[dan3676]

I'm upgrading my sosylva diy cnc from all thread to acme 3/8-10 2 start acme rod purchased from McMaster and anti back lash from dumpstercnc. Even after running the anti back lash nuts about 20 feet worth of the rod on a drill at about 400rpm they're still a little difficult to turn on the rod. Easier than when I started but not near as easy as a nut and bolt would be. I'm concerned with excess drag on my motors? Is this normal with the tension in the back lash nuts? I've double checked threads from couplers I've purchased and everything seems to be lining up correctly. Any ideas?

[awerby]

Have you oiled the screws? That's important, even with Delrin nuts. Yes, the anti-backlash feature will make the screws harder to turn; that's because it's pushing on the threads in more places, and the spring has to be fairly strong if it's not going to be overwhelmed immediately by cutting forces. But as long as the motors aren't losing steps, you're probably okay - steppers don't mind working fairly hard, up to a point.

[dan3676]

I haven't oiled them. I was a little concerned and unsure about lubrication because the lead screws are exposed. I didn't want to worry about debris sticking to the lead screws and getting caught in the nuts. I was having a little trouble with stalling with my all thread but after removing my old lead screws and moving things freely I think I was pushing the rpm of my drive motors. If I'm using xylotex drives and 200 oz motors with 3/8-10 2 start lead screws what dot settings need to be in Mach 3?


Sent from my iPhone using Tapatalk

[ger21]

8000 steps per inch.

From Dumpsters website:

What if my nuts are too tight?
Umm... ok... They shouldn't be too tight, only snug and you should be able to turn them with a finger and thumb, you shouldn't have to grip them for them to turn. All lead screws are not created equally, if you use a different supplier than I recommend (sometimes even the same supplier!), they may fit differently. We've had good luck when I suggest to file 2 or 3 notches around the last 2-3 threads of the lead screw and force it through the back of the leadnut only under the flange area, basically working as a custom tap; it usually doesn't take much. You are welcome to try this, otherwise, you can return them and I'll see if I can open them up here.

I use pneumatic tool oil on my dumpster nuts.

[arizonavideo]

The lead screws need to break in. After a good amount of running they will smooth out to a glass like surface. At this point they will have vary little friction and the nuts will last almost forever with no wear.

Unfortunately the break in process will wear the nuts a lot. I ran mine in at too high of a speed and over heated the nuts so they ended up with backlash. I bought a few new ones but ended up making a tap and making a new set of nuts. Now that the screws are broke in I have low backlash and the nuts don't look to be wearing at all.

I have had the router table running for 4 years with the same screws. The 8 TPI screws are a bit slow for a few things and the longer 20" axes do like to whip over 100 IPM speeds.

[wizard]

The lead screws need to break in. After a good amount of running they will smooth out to a glass like surface. At this point they will have vary little friction and the nuts will last almost forever with no wear.

Part of that break in involves the leads crew itself. It isn't just the nut theta gets polished but the burrs and roughness on the lead screw get polished up. This is why break in over the entire length of the lead screw is ideal as you don't want to leave any roughs spots on the screw to mess up the backlash adjustments. There is some argument to the idea of having two sets of nuts so that after the initial break in and wear you can fit up a tighter fitting pari of nuts.

Frankly manual lathes run into a similar problem where the lead screw for saddle drive will often wear the half nuts faster than you would like.

Unfortunately the break in process will wear the nuts a lot. I ran mine in at too high of a speed and over heated the nuts so they ended up with backlash. I bought a few new ones but ended up making a tap and making a new set of nuts. Now that the screws are broke in I have low backlash and the nuts don't look to be wearing at all.

One thing people need to watch out for is thermal expansion in the plastic nuts. A closely fitting nut can actually tighten up under load (heat and other factors) grabbing the lead screw. Then you end up with a stiction like problem.

I have had the router table running for 4 years with the same screws. The 8 TPI screws are a bit slow for a few things and the longer 20" axes do like to whip over 100 IPM speeds.

[Devastator]

I polished my leadscrews by putting them in the lathe and running ultra fine steel wool over them forward and reverse in the threads while it was turning. I made my own AB nuts and they have some serious production time on them now, still backlash free. Using the plastic nuts to smooth out a steel screw seems a bit silly to me. Smooth them out first before they have a chance to wear out your nuts.

[wizard]

I polished my leadscrews by putting them in the lathe and running ultra fine steel wool over them forward and reverse in the threads while it was turning. I made my own AB nuts and they have some serious production time on them now, still backlash free. Using the plastic nuts to smooth out a steel screw seems a bit silly to me. Smooth them out first before they have a chance to wear out your nuts.

Actually you have good point here. This is often suggested for leadscrews on chinese imported lathes. Not so much steel wool as it can catch easily but a little hand work to remove burrs and such can do wonders. The leadscrews on these lathes are especially bad due to the burrs from the keyways.

Given that you will still find that your nuts will wear in on a new lead screw even if you have cleaned it up ahead of time.

[crob09]

One suggestion would be to look at the RepRap and how they use threaded bar stock but still achieve accuracy of .01 MM or better.

Also if you need to calculate steps per inch or steps per mm you can find a handy free tool at steps per inch mach3 wizard Conversion tool just enter TPI and Step angle to get steps per MM

[bhurts]

Rep rap may achieve a resolution of .01mm but there is no way it is that accurate.

Ben

[crob09]

Rep rap may achieve a resolution of .01mm but there is no way it is that accurate.

Ben

Why not?

Mine is.

Need read back? (Closed loop) look into Hall Effect Sensors, sometimes it's not about the hardware it's the math behind the software.

[ger21]

Unless your using screw mapping, and a way to measure the entire length of the screw to .01mm accuracy, then no, it's not that accurate.
Typical leadscrews have a tolerance of ±.009" per foot. So in 12" of travel, it can be out .22mm.

Now, it can certainly be repeatable to .01mm, but it's not accurate to .01mm.

[crob09]

Yes you get it!!!

Hall Effect Sensor, this device will give Read-Back to the PLC/ PC, you can use it with magnets to create a Closed Loop System. (Look at Digital Micrometers)

If you want accuracy you require a closed loop system.

However most applications don't need this, my Printer is pretty accurate and that's after it's completed how many passes over and over and over to build the part, the Z-Axis will lift a whopping .01 MM each time without fail giving the resolution, this is with threaded bar stock from Home-Depot or Home Hardware, two nuts are pushed apart by a spring, taking up the slop and removing Backlash.

The solution to the question asked is the Rep-Rap and how the printer was engineered.

[ger21]

the Z-Axis will lift a whopping .01 MM each time without fail giving the resolution

It probable moves .008, or .011, or .009, or .012 most of the time.

[sudip]

I think this post can help you.

http://cncmachinetools.org/ball-screw-in-cnc-machine/

[crob09]

You have one right?

So you would know this?

from my measurements it's accurate, I have three of them.

Also have you considered losses in other areas, for example the bearings, it's not all the screw.

As noted before Read-Back (Closed-Loop) is what a person will need for accuracy when cutting/ printing/ carving over a large work area, small machines won't have the same problems. "Slop" is compounded over distance, consider a triangle, where the lines almost meet/ intersect there is very little distance between them, as you move away from the intersect the distance increases, it's the same concept with the CNC machine.

Check out a Tangential Knife Cutter with Ultrasonic Blades, these machines are massive in footprint, I've used one 10Ft x 15Ft all servo driven on a pinion rail gear, no screw, but the machine is accurate in everything it cuts because of Read back, in it's case an Opto for the tangential blade and encoders for the X-Y axis.

[bhurts]

Do you have a way to measure .00039 inches? Was your machine but together and measure by someone qualified to certify that tolerance for squareness, distance traveled, perpendicularity etc... Can you honestly say that the machine is rigid enough to never flex .01mm or that the materials it's made of won't grow or shrink that much when it gets warmer/colder. Rep rap is a fine machine for what it is. Does it hold the tolerances of a machining center. No way.

Ben

[bhurts]

If you have a link to the tolerance info you are stating I would like to see it. I have not done extensive research but from what I have looked into I can find no stated tolerances of .01mm. As far as I can see .004 inches tolerance and .012 inches is the minimum build height. Also a link to there read head would be nice.

Ben

[crob09]

Do you have a way to measure .00039 inches? Was your machine but together and measure by someone qualified to certify that tolerance for squareness, distance traveled, perpendicularity etc... Can you honestly say that the machine is rigid enough to never flex .01mm or that the materials it's made of won't grow or shrink that much when it gets warmer/colder. Rep rap is a fine machine for what it is. Does it hold the tolerances of a machining center. No way.

Ben

Yes you can measure .00039 if you like, it's truly amazing what a processor can do.

No a RepRap will not work as a milling center but the concept used on the Z axis will work on other units, convert a conventional mill.

You need to determine what the material is going to do, that's part of your design process, all material is different, even the spindle speed causes changes.

You start the machine at Zero, square from that depends on your programming and machine, again the Concept used for the Z-Axis on the Rep-Rap is all I was referring to, I also noted that Red-Back would be required to insure accuracy.

Some of what you asked makes me question the experience you have operating a machine.

[crob09]

Look at Slic3r...
You can see it there.
Also a good idea would be to invest in building one, or even procuring a 3D printer, this way you can see first hand and gain valuable experience.

Also .004mm is finer than .01mm, and .012mm is larger than .01mm. So not sure what you're getting at here.