New SO3 Z-Axis

Assembly questions and answers for the Shapeoko 3 CNC Mill
zamp
Posts: 44
Joined: Fri Jan 02, 2015 8:54 pm

New SO3 Z-Axis

Post by zamp » Sun Jul 23, 2017 9:15 pm

My Sunday afternoon project was taking one of these eBay z-axis assemblies...
new z.jpg
new z.jpg (226.44 KiB) Viewed 1599 times
...and replacing z-axis on my Shapeoko 3.

I removed x-axis plate from my SO3 and removed everything related to the old z-axis. Because I wanted to minimize how much further out the new z-axis would stick out form the x-axis plate, I then removed the back plate off of the new z-axis assembly. I then used that back plate and a transfer punch to mark on the SO3 x-axis plate for new holes to attach the new z-axis to the SO3 x-axis plate. Two new holes were needed at the bottom of the x-axis plate, and two existing holes at the top needed to be slightly enlarged. I countersunk the back side of the plate for the two new and two existing holes, then mounted the new z-axis and checked for clearance from fasteners. I found I needed to shim out the new z-axis from the SO3 x-axis plate using a washer at each mounting screw. Mounted this way, it seems pretty ridgid, though I think I detect some slight flex of the SO3 x-axis plate now that the the custom extrusions are no longer mounted to the x-axis plate and effectively stiffening it -- I may end up needing to mount some bar to the x-axis plate to stiffen it, much like we used to need to do on the original, flate z-axis plates.

I then pulled off the new front slider plate from the new z-axis and drilled it for the SO3 router mount. I found I needed to file down channels in the bottom end plate on the new z-axis in order to provide clearance for the screws that attach the router mount. Instead of filing these clearance channels, I could have bought new flat head machine screws and countersunk the back of the new z-axis slider plate for those screw heads. But Fastenal is closed on Sunday and I didn't want to wait on the project.

That's pretty much it, other than re-hanging the x-axis plate on the SO3 gantry, attaching the pulley provided with the new z-axis to the old SO3 stepper motor, attaching the stepper to the stepper plate that came with the kit, attaching the stepper/plate assembly to the new z-axis and tensioning the new z-belt.

The final result is a new, spring-less z-axis with 5" of travel (it's easy to reconfigure the new z-axis for 6" travel) that doesn't sag when power is off, but that sticks out from the gantry a total of 1/4" more than the original z-axis did. And my Suckit dust shoe works with the new z-axis and still lines up with the router collet after slightly adjusting the dust shoe magnet plate.

What is left to do is finding a new way to mount my old x-axis cable chain/guide and a new way to mount the z-axis limit switch. And I'll probably re-drill the z-axis slider plate to attach the router mount in the middle of the plate instead of the bottom of the plate and then cut off the bottom of the slider plate. That will give me the exact same z-axis range and reach, but will allow just a little more clearance from the top of workpieces to the bottom of the z-axis.

And here's what it looks like (sorry for huge pictures - right-click and view image if you want to see one of them without scrolling):
top s.jpg
top s.jpg (612.62 KiB) Viewed 1599 times
quarter inch more stick out s.jpg
quarter inch more stick out s.jpg (634.8 KiB) Viewed 1599 times

zamp
Posts: 44
Joined: Fri Jan 02, 2015 8:54 pm

Re: New SO3 Z-Axis

Post by zamp » Sun Jul 23, 2017 11:41 pm

I went ahead and attached the router mount to the middle of the slider plate instead of the bottom, and then cut off the bottom. At the bottom of the z-axis you can see where I filed some channels to allow the router mount screw heads to pass by the bottom plate. WIth the router mount attached in the middle instead of at the former bottom of the slider plate, there is no need for these channels to be filed out as the screw heads no longer need to pass by the bottom plate.
cut off.jpg
cut off.jpg (80.34 KiB) Viewed 1595 times
With the router mount in the middle, the router can still reach all the way down to the spoil board. The whole slider plate and mount can be easily removed and flipped upside down so that the router could be attached higher on the z-axis.
reach.jpg
reach.jpg (44.93 KiB) Viewed 1595 times

Dan_the_Chemist
Posts: 76
Joined: Thu Nov 10, 2016 2:52 am

Re: New SO3 Z-Axis

Post by Dan_the_Chemist » Mon Jul 24, 2017 1:30 am

I am looking forward to a use-report in a week or two after you've had a chance to use it (and challenge it). I was in the middle of a 3 hour job this weekend when the Z-axis belt slipped ... :cry: Perhaps I should look on this as a benefit because if it wasn't for occasions like this, I'd never bother to learn how to edit raw G-code.... :roll: The Z-axis is the only one to give me troubles. If you get good results, I'll be upgrading.

fiero1
Posts: 2
Joined: Fri Jul 28, 2017 1:59 pm

Re: New SO3 Z-Axis

Post by fiero1 » Fri Jul 28, 2017 2:55 pm

This is great. I was just looking at this same setup on ebay. So I did some googling and it brought me here> I mounted a 2.2 kw spindle on me SO3 xxl and my z belt just broke :( I really think
I just need to do this exact thing. Have you got it up and running and calibrated yet? Ray

zamp
Posts: 44
Joined: Fri Jan 02, 2015 8:54 pm

Re: New SO3 Z-Axis

Post by zamp » Fri Jul 28, 2017 3:56 pm

I've reworked the x-axis cable chain mount and reattached the cable chain, fabricated a new z-axis limit switch mount and mounted a new (different style) limit switch, and rerouted the cable for the z-axis limit switch. Tonight or this weekend I'll solder the cable to the z limit switch and calibrate everything.

I had to replace my Carbide Motion controller as that seems to have died in the last half year while the SO3 was sitting unused, so I'll also find out if I saved my most recent GRBL config and calibrating only the Z, or if I'll also be calibrating X and Y.

I'm sure I'll get it all calibrated and running this weekend, and I'll also try a test cut, but probably nothing big.

I'll also push and pull on the z (using the router collet as a handle to push and pull) to see if I get much deflection of the Z axis or not. I can definitely get some minor deflection when I push or pull at the very top of the new Z-axis assembly, but that is very far above where the z slide plate and router itself will be when I'm cutting.

I did already use the stepper motor to run the Z up and down, and that seems to be running smoothly. And of course when I power off the SO3, the new Z and router just stay right where they are.

I'll write more once I've checked Z-axis deflection (in the Y or front to back direction) and done a test cut.

I'm afraid I've not made much use of the SO3 up until now as I often seemed to lose steps on Z because of the springs. I occasionally lost steps before I replaced the original Z plate with the newer Carbide3D Z plate, but the different spring mounting points for the new Z plate (and/or maybe the Suckit mounts?) seemed to cause lost Z steps far more often. I've always lost too many Z steps to feel confident enough to run a long job. I know the springs work out OK for almost everyone, but they never did for me. The X and Z axes have always worked great for me on the SO3, and I hope I will finally have Z working as well as X and Y.

Dan_the_Chemist
Posts: 76
Joined: Thu Nov 10, 2016 2:52 am

Re: New SO3 Z-Axis

Post by Dan_the_Chemist » Fri Jul 28, 2017 4:17 pm

zamp wrote: I'm afraid I've not made much use of the SO3 up until now as I often seemed to lose steps on Z because of the springs. I occasionally lost steps before I replaced the original Z plate with the newer Carbide3D Z plate, but the different spring mounting points for the new Z plate (and/or maybe the Suckit mounts?) seemed to cause lost Z steps far more often. I've always lost too many Z steps to feel confident enough to run a long job. I know the springs work out OK for almost everyone, but they never did for me. The X and Z axes have always worked great for me on the SO3, and I hope I will finally have Z working as well as X and Y.
My experience exactly. X and Y are great. Love em. Z causes me a lot of headaches between the springs and belt slip. I removed one of the springs and that helped a lot. The other thing I do is to break long jobs into smaller jobs. That way I can restart easily if I have Z-slip... I also take advantage of tool-changes to readjust the router in the holder to keep the Z-axis in the sweet-spot of the first half of the spring travel.

Grumpa
Posts: 2
Joined: Thu Jan 05, 2017 2:12 am

Re: New SO3 Z-Axis

Post by Grumpa » Fri Jul 28, 2017 4:48 pm

I also take advantage of tool-changes to readjust the router in the holder to keep the Z-axis in the sweet-spot of the first half of the spring travel.
Hi Dan, what do you mean?? Is there a router height in the mount that depends on location of the Z axis plate like trying to keep the z axis centered on the rails vertically???? Thanks Jude

Dan_the_Chemist
Posts: 76
Joined: Thu Nov 10, 2016 2:52 am

Re: New SO3 Z-Axis

Post by Dan_the_Chemist » Fri Jul 28, 2017 7:29 pm

I find that the Z axis seems to work best when I avoid the bottom 1/4 of the travel ... Since the lifting force on the spring is proportional to the distance traveled... when the Z axis in in the bottom 1/4 of it's travel the spring has 3-4 times the lifting force as it does when the Z axis is in the top 1/4 and this seems to increase the likelihood that a Z-axis skip might occur.

So lets say I am switching from a long router bit to a much shorter one ... say a 3/8" roughing mill with 1.25" of flutes and a 1.5" shank to a 1/16" ball with 0.4" of flutes and 1" of shank. With the 3/8" I will have perhaps 1.75" inches of stick-out past the collett, while I can't get much more than 0.7" with the smaller bit. That means that I would have to adjust the Z = 0 down by at least 1.05" to get the same depth of cut. That would increase spring tension, and increase the possibility of skip.

So, instead, IF the new tool is substantially shorter than the old tool, then I change router height in the holder when changing tools. I put a 123 block on the wasteboard, jog the router over to the 123 block, and drop the Z slowly until the bit just barely captures a sheet of paper... Then I note the Z value (just for grins and chuckles, really - you won't actually need it). Then I pull out the old bit, and put in the new bit. THEN I release the tension on the large aluminum block that holds the router and I allow the router to descend until the router bit touches the 123 block (with paper). Tighten up the holder and Bob's your Uncle. The new bit is at the same height as the old one, and I don't have to adjust the Z axis values in the code.

It would be easier to adjust it in the G-code sender, and it would avoid possible XY shifts as the router is loosened and tightened, but right now I want to avoid the bottom 1/4 of Z axis travel until I get the slippage problem solved. This might be a unusual problem since I am using stock up to 3.25" thick, and have very large Z differences across the work.

zamp
Posts: 44
Joined: Fri Jan 02, 2015 8:54 pm

Re: New SO3 Z-Axis

Post by zamp » Sat Jul 29, 2017 12:04 am

I also found that with springs and a Z limit switch, having the switch trip when Z was almost at the top allowed me to home OK even though the springs were no longer supporting the router at all (they were actually slightly compressed at the top of the Z travel), but as soon as the z started down it would drop like a rock until the springs started taking up some load. When I ran into that, I ended up having to lower the Z limit switch as much as possible and then bend the switch's leaf spring down. I lost about a half inch of travel at the top of the Z's range, but I was once again able to home the Z without immediately losing steps. But I was annoyed by the loss of range and knew I'd still lose steps if I tried to cut too deep with, so I ended up just shutting the SO3 down for months until I could figure out how to build a Z with leadscrew and no springs. I finally lucked out and could buy and modify the Z assembly instead building the whole thing from scratch.

I'll write again as soon as I've calibrated and cut with the new Z.

zamp
Posts: 44
Joined: Fri Jan 02, 2015 8:54 pm

Re: New SO3 Z-Axis

Post by zamp » Sun Jul 30, 2017 9:45 pm

I got the SO3 with its new eBay leadscrew z-axis assembly calibrated, but I've not had time to draw anything for a test cut. It may be a few more days before I get time to do a test cut.

A change with this new z-axis is that the z-stepper has to turn a lot more to make the z-axis move. It now takes 1012.5 z steps to move 1 mm on the new z. Compare that to about 40 steps per mm for the x, y, and old z. If I try to move the new z much faster than about 500 mm/min, then the z moves pretty roughly and I think I'll lose steps. At 500 mm/min the z movement is nice and smooth. I didn't actually try the z anywhere between 500 mm/min and 550 mm/min but 500 was smooth and 550 was rough. I might be able to push the z just a little but faster. than 500 mm/min.

Following are my GRBL 1.1f settings after calibration. My x travel limit is smaller than on a standard SO3 because the mounts for the Suckit dust plate limit x travel a bit. But you can see my Z travel at 112mm is larger than a standard belt-drive z. I would have a little more z travel than that, but I left a cushion at the top and bottom of the z.

Code: Select all

(Grbl 1.1f)     
$0=10           (step pulse, usec)
$1=255          (step idle delay, msec)
$2=0            (step port invert mask)
$3=6            (dir port invert mask)
$4=0            (step enable invert, bool)
$5=0            (limit pins invert, bool)
$6=0            (probe pin invert, bool)
$10=255         ($status report mask)
$11=0.020       (junction deviation, mm)
$12=0.010       (arc tolerance, mm)
$13=0           (report inches, bool)
$20=0           (soft limits, bool)
$21=1           (hard limits, bool)
$22=1           (homing cycle, bool)
$23=0           (homing dir invert mask)
$24=100.000     (homing feed, mm/min)
$25=1000.000    (homing seek, mm/min)
$26=25          (homing debounce, msec)
$27=4.000       (homing pull-off, mm)
$30=1000        (max spindle speed, RPM)
$31=0           (min spindle speed, RPM)
$32=0           (laser mode, Boolean)
$100=39.905     (x, step/mm)
$101=40.000     (y, step/mm)
$102=1012.500   (z, step/mm)
$110=3000.000   (x max rate, mm/min)
$111=3000.000   (y max rate, mm/min)
(note: $112 z max rate much slower than x and y rate)
$112=500.000    (z max rate, mm/min)
$120=400.000    (x accel, mm/sec^2)
$121=400.000    (y accel, mm/sec^2)
$122=400.000    (z accel, mm/sec^2)
$130=386.000    (x max travel, mm)
$131=425.000    (y max travel, mm)
$132=119.000    (z max travel, mm)
I'll report back when I finally get a chance to do a test cut.

Post Reply