I am having some nasty problems with my ballscrews, I must be getting junk in them or something, Here and there a few of the axes will want to go somewhere usually during a rapid and then I usually rapid into a peice, like when I was just peck drilling, I have it peck .01 down .1" at a time then rapid up over the peice then back down, yeah, I rapided up and it didnt go up, then slammed into my peice of steel.

I oiled my z axis ballscrew like crazy, just dumped and dumped, it is still doing the same thing. thats the one that slammed,

shortly before I was cutitng some other stuff and it happened but with the X and the Y at the same time, is it maybe my motors or something? its really bothering me and I am scared to mill stuff to find myself breaking things.

Jon

Hi Jon,

If you are using steppers, maybe you are losing steps from overloading or binding of a slide?

What kind of nc code are you running? Is it correct?

I am running G-code and it is correct, I hand wrote it:p I used the same code to drill my 3/16, 1/4, and 5/16" hole then I was working on the 3/8" hole(stepping up the drill bits in steel)

Its likely that I am having problems in the Z axis because the head is kinda heavy, the others I really dont know what the deal could be, here and there I think I do hear something running past the ballnuts.(in the X axis)

Jon

Try lowering the step rate a little or lowering the maximum rapid speed.Then see if it still stalls.

sometimes when I am moving the head up and down it sounds a little different than the regular stepping sound, kinda like its binding or something, could it be something else like the head weighing too much? shoudl I get that torsion arm back on? oh yeah, I really cant becasue my ballscrew is in the way of the bolt that goes through.

I really dont know what to do atm. I was planning on machining some parts for my brothers motorcycle I still hope I can do that monday.


Jon

You can counterbalance the Z with a spring. You should retract the z from a hole at feedrate, then rapid once you are out of the part (if at all).

Eric

Jon,

Try the counterbalance idea.

It is possible that a ball could be damaged and a fragment is getting jammed on occasion. I would remove the ballscrew, and run the nut back and forth manually, just to see how it feels. Then, apply some end pressure to the nut and see if there are any catchy areas.

Inspect the ballscrew for damage in the threads.

If you still cannot find anything, then it might help to wash the ballnut and screw. This you would do by dunking it in solvent, then running the nut a little and then dunking it again. This would be an attempt to wash a chip out.

You did not mention the construction of the ballnut, but it is also possible that it was incorrectly assembled, and one (or more) balls is not in any of the ball circuits at all. If it is a dual ball circuit (with two return tubes), it is possible through improper assembly, that one (or more) balls can get trapped in between the two tubes. It will roll freely for a ways after reversing direction, and then suddenly jam up when the stray ball hits one of the return tubes. The ballscrew will have to be disassembled and then reassembled correctly.

I think I will have to try dipping it in solvent and runing the nut back and forth. I cannot take these nuts off the screw without balls flying everywhere right?

I dont get what you mean by the construction of the ballnut. Liquidrulez(possibly from these forums and others) build pretty much everything on this, he has built his own and I think a few others, I think he might know what he is doing but I really am not sure. He did dissapear with no trace though, thats after I got pretty much everything from him.

These ballscrews and ballnuts are thompson from reid tool.

On the Z axis, it is 2 ballnuts preloaded against eachother.


I dont know how I will rig up a counter balance, thats the only problem I am seeing.


Jon

Jon,

That's right, you cannot take the nuts right off the screws, unless you are prepared to catch all the balls and then painstakingly reasssemble them.

It could be that the preload is just a little bit too tight. Usually there would be a two piece shim between the two halves of the nut. If you take the keys out of the nuts, then you can remove the half shims and try to shim it differently. This would be a highly precise shim modification required.

I would try the counterbalance first, just by rigging up a cable and pulley over top of the machine with a bucket attached to the cable, to which you can add counterbalance.

Counterbalancing is definately a good idea. Can you use gas struts from a car hatch?
Also try to not rapid when going up on Z. (If you use steppers). They have less torque on higher speeds, and if they loose step(s) the controller will believe that Z is higher up not knowing it lost some. Then when going down again it will go too far down, hopefully not into your table. If you go slower, the motor will be stronger thus not loosing steps. If that helps, higher motor voltage or bigger motors should help.

If using servo's, forget what I just wrote, they should just fail out. Or if error within error band they will catch up again when going down.

But first of all: you said you oiled and checked the ball screws. Did you do the same with your Z-slides?

I beleve the preload is fine, I can hold the screw asembly and watch the nut spin around all the way to the bottom. I think i will try the pullie and cable thing sometime soon, maybe get ahold of a gas spring when I make my new motor mount and pully setup.

I am running steppers and I beleve thats my main problem, can I change the rate they rapid at? I think I might have seen it in turbocnc but I am unsure.
I do have the Z dovetail greased up real good with white lithium grease.

I was hoping to win those servos and solve my problem;)

What about sending them through a parts washer? do you think that maybe that could help?


Jon

NO! Then you would have to take the ball screws apart first. And that is only recommended if you know how to do it, and most importantly how to put them back together. Just washing them as a whole will probably just remove the lube but any debris will not come out. If going carefully and making sure you don't loose the (very small) balls, it can be done with most ball screws. If there are more than 1 ball circuit, make sure you count the balls out of each so that you get the same number back in.
Use a large plastic tray to do all the disassembly in so that the balls don't disappear on the floor. I had to do that with one of my small ball screws, and found that the debris was actually part of the ball guide edges broken off. They were ground to a knife edge and so it was bound to happen. In other words a bad design from the manufacturer. That's what you get when the manufacturer does not put his name on the product I guess. It was done in about an hour, and now runs smooth as any.

A digital camera is very nice to have when dismantling mechanics that you don't know well. It may not be obvious when you take it apart that the 2 ball guides are different on some designs Pictures at each step could tell you which goes back where.

ok, what do you think about reducing the frequencies of the laptop so my max ipm for the Z axis is around 10-20IPM and the max for the others is around 30? and reduce acceleration a lot?

Im thinking thats what I might end up doing along with a counterweight. These motors do have a rather high RPM I woudl guess I am shooting rapids around 60+IPM(turbocnc doesnt tell me)


Jon

Just to test if that's the problem can't you just edit the G-code?

yeah I could, but I think that if I change those it would be simple to change, and I can use the same G-code.


Jon

I think a good place to start would be to lower the max speed and acceleration for the Z axis. The easiest place I've found to do this is in the turbocnc.ini file. It looks something like this:

[AXIS1]
Designator=X
StepIncrement= 3.12500000000000E-0005
IsLinear=True
IsStep/Dir=True
PortAddress=$378
StepPin=4
IsActiveLow=True
Pulsewidth=0
DirPin=5
LowIsPositive=True
Acceleration= 4.00000000000000E+0004
StartSpeed= 5.00000000000000E+0003
MaxSpeed= 1.60000000000000E+0004
ABSPrescale= 1.00000000000000E+0000
Motor Resolution= 200
Leadscrew Pitch=20.0
Controller Microsteps= 1
Gear Reduction= 1.0000

It's been a while since I was playing with this, but I believe all of the speed values are in pulses per minute (I could be wrong, you should double check that). Basically, a higher value is faster, lower value is slower. Also, the X, Y, and Z need not have the same acceleration/MaxSpeed values.

that stuff can be edited in turbocnc too. tahts what I think I will do, actually I think I will lower it for all of them because I have had the others give me crap too.


Jon

well I decreased the rates and acceleration, I beleve it was the acceleration that was giving me problems. The z axis is going really slow, about 15ipm max wich is fine with me. the others are around 20-30 Im guessing. I cut the acceleration in half for x and y and from 5k to 1k on the Z axis.


Jon

it tells you in tcnc setup what your rapids are based on your settings. Have a look under "motion setup" I think it is.

Eric

yeah, it said around 60ipm or something like that but thats a lot. I guess it could be.

Jon

Well, I found out what my problem was, I have been working with the mill on slow rapids, I took it apart a week or so ago and have been trying to pinpoint the problem, I think I have found it.


Have you ever noticed how that when you tighten the regular leadnuts that it removes some backlash but it makes it harder to turn? The slots are slightly angled. When the ballnuts have that angular force on them they bind.

If you have skimmed any of this post, read this:
I think my ballnuts might be wrecked, but they still might be good, if they are bad Ill have to buy new ones, but heres why I think they might be wrecked, I have about .005" or more of backlash, that does not seem right for ballnuts.


let me know how much backlash you guys have,
Tomorow Im going to take that part of my mill to work and squair up the slots on one of the bridgeports.

Jon

Jon,

I'd be looking critically at the mounting surface that the ballnuts are being tightened against. It is more likely that there would be an out of square condition there, than that the ball nut flanges would not be true. You might even stoop to trying shimstock under the bolts on one side, in order that it not be binding when tightened up.

Inexpensive rolled ballscrews are likely not going to be very tight. That is why guys use two nuts back to back, to try to make an antibacklash, preloaded arrrangement. With the correct spacers between the two ballnut (and lots of trial and error) you might be able to reduce the backlash somewhat.

Yes, thats waht I have said, and Im bringing it to work to true it up in a mill, just shave a little off and itll be all good,

I have Thompson Saginaw ballscrews and nuts, I have my Z axis preloaded, but there isnt room in the x or y for preloaded nuts.

Jon

I took it to the shop, milled it and its rollin smoothly:D :D :D :D



Jon

Jon, I just read thru this entire thread for the first time. In your first post you mentioned stepping a drilled hole out in steel from 3/16 to 1/4 to 5/16 etc. If you'll check the distance across the web of the drill that fits the final hole size, and drill your pilot just slightly larger than this, then go right to the final size drill you'll find your drills will last a lot longer. It will keep the cutting edge from chipping at the edge of the previous hole like it often does when stepping out a hole. I'll be darned if I could explain why it works, but it does.