2UVR to CNC

[Flyinfool]

I have a Tree 2UVR that I would like to convert to CNC.
The X and Y should be pretty straight forward.
The Z axis is where I am not sure how to attack.
The easy way, but not the best way would be to drive the knee up and down.
I am at a loss as to how to drive the quill without a ton of backlash.
Any one try this yet, or am I breaking new ground.

[cadman]

Your in the same boat I'm in. I want to convert my Tree to a full 3 axis but I'm not sure how to go about the Z axis either. Someone posted something about this a while back but I can't find that thread.

[Adobe Machine]

Did a retrofit on my Tree 200 Tradesman. (1979 model)Thought about using the knee as my Z axis, but calculations showed it way heavy...so I built a Z box with a ballshaft and nut, hooked directly to the quill.Since I use the 2uvr parts book, assume your quill has the same guide as mine does(right in front) that is attached with 2ea 1/4-20 allen bolts attached to the quill housing.I duplicated the quide, but added a horn to connect the ball nut flange to. Does your head casting on the right side(facing the machine ) directly below the manual quill handle have a flat small raised pad ? If so that is where I drilled and tapped 2ea 3/8 18 th holes and mount the right box plate.On the left side, below the quill clutch knob there should be enough material to drill and tap 2ea 3/8 holes for the left hand plate.Used 1/2 " 6061 T6 plate,so drilled and tapped (1/4 20) into the left and right plates and fabricated the front plate.The top plate is the thrust bearing plate,used 2ea Barden 304HDL bearings and calculated about 0.01 "crush"( preload), the bottom bearing is just for alingment.Drive the ballshaft on the bottom with a the same type servo as the x and y axes thru a timing belt. Accuracy after tuning = 0.0003 to 0.0008 in 7 inches.Some pictures of the my Tree are in the photo Gallery ( Adobe Machine)Works good.
E mail me if you want more pictures or info on the retrofit.

[daking]

To each his own, but converting a 2UVR or worse yet a 2UVRC to CNC seems a waste of a fine manual machine. The collet system, the nice smooth feel of the quill and micrometer adjustable stop. The gearbox power feed and power knee. The nice full round handles and smooth solid feel of the table and saddle. Thats what sets this machine apart. Add a couple of Trav-a-dials you've got a great machine. A retro-fit would be just another conversion. Do that to a Bridgeport or some other generic old mill. A Tree should be left in it's natural state. Just my 2 cents worth.

Dave

[Adobe Machine]

Unless some one operated a Tree Mill after operating a Bridgeport or clone, I doubt they would appreciate the how smooth a Tree is, but mine is one of the 1st attempts by Tree at NC in 1979,and does not make a good manual machine. With new electronics and adding a Z driver for true 3 axis CNC, these units are exceptionally stable and accurate.
I have seen very few Tree Manual machines on the market, which indicates people that have them ,keep em ! Not sure about taking a manual to CNC, as the cost of ball shaft and ballnuts may be prohibitive. I have not seen any "kits" advertised except for Bridgeport and clones.
Certinly before attempting to retro any manual machine a person should do a carefull cost analysis as opposed to buying a purpose built older CNC ,that updating electronics and going to a PC operated system may be more economical and quicker.

[Flyinfool]

The difference is that this Tree has been in my basement for the last 7 years now. Its not like I am looking to go out and buy a mill to convert.
And I certinally do not have room for a second mill or the desire to haul 3000 lbs of cast iron up or down the basement steps again.

My intent with the conversion is to maintain the ability to do manual machining or CNC. I will use ball screws with the same pitch as the original lead screws so that the hand wheels are still accurate, I will also leave on the DRO.

I have run into a bunch of projects that took days of setup to complete manually and would be handled in minutes with CNC.
Especially when a good friend looks at it and says "cool, can you make me one?".

[daking]

Originally Posted by Flyinfool The difference is that this Tree has been in my basement for the last 7 years now. Its not like I am looking to go out and buy a mill to convert. And I certinally do not have room for a second mill or the desire to haul 3000 lbs of cast iron up or down the basement steps again.

I'm not going to try to argue with that. Those are solid reasons. I cant even imagine trying to move one of those things down stairs. My argument is more for those doing it in a business than a hobbiest or someone moonlighting. Sounds like a pretty good challenge though. Especially the quill.

Good luck

Dave

[Flyinfool]

Originally Posted by daking Those are solid reasons. I cant even imagine trying to move one of those things down stairs.

Half of the fun of that machine is the look on peoples faces when they see the Tree and look at the basement steps with 2 90* turns.

I have a couple of new thoughts on the quill.
If I am willing to give up .50 - .63 inch of travel I might be able to extend the quill down a little and clamp onto to it and then drive the clamp with a ball screw.
The clamp would have to have a quick release of some kind so that the quill could be raised all the way up to change collets or I will need to devise some other method of operating the collet closer.

or

I removed the Tree logo plate on the front of the head and under there is the worm gear from the power spindle feed. if I remove that worm gear and attach the servo motor directly to that shaft it could drive the quill up and down with a 20:1 gear reduction that is already built into the head.
The only problem is getting rid of a lot of backlash in those gears.
Maybe I could replace the spring that now is a counterbalance for the manual down-feed with a much heavier one to preload the gears.

Ok I'm just thinking out loud here, feel free to shoot my thoughts full of holes.

[Adobe Machine]

Did evaluation on that system too; Problem with the backlash in the rack is you would have to load the quill "up" and the "pinion" down,thereby loading the gears and possibly eliminating backlash,but milling will tend to pull the tool down opening the loaded gears and causing back lash.
Any system like this will work as long as the movements are Z minus (down) and you have calculated the software ratio correctly, but the minute you have Z plus (up),there is so much backlash it will take over 0.020 to move the quill up,thus losing any referance as to position.This will destroy pocketing movements , countersinking etc,and trying to run a peck drill program with this much backlash is real frustrating
(and dangerous).
Wow, how did you get that mill in a basement down stairs..That must be some story.
Any exciting moments ?

[Flyinfool]

Originally Posted by Adobe Machine Wow, how did you get that mill in a basement down stairs..That must be some story. Any exciting moments ?

It was well choreographed and planed out and as a result quite boring.
Even though everyone helping was sure it would not work.

To bad I didn't think to take any pics of the process.

I will have to look closer at the spindle drive gears in the head to see if there is some way to get out the backlash.

[quintc]

I'm in the same boat. Just got a 2UVR and can't wait to strap my stepper motors to it.

I started analyzing where the slop in the quill lies and how much. The places I can find slop starting from the tool are:

1. Quill to the hand feed lever "pinion shaft"
2. "Pinion shaft" to "horizontal worm shaft" behind the Tree emblem
3. "Feed worm wheel" (the gear behind emblem) to "verticle worm shaft"

To measure #1 I put pressure on the hand feed lever to lock up the rest of the mechanism and with an indicator mounted under the quill, lifted up on the quill itself. Suprisingly I got only 0.003-0.004" movement.

I thought #2 was where the bulk of the slop was until I pulled the Tree emblem off and saw the gear and shaft sliding significantly in and out with the hand feed lever movement. With the gear and shaft retracted I put pressure on the gear with my hand while actuating the hand lever, looking for movement on the indicator. I saw only 0.002" movement before the gear and shaft would begin to push out against my hand.

#3 has so little if any slop and is geared down so much from the quill that I don't think I can even measure the effect on the tool.

Conclusions:
It appears to me if you could constrain the horizontal worm shaft with a bearing you would have only 0.005-0.006" backlash in the quill with a motor mounted to the shaft. If you can add a stiffer spring to the feed lever as mentioned above, strong enough to overcome negative Z forces from helical tools, you could take out the 0.002" and potentially only have the 0.003-0.004" in the quill rack. Adding some kind of external HEAVY spring to the quill itself could take out all the slop, but as previously stated you'd be interfering with the collet mechanism.

I'm only going to be using this for home projects and I'm not even going to bother changing out the lead screws for ball screws. I'm going to simply let my software adjust for backlash and call it good.

What do you guys think? Have I missed anything?

Quint

[quintc]

I looked into the horizontal worm shaft a little more and realized the whole thing was being held in with a set screw on the left side behind the housing containing the gear. The set screw sits in a divot in the side of a bushing on the shaft. The set screw retains the bushing which is constrained by a step on the shaft and the gear in front. It turns out my set screw was a little loose, but even once I tightened it I still had 0.008" of travel on the shaft. Near as I can tell this is just clearance for the bushing. I suppose I could just machine the bushing down enough to tightly fit a thrust bearing between it and whatever coupling I end up putting on the shaft. The end of my coupling could even have some sort of lock nut for taking up the slack. I am suprised to see a set screw being the only thing supporting the thrust of the Z-axis, but the manual says the clutch on the hand feed pinion shaft is designed to slip in case of the quill bottoms out against the stop.

Quint