I will motorize the knee of my CNC mill soon

Since I got the 4th axis working, I am going to finish the mill's CNC conversion. I am 100% convinced that the knee needs to be motorized,
so that I can change tools and alternate between short tools (small end mill) and long tools (long drill bit in a chuck), while retaining all offsets and relative Z positions.
Otherwise I cannot really do a lot of jobs nicely.
With the knee motorized, I can machine more or less anything that fits into the mill's envelope.
The axis with motorized knee is called W. It is parallel to Z, but has the opposite sign, which actually makes perfect sense if you think about it.
Chris Radek from EMC mailing list gave me a great suggestion, which is to store W offsets in the tool table. This way, when I change a tool, the knee automatically moves to accommodate the tool height.
The best picture of the knee of my mill is from a while ago:
http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/34-Adding-Servo-Control-to-Knee/
The page is kind of empty, but I will populate it as I go on.
I already have a servo capable precision DC gearmotor ($45 on ebay), with a shaft for encoder sticking out of the back. This gearmotor goes up to 4k RPM and is reduced 15:1. So, at 3k RPM it would give me 200 RPM of the knee crank, which is all I care to have, it would be about 60 IPM speed if I got my math right. I can probably limit it even further and will still be great.
I bought an encoder from US Digital, hopefully I will get it this week.
Additionally, I have a dual limit/home switch that I also need to install on the knee. I am pretty sure that it will take me a month at least to get done, maybe more, as I barely have any time.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Oh, and the knee has a pneumatic assist, so cranking the table does not require enormous torque.
i

http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/34-Adding-Servo-Control-to-Knee/
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Are you going to use a ball screw on the knee?
If not, then adding the position to the Tool table will be moot as accurate positioning will vary greatly.
With a ball screw it will also need an electric brake to prevent auto lowering.
I went through this all about 6 years back and sold my knee mill and bought a bed mill with 20" Z ball screw. Solved the envelope problem.
Dave
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Dave, if you code the process so it always lifts toward the target height, use a linear encoder for absolute positioning (not the shaft encoder on the motor, no), and add the brake, you can get precision even with a conventional lead screw.
If not, DROs would be useless.
Backlash in this case is not an issue, because he won't be articulating the knee during milling operations, only before or after a tool change; then locking it for the duration of that tool's residence in the machine.
LLoyd
LLoyd
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

no, I wil use the original ACME screw.

Why is that so?

That does sound nice.
But, I think that this screw is a precision screw, the dial on the crank is graduated in thousandths, this is probably not crap. The backlash does not matter due to weight of the knee. Why is accuracy going to be that poor?
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Tue, 15 Feb 2011 11:17:13 -0600, Ignoramus20463

probably depends on your needs. If you're a mold maker this would never do. If you're a fabricator this is plenty good. I fall in the second category. For those few jobs that need extreme accuracy you can always tweak your offsets to what the machine is doing.
This has been on my real nice to have list for some time. You're way ahead of me as i don't have the gear motor. I did get the air cylinders, but they aren't installed. I'd also use W axis for offsets.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I am more of a fabricator, too. I just make stuff for my own use and usually it does not involve extreme precision.
And like you said, this motorized knee essentially is an optional feature.

I have very high hopes for this. The motor is a DC motor with a precision gearbox. It was $41.
http://www.soelma.it/prod_mo_tot.html
(with a gearbox that is not shown there)
I think that after this motorization, the mill will be a total kick ass machine. I could change tools and not lose position, plus, I do not have to crank that table handle all the time.
I would use W for offsets only, not to jack it up and down for actual machining operations, to save the life of the screw.
I will have questions, quite soon, on how to mount this gearmotor, but first I need to take some pictures.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Were I you, I'd put a clutch of some sort in the system, and make the rig so you can hand-crank it also. If you need to fiddle a little with offsets or depths of cut, that would be a boon.
Just remember to remove the handle (or put a cut-out switch so the handle in place disables the drive), or you could be facing a 200rpm billy club made of metal!
LLoyd
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 2011-02-15, Lloyd E. Sponenburgh <lloydspinsidemindspring.com> wrote:

Lloyd,
I have not, yet, fully decided on anything.
But I think that what I will do is, I will completely remove the handle and put on the motor permanently. I do not need handles on X, Y, Z, A, so why would I need one on W?
If I want to fiddle a little with X, Y, Z, A, I use the keyboard or the joypad.
And if I want to fiddle with W, I will use the same thing.
I agree 100% about the handle on a motorized knee being a club, whacking my private parts wildly, but also, it always gets in my way anyway.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Feb 15, 2:13pm, Ignoramus20463 <ignoramus20...@NOSPAM. 20463.invalid> wrote:

Iggy, Are you bragging that your private parts get in your way, anyway?
I do agree that you should not need a handle with a jog control.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Feb 15, 10:17am, Ignoramus20463 <ignoramus20...@NOSPAM. 20463.invalid> wrote:> Why is accuracy going to be that poor?
Well, there are so many things involved in lifting that amount of weight, especially if it is air-assisted. In fact the air-assist would probably add to the inaccuracy as the load on the thrust bearings and dovetails will vary. The knee will sag differently every time.
I would suggest before removing the knee-handle and assuming accuracy, you attach a dial indicator to the spindle then zero both the indicator and the knee-handle-dial. Lower the knee down a good way, 5" or so then come back up to the knee-handle-dial zero and read the indicator. Do that a few times over a couple of days and report back. Then maybe consider a ball-screw.
Dave
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I think if you lower it down below the desired level and then raise it back up it would be more repeatable than just running it from current location to desired height. I think the addition of a counter weight, air assist or, gas spring would make that harder to achieve. If I were going to motorize the knee I think I would consider going to gear reduction to get the torque to raise the knee rather than a gas spring or other assist. Since I am working on an old Hurco I've been thinking about this issues as well. I don't think the slower movement of using gear reduction would be a huge issue since its not being moved constantly, but rather once to position properly for a tool change.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I don't agree. If you took all (or almost all) of the load off the knee, it might be inaccurate, but if you left some healthy positive weight on it, and the gibs and ways are smooth, you should have no problem whatsoever repeating a lift.
I do agree that dropping below the desired height, then always raising to the target will give the greatest degree of accuracy. But that's just software, if he does it in the W axis under EMC^2.
LLoyd
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 2011-02-20, Lloyd E. Sponenburgh <lloydspinsidemindspring.com> wrote:

Exactly. I can always write a subroutine lower_the_knee that would take the intended value of position W, drop rapidly to W+0.1, and then go up by 0.1 inch at some slower feedrate. That takes out all backlash and is not too hard to use.
I will not use the W axis as a machining axis, at least unless I need to drill holes over 5 inches deep. I will just use it to accommodate larger and smaller tools occasionally.
Just yesterday, when I was making this dog clutch, I had to move the table up and down a lot (edge finder/small endmill/large drill/etc), and really wished that I already had the W axis working!
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I am not sure if I fully understand "the issue".
The air assist is a pneumatic cylinder that maintains costant pressure in it. It works the same as a big counterweight (without the inertia added by the counterweight).
The amount of upwards force is clearly much less than the weight of the table (with the rotary table on top of it).
What the air assist really accomplishes is two things. First, it makes it easier to move the table. Second, it reduces wear on the knee screw due to reduced pressure.
Since the table leans on the screw at any time, there is no backlash due to screw. The only backlash here is in servo motor coupling and the 90 degree or whatsver drive that turns the screw nut.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

http://igor.chudov.com/projects/Bridgeport-Series-II-Interact-2-CNC-Mill/34-Adding-Servo-Control-to-Knee/
That sounds like a nice feature. Since the quill and knee both operate on the Z axis, is there a way to optimize for minimal knee travel? I guess you're going to do it manually by programming a W axis but I was just wondering if there was a way to do this automatically, positioning the knee to the right height where it moves the minimum amount during a part run.
OK, have you got plans for tilting the head yet? Or is the automatic tool changer next? :-)
That's what I love about EMC2, these kinds of upgrades would be expensive on most other controls I know of.
RogerN
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

The way I think about it, which is slightly too simplistic, is that I would use W for tool changes only, which will not amount to a lot.
Use of W for actual machining, is necessary only for drilling holes over 5 inch deep, which is pretty rare.

Definitely not.

Me too. For all other controls, all the above would be expensive add-on features.
i
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Ignoramus20463 wrote:

On some controls the w can be added to the Z axis to get a tool position readout the is the distance of the tool tip to the part. fanuc charges 3000 bucks for that option I believe.
John
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.