills from .015" to .125" can be used. The greater hp of this tool is much more than needed for a 1/16" trit slots but it will be useful for more traditional jobs like milling long slots where the tool can be fed anywhere from 50 ipm to 150 ipm. With the cost of this tool only $5 more than the .2hp tool the choice was pretty easy.

 

[precisionworks]

The 2" hole drilled yesterday is a clearance hole for the mill spindle & now the counter bore is being cut to slip over the end of the quill:

The ID was brought to .0015" over size so it would be a pretty easy slip fit:

Surely is nice to see a stripped clean face plate :)

Set up to drill the tap hole for 3/4-10 thread. Drill is a "long boy" that was picked up in a set on eBay for next to nothing, about $30 for eight drills IIRC:

Sometimes it's a life saver to have a pulley tap. The tap passes through the 3" bolt hole before threading the bottom 3" of the block:

The block locked down tightly with half a turn of the allen head cap screw:

Just need to bore a guide bushing & bore for the air spindle & it's done.

 

[ICUDoc]

I learn a lot from your photo threads- thanks mate.
Looking forward to the next update

 

[wquiles]

+1

Nice project Barry!

 

[StrikerDown]

Always interested in your projects, this is no exception!

Cutting the counter bore probably took more than a couple passes!
Question: How do you make the bottom of the counter bore nice and square and smooth ?
Do you stop a little short on each pass then make a cleanup cut from the center out to clean up the bottom?

 

[precisionworks]

Cutting the counter bore probably took more than a couple passes!

That it did ... faceplates depend on the little bit of pressure applied by a few clamps to hold the work in place & I managed to bump the part over by taking too aggressive a cut at the start. Brought the pipe center back up to the part to hold it in position & re-tightened the clamps, then started taking only .100" radial DOC with a slow feed. To go from 2.000" to 3.750" took nine passes at glacial speed with the spindle turning 115 rpm - plenty fast enough for me.

How do you make the bottom of the counter bore nice and square and smooth ?
Do you stop a little short on each pass then make a cleanup cut from the center out to clean up the bottom?

Exactly what was done. Each pass was stopped at 1.495" depth and the very last pass from the center out was at 1.500". This would have been an easy job on the mill with a Narex VHU or a Wolhaupter facing head but I don't yet have one.

----------------------------------------------------------------------------------

Different positions were tried until one was found that looked the best & the part was marked out for trimming. The band saw removed most of the waste & the Burr King cleaned up the cut:

The black circle shows the approximate location of the spindle. A Delrin bushing was turned & bored to hold the spindle, then split on the table saw:

The mounting plate was set up on the mill & a 1" starter hole was drilled. Then the boring head was used to open the hole for a size-on-size fit with the Delrin sleeve (1.258"):

Because the air spindle is so small it isn't in the way of the spindle lock:

Still have to install a 12" piece of 5/8" diameter Thomson shafting & a pillow block linear ball bearing to prevent side to side quill motion. The the Mitu vertical DRO goes on and it's all done ...

 

[precisionworks]

The mill mod was scheduled for last Friday but an all day portable welding job came up. Saturday was spent installing a 3ph meter loop & finally got back on the mill today. First located the position for the hardened linear shafting & bored a press fit hole, then secured it even more with three set screws. Next located the position for the linear ball bearing pillow block & tapped four holes:

The rod had to center in the hole where the depth stop rod used to go:

Installed the pillow block after fitting it for height. Pillow block height has to be 0-0 & that means either milling or grinding the back for the correct fit. Left-right and front-back can be out by as much as 1° in any direction as the cartridge floats within the shell.

All done & smooth as silk. There is zero side to side quill movement, which is the whole purpose of the linear shaft & bearing.

 

[wquiles]

WOW - very impressive!

 

[precisionworks]

The last step for the mill mod is adding the quill DRO. Started by locating the best position.

Cut a piece of 1/2" aluminum plate to 1 1/8" wide by 9" long so that it just fits the head casting.

There were already M8 tapped holes in the head but M8 screws are tough to find when you live in the boonies so I retapped two of them for 1/4" x 20 tpi. The front of the casting is not flat & a washer was placed behind the upper screw so the plate would not bend.

The DRO bracket was positioned & the center of each hole marked with a Sharpie marker.

#10x24 tpi holes were tapped to hold the scale.

All that's needed to finish this is a strip of sheet brass 11" long by about half an inch wide ... but there's none in the shop. Just called McMaster & ordered a 12"x12" sheet for $50 delivered (.062" thickness). A number of sources were checked and that's about the going price everywhere. It will be here tomorrow afternoon & this will be finished :)

 

[StrikerDown]

I have noticed the quill on my RF-31 rotate slightly on occasion. It seems there is a small amount of slop between the quill and it's indexing pin. It' isn't a problem with the quill being coaxial to the spindle, but I was wondering how you were going to deal with that... Looks like a good solution for the light milling loads of the offset sub spindle.

I mounted a similar DRO in the same spot too! Mine isn't a Mitu though! :mad:

Great work there Barry, I is always a pleasure to see how you do things.

PS: I love the DRO... my RF-31 is so bad about positioning the Z axis, lock the spindle and go back to find the dial is not even close to where it was left! Now it doesn't matter because I know exactly where the spindle is.

 

[precisionworks]

Thanks Ray :thumbsup:

My normal milling sequence is to lower the tool, lock the quill, make a milling pass, unlock the quill, lower the tool, etc. Because there's nothing but a dog point set screw riding in a milled slot I have zero confidence that the quill will not rotate as it is lowered. Since the tool center is 6" offset from the spindle center there was a huge potential for error. On a knee mill there's no problem since the quill stays locked and the knee is raised to deepen the cut but this is not an option on a mill-drill. IMO the linear shaft & linear bearing is an elegant solution to a tough problem.

The DRO is a welcome addition, especially when moving the tool down only .006" per pass in Ti 6-4. I will not miss the dial at all.

Mine isn't a Mitu though!

$110 delivered from a member on the PM forum. Brand new old stock, about $250 retail.

 

[StrikerDown]

Dog point set screw is a better name for it than I used!
On mine the quill can rotate enough that with a 6" arm the shift could be as much as an half inch. There must be a way to tighten it up, maybe a larger dog point would help, but only so much, too tight and it would drag.

The DRO:

 

[precisionworks]

Your photo reminds me of something else I did to mine which was to put a sharper bend in the quill locking handle. It bends easily when heated red hot & the "straighter" handle is a lot easier to operate. Then it's a matter of getting the handle to unlock in the 11:00 position & lock up around 2:00-3:00.

 

[precisionworks]

Used the new SST (Super Spacer Transporter) to move the SS off the bench. The SST is a plywood box with corner reinforcements & a tempered Masonite top that snugly fits on top of my hydraulic lift table.

Because the wheels are located at the very end of the table the SS can slid on or off without tipping the table.

Raised up to mill table height:

Attached a connector bar to the DRO head:

Made a spacer from linen Micarta & shaped the back to fit the curve of the base:

 

[Philabuster]

+1

There are quite a few dedicated sub spindles available & some are even set up for CNC machines with an automatic tool changer. The dedicated spindles cost from $2k-$5k so I wanted to find a hand tool with a straight forward section to make mounting easier. In the ATT line that means either the 201SV (.2hp for $370) or the 202SV (.5hp for $375) so the higher hp unit was selected. Early on I'd considered using my 200SV (.3hp @ 50k rpm) but the shape does not lend itself to sub spindle use.

I applaud you efforts, but these hand finish motors are really not designed to be used as a spindle. The spindle models you saw have much heavier internals (bearings and center shaft) than the hand finish motors and they are priced higher as a result of the components.

Our work tried to automate a hand finish operation by placing the same air turbine motors the hand finish guys were using onto a Fanuc robot arm. What a disaster. They did not listen and would not buy a proper spindle. Ultimately, the project was a total failure 2 years later and many dollars spent. :shakehead There is a big difference in side bearing loads when used as a spindle vs held by hand.

Have you disassembled an old air turbine yet? The body is not much thicker than my Quark Mini AA2 and the bearings look like they came out from an old RC car. The bearings in a good quality vane motor are 10X times as sturdy and while not able to spin 50K RPM, would be much more rigid in this application.

One major drawback is the CFM required to spin an air motor. The air compressor consumes much more electricity than mounting a small electric sub spindle. Some people have found a cheap $100 Porter Cable or Bosch trim router (35K RPM) works great in this application:

http://www.amazon.com/dp/B0000222Z6/?tag=cpf0b6-20

http://www.amazon.com/dp/B000ANQHTA/?tag=cpf0b6-20

 

[precisionworks]

these hand finish motors are really not designed to be used as a spindle.

Most hand grinders are not designed for this application but this one is :nana:

ATT makes the 201SV that I purchased with a knurled barrel for secure grip when hand holding the tool. They also make the almost identical 202HD:

Their description of the 202HD says: 50,000 rpm, 0.50 HP, 1/8" Motor/Fixture Mount. The Model 202HD is a long steel barrel type fixture mount motor with constant high speed and 1/8" collet capacities is available with several optional collets for smaller applications. The 202HD has the advantage of a stainless steel barrel designed for accurate fixture mounted applications and many clamping applications for robotic metal finishing, routing plastic, etc. Constant high speed results in greater accuracy.

The difference in construction is that the 202HD has a smooth stainless steel barrel versus the knurled aluminum barrel of the 201SV. Barrel OD & ID are identical & bearings are identical according to the manufacturer. The smooth barrel would be nice as it would more easily slide into & out of the Delrin spit bushing, but the additional $300 for this model was impossible to justify.

Some people have found a cheap $100 Porter Cable or Bosch trim router (35K RPM) works great in this application

From post #4 in this thread: The other option that initially looked promising was to use my Bosch Colt trim router but there were issues with that approach, primarily the runout of the factory supplied collet. That can be addressed with an aftermarket collet & nut for $75 & is not a bad choice. The size is huge compared to the ATT spindle & it looked like there would be no way to get a 1/16" - 1/8" tool anywhere close to a flashlight body fixtured in the Super Spacer chuck. For flat machining like O-ring grooves this would work well.

There is a big difference in side bearing loads when used as a spindle vs held by hand.

With the same axial DOC and the same IPM feed the radial bearing loads are identical. If anything the bearing loading is less severe in a fixtured spindle whose DOC & IPM are controlled by a hand turned feed wheel. One reason for choosing the Air Turbine brand is my past experience with their smaller 200SV which has endured over 20 years of hard use.

One major drawback is the CFM required to spin an air motor.

+1

Part of my business is rebuilding high end air tools and air motors. Most of the ones I see are $1000-$5000 replacement cost & often need a large air flow for full power. The three vertical compressors in my shop will supply over 30 SCFM so running the small air spindle is no problem.

 

[precisionworks]

The air compressor consumes much more electricity than mounting a small electric sub spindle.

That it does but small electric motors come with their own baggage. Electric die grinders & routers use a motor design called "universal AC/DC" and that's exactly the same motor style used in a shop vac or kitchen mixer. Their greatest attribute is that they're small & cheap to make so they are perfect for a trim router with a $100 price point.

On the negative side ...

A router spins at 25k-35k rpm & uses less than precision bearings. Spindle TIR often runs .003"-.005" but can sometimes be reduced by adding a precision collet & matching nut & remachining the collet bore. The TIR on my 201SV measures only .0003"-.0004".

The noise level is very high, about 90-95 dBa at three feet. The ATT spindle produces only 67 dBa.

Even a small trim router is much larger than the 201SV/202HD which makes mounting more challenging.

Spindle projection from a router is very short & most of that would be taken up by the mounting base. This makes it nearly impossible to position the end mill over a light body held in the 3-jaw chuck of the Super Spacer.