What's new
What's new

Tree Journeyman 325 ballnut rebuild

MechWerks

Cast Iron
Joined
Dec 11, 2007
Location
Nebraska
Went to set the backlash comp on a new to me Tree Journeyman 325 and was not happy with the numbers already set into the controller...time for new balls.

The maintenance manual (only one I got witht the machine) describes the X and Z ball screw removal, but skips the Y axis (an ominous indicator).

Just for anyone else wanting to travel the path...here is what I found.

Everything related to Y axis:
Remove the drive belt cover, remove the servo motor and belt.
Remove setscrew at front of metal way covers and slide covers out front of machine (ref picture). My machine doesn't have soft way covers yet...so you will have to deal with those as necessary. Note that the picture shown already has the ball nut unfastened and is hanging...at this point, just screw the screw out the front of the machine and you will have decent acess to the ballnut fasteners and oiler nut through the top of the knee opening. Gifted with long arms, I could touch the end of the screw while turning the screw out to ensure I stopped flush with the ballnut and not out of it just yet.
 

Attachments

  • Ynutout2.jpg
    Ynutout2.jpg
    12.8 KB · Views: 1,184
Now there is good acess to the oiler nut (remove first) and ballnut fasteners. An extension and allen socket on a ratchet takes the fasteners right out. The tough part is the oiler!
 

Attachments

  • Ynutout1.jpg
    Ynutout1.jpg
    14.2 KB · Views: 419
After the ballnut oiler and fasteners are removed, the screw can carefully be moved to the rear of the machine and the ballnut will stick out the top of the knee opening (take care not to whack the oiler flange fitting).

I turned down a "tool" from 1-1/4" tube with a length that fit inside the ballnut to 1-1/8" diameter. Put that "tool" to the end of the ballscrew and turned the nut off onto the "tool". That kept the balls from exiting the ballnut until I was ready to see them.
 

Attachments

  • Ynutout.jpg
    Ynutout.jpg
    14.8 KB · Views: 911
Now the "tool" can be pulled out of the ballnut, dump the old balls and flush out the years of chips and crud.

The picture shows the reloaded nut ready for install.

The nut on this Tree is a two circuit ballnut. The ball returns are removed (three screws, remove top clamp, and very carefully remove ball tubes...didn't even want to think about sourcing replacements!).

Got .125" grade 25 chrome balls from Enco for $2.29 per 100. Each ballnut takes about 250 balls (242 came out of this one). Take care with the replacement ball "grade". McMaster had grade balls that were +/- .0001" which seemed sort of a wild range...grade 25 is quite a bit tighter and price was not significant.

Reloading the nut is done on the removed Y screw. Screw the ballnut down the screw about half way. The seals will require threading the ballnut on and is best if they are installed prior to reloading. If you took the seals out of the nut...you will notice that they have to go back the exact same way or the nut will not go on..best to learn before loading balls.

Loading the balls is as simple as drop in a ball and turn the screw to suck in that ball, drop, turn, drop, turn. Or, I took a short length of clear tube, layed it in a tray with the balls, let balls roll into the tube, pinch the end, stick end into ballscrew opening and let the "magazine" load balls as the screw is turned. Dramatically speeds up the process!. Keep turning in new balls until they appear at the end of their respective return hole. Remember, this is a dual circuit...so you only want to fill one run at a time. Don't let any balls past the circuit they should be in. Put some grease, vasilene, etc. into the a ball return tube, pack with balls, carefully push into the correct loaded circuit and use the clamp and a screw for a temporary hold. Do the second circuit the same way.

You will notice the nut has significant drag now. The old balls were only about .0001 under on the Y on this machine but caused .0011 backlash. The X balls were .0004 under and causing .0047 backlash. Now the Y is at .006 and X down to .0011. After the comp switch, each .0001" step seems to indicate pretty decent. The old dog will be close enough for what it will be asked to do.

Roll the complete ballnut off the screw onto the "tool". Stick the screw back into the knee and poke the end up out of the knee opening. Screw the ballnut back on to the screw, taking care to keep the "tool" tight to the end of the screw until balls thread on as they should and run the nut on until past the seal. Pull the ballscrew back out the front of the knee and take great care to not ding the screw or hit the oiler line near the ballnut. Now the toughest part is getting the oiler nut back after reinstalling the ballnut fasteners. A pretty straight forward reversal of previous steps to put every thing back.
 

Attachments

  • Ynutclean.jpg
    Ynutclean.jpg
    12.2 KB · Views: 677
If doing the X axis, the nut can be rebuilt on the screw without the transfer "tool". Remove the screw assembly but leave the prealoaded bearing mount in place. Take off the ball returns and eject the balls by turning the ballnut. My seal was broken on the pulley end...so it would expand and come off the pulley end shank. It appears the seal is still serviceable so advantage of the situation was taken...The ballnut will come off the pulley end of the shaft over the unthreaded shank without the balls making cleaning easier. It could have been cleaned in place if the seal had not been broken buy screwing seals out of the way and flushing. Then the clean nut is reloaded on the screw with the same drop in a ball and turn process described above to fill each circuit. Hint...put on the loose oiler tube to near tight with the ballnut screwed to the preloaded bearing mount end. Then turn the ball screw to draw the ballnut towards it's mount. Really saves trying to get the oiler tube back on the nut...works the same for getting it off.

Maybe a lazy way to get the job done, but avoids messing with the preloaded bearings (didn't have the spanner wrench to mess with the preload nuts right now anyway) and maybe there is some more backlash to pull out of that too, but right now things are looking much more functional then they were and one of the cheapest projects related to putting this mill back to work.
 
Steve, was there a lot of swarf in the nuts?

While the lines are off that is a good time to flush them and make sure that oil is getting to the nuts.

I had one of those round hook or pin spanner nuts break in half on me when removing the screw. It had pretty clearly been fractured for some time.

I've never checked to see if there is any float in the angular contact bearings. I suppose I should find the time, just to be sure.

cheers,
Michael
 
Yes there was trash in the nuts. May have actually been helping the backlash on the old balls.

Yes, flushed the lines...don't want to go back anytime soon.

The load bearings should be easy enough to check with an indicator against the screw end and reverse travel to see if it moves.

Gotta thank you guys for posting up enough information to make one of these machines look like an option as a starter machine. There may be some issues but the bones are solid and hopefully it will at least pay for itself in education. You guys have shared quite a bit...my turn to give a little back.
 
thanks!

I'm actually pulling the trigger on a 325 this week as well, and the backlash looks to be right around .0015-.002 on both axis, so this may be be one of the first tasks when i have downtime waiting for a phase converter.

do you have any idea if the tightness of the new balls is excessivly tight to cause flaking of the screw or nut races? i.e. surface fatigue failure

also, were all of the old balls the same size? i read on a posting that some ball screws use alternating ball sizes to minimize binding. would it make sense to alternate using new and old balls to get that same effect?

Mike
 
Mech

Keep turning in new balls until they appear at the end of their respective return hole.

I just wanted to ask you about that. Did you manage to get about the same amount of balls back into the nut?

No offence, but I see this often enough when some one sends in a screw for repair. You have to be careful particularly on the first circuit when the nut is loose. If you have a couple of turns in the ball circuit, meaning they go around the screw twice or 3 times before they pop out at the ball return tube hole. As the balls come over top dead centre, they will often fall down the screw, leaving gaps in the ball train.

I call that the gopher ball, when you see that starting to come up the flight of the screw approaching the ball return hole. Put a nice fitting pin punch or piece of stock down against that shoulder that the return tube sits on, then roll the nut some more just as you had done. The balls will all gather up again, and you ensure there’s no gaps in the ball train. You will often get another dozen balls to go in your filling hole, but the gopher ball is trapped by the pin punch.

Also the balls once they are well into the nut and making good contact between the nut and the screw flanks will roll away faster than the ones waiting to drop into the flights. That also causes gaps.

Then fill your return tube as you did, subtract one ball, because they can’t be tight, and you’re good to go.

Well done on making up that loading sleeve. It’s a little known trick.

Regards Phil
 
I did a practice run with the old balls just to see what to expect in loading. I also did stop the end of the circuit to make sure the circuits were completely packed, less a ball for clearance. I actually put back more balls than original...it didn't appear the original load was all that accurate. Then again the machine is supposed to be new in 1991...a lot of history can befall a machine over that sort of service life. The idea was to make it do some more.

"backlash looks to be right around .0015-.002 on both axis, so this may be be one of the first tasks when i have downtime waiting for a phase converter. "

Just since I have "been there" recently, be sure you go into the backlash compensation setting of the controller (assuming original Dynapath..but should apply to any) and check for existing backlash comp...set RX and RY to 0, then check backlash. I was followign the setup and maintenance manual and saw a small backlash, until I went to set comp and found some big numbers already waiting there...the controller made the backlash look pretty good initially. I am guessing the original balls were all .125" as they were pretty evenly sized per the screw they were on. I realized that the machine could have spent significant time in one spot of any given screw (meaning loose in one spot and tight in another). I checked several places in travel and was happy (sort of) to find pretty much the same backlash all over. In order for a ballscrew or any other preloaded bearing situation, there is going to be some tightness to take out any slop and that is normal. These machines are rated at 1700# axis force. The old balls let the ballnuts spin quite easily. New balls required a couple ounces of torque to turn. Yet there was still some backlash...so it couldn't have been too tight. Original accuracy was supposed to be .0005". On a machine with this much age on it, I would be very suprised if replacing balls could really help that much and obviously new screws would be the correct repair.

This is just a case where high accuracy is not a significant issue, balls were cheap and things are better than the were for just the cost of wrench time and a few bucks in parts.
 
so i did my backlash measurement roughly, and probably not that accurately. i didn't think it through until after i got home form inspecting the machine. i had a test indicator on the knee and moved the table a total of about .002" in the Y direction by pushing and pulling on it by hand (very hard). then with a mag base on the table and indicator on the side of the vertical surface of the Y axis box way to measure X. the backlash parameters were all between .0004 -.0006 so i do have some room to adjust before digging into the ball nuts.

from what i understand, the accurate way to measure backlash in the y direction for example is to mount the indicator base on the way or side of the knee. then move the table to a certain position in one direction (say the negative direction). zero the indicator on a vertical reference surface, then cycle the y axis in the negative direction for .1" or so, then back to the original indicated position in the positive direction. the difference displayed on the test indicator should be the backlash. is that correct?

it would seem important to mount the dial base on the other side of the way, not on the spindle or other place to isolate the backlash to that particular axis. i.e. no other ways/ballscrews/general sources of error in series with the test indicator.

what were your original backlash parameters?

thanks,
Mike
 
I set a .0001 indicator on a mag base on the Y ways to check table movement for Y and X. Soild spot against a solid spot. In the manual, it specifies to move the table one direction until movement is noticed then use the .0001 jog to move the opposite direction until the indicator moves. The number of times you click the switch (also on the read out) is the backlash. You will want to set up on the end of the screws first to see if you see any backlash there. Be sure to clear out the "setup" backlash before hand to keep the controller from tossing in that compensation. Then you get to see the full amount of direction change needed to move the indicator.
 
Bumping an old thread to get more information. While it is my intent to re-ball my screws I found that my Y-Axis, on a J325, is very tight. It moves but nowhere near the ease of X or Z. I would like to understand why before I replace the balls. What are the items that cause drag? The ways seem to be well oiled, the ball screw is clean. My questions is are there other adjustments I should look at before the re-balling? I see there is a grease fitting on the Y-Axis shroud - I have never greased it could the bearing be an issue?
Sorry for all the beginner questions but I am in the process of starting Retrofit 2.0 and want to cover as many maintenance items as possible.

Thanks

Bill
 








 
Back
Top