Powered Knee tool offset?

[Jet A]

Hey,

I am pondering an idea that would make my life a lot easier....i think.

I currently have a series 1 converted to a cnc. The problem is (as i am sure some of you have found) without a lot of travel in the z axis, its almost completely ineffective to do tool offsets, as the knee must be moved in order to accommodate a much larger tool. example, switching out from a 1/2" end mill in a holder to a 3/8" drill bit in a jacobs. the difference is nearly 4" giving me only a 1" max drill depth, not counting travel room. Anyway, back to my idea. If one were drive the knee with another stepper motor (i am sure there is a way somehow) all standard z axis motion would take place in the quill, and the tool offsets could be done on the knee.

I am not sure if it is a great idea to drive the knee, but seeing as you would not be doing this all the time, its no more than doing it by hand. I would probably lighten up the knee with some gas shocks.

I am interested if anybody sees any potential problems with this. I am very new to cnc. will the knee have enough precision to justify this? do i need to consider a pneumatic or other locking device? am i being a big baby? I am just trying to reduce errors and time by resetting z's every time i have to move the knee.

[HimyKabibble]

I not only don't see a problem with it, that's exactly how I've been running my machine for some time...

Regards,
Ray L.

[79TigerPilot]

That is what I plan on doing after seeing Ray's setup. I put a stepper on my knee. Used an 1810 in-oz size 34 stepper and a 3:1 pulley that drived the Z lead screw directly (not the crank handel shaft). The belt exits the right side of the knee through a 3" x1" window cut in the casting. The stepper is mounted to the side of the knee. I added two 200# gas springs in the center directly behind the lead screw to take off some of the dead weight. I get between 20 and 25 IPM rapids but above this the stepper lockes up as it looses torque with RPM. The net ratio is 3:1 for the pulley and 5:1 on the lead screw for a total of 15:1. The pulley ratio is limited by the width of the window, I didn't want to cut through any ribs in the knee. I think a servo with some gear reduciton would be the way to go as you don't loose torque with RPM.

Craig

[Jet A]

Interesting comments. Ray, where can i find out more about your set up?

I am not interested in super fast rapids, just a little more accuracy and less labor intensive than spinning that stupid crank! any downside to driving the crank?

That seems like a pretty massive stepper, i suppose there is quite a bit of weight there.

Either of you have an pictures>?

[loose1]

YouTube- bridgeport cnc knee z axis

[HimyKabibble]

Originally Posted by Jet A Interesting comments. Ray, where can i find out more about your set up? I am not interested in super fast rapids, just a little more accuracy and less labor intensive than spinning that stupid crank! any downside to driving the crank? That seems like a pretty massive stepper, i suppose there is quite a bit of weight there. Either of you have an pictures>?

My machine is a Taiwanese BP clone. I did a full servo conversion on it - X, Y, quill and knee. I drive the knee through the crank shaft, with a 4.8:1 XL belt reducer, and run it at 50 IPM (vs 300+IPM for X,Y and quill), with a total reduction of 48:1, or 96,000 steps/inch.

I wrote Mach3 macros that let me mount all the tools for a job in sequence, and the macros do touch-offs to measure the tool length, calculate the offset, and load the offset into the proper slot of the tool table. I use custom macros (M843 and M849) in place of G43 and G49 to enable/disable tool length compensation. The compensation is done with the knee, with all moves being done so the final move is upward, so any backlash in the knee is irrelevent. When tool length compensation is turned off, the knee is driven to its lowest position, as determined by the current "reference" tool, for which I use tool #253. My "reference tool" is an endmill holder with a piece of drill rod mounted in it, that is slightly longer than my longest tool. This tool is always probed first, when setting up the tool table. This is necessary to ensure that turning tool length comp on and off can never drive the tool into the table or workpiece.

A few (old) pictures of my machine are below. The quill drive now has covers over the belt and ballscrew (this one was taken the day I built it, so it was not complete). The servos on all axes are 850 oz-in peak, 72V, 8.5A, 4200 RPM. X and Y drive 5-pitch screws through 4:1 GT2 belt reducers. The quill has a 5-pitch screw, and, IIRC, a 2:1 GT2 belt reducer. All run on a pair of 68V/15A continuous supplies, using Gecko G320s or G320Xs, and CUI capacitive encoders set for 2000PPR, and a SmoothStepper.

Regards,
Ray L.

[BobWarfield]

Originally Posted by HimyKabibble All run on a pair of 68V/15A continuous supplies, using Gecko G320s or G320Xs, and CUI capacitive encoders set for 2000PPR, and a SmoothStepper. Regards, Ray L.

Ray, didn't you start out with HEDS encoders? What prompted the change?

Cheers,

BW

[CalG]

With encoders on both knee travel and quill, wired into a "combiner box" feeding the Controller. Your tool offsets could be maintained with any knee position.

Don't ask me how to do it though ;-)

[HimyKabibble]

Originally Posted by BobWarfield Ray, didn't you start out with HEDS encoders? What prompted the change? Cheers, BW

Bob,

Actually, I have a mix. HEDs on X, Y and knee, with G320s. CUI on the quill, which has a G320X. The HEDs worked fine on the quill, but did need the pullup resistors. At this point, I'd use nothing but CUI - half the cost, adjustable resolution, smaller, sturdier, mostly metal, construction.

Regards,
Ray L.

[79TigerPilot]

Pictures of my restorationand conversion can be seen at:

http://picasaweb.google.com/cschaffter/CNCMillProject#

Z axis pictuers are around #53 and #133

Craig

Originally Posted by Jet A Interesting comments. Ray, where can i find out more about your set up? I am not interested in super fast rapids, just a little more accuracy and less labor intensive than spinning that stupid crank! any downside to driving the crank? That seems like a pretty massive stepper, i suppose there is quite a bit of weight there. Either of you have an pictures>?