[smallblock]

On my machine, I found the Y coupler got loose after a while, so I swapped out the factory set screws with some top quality dog point units that were about 5mm longer and put a jam nut on each one- since then they have not loosened up. On the X, if the belts and pulleys are tight, the next common source of lash is the thrust bearing on the tailstock end. Pull off the rubber dust cover and stand at the end of the machine, grab the cross slide table with a hand on each end and push and pull toward and away from you. Watch the thrust bearing for any back and forth motion- if you have any, then tension the nut. Those Chinese ball screws are usually good for about 0.001-0.002" lash, but with backlash compensation, you can get it right on- here is a video showing the machine travelling 10" and then coming back 10" to zero.

[n1tr0]

The ball screws are metric- very close to 5 TPI, but not exactly.

Being metric, you can use the screw's lead and steps/rev to get some exact values; an inch has been exactly 25.4mm since 1959.
It's easier to have metric hardware setup in Mach3 with metric as the default units, then steps/mm become super simple, (y)400, (x)800, and (z)1600 for the 3 axis and setting the native resolution for each axis in the cam software get's really easy as well; 0.0025, 0.00125, and 0.000625mm. With backlash compensation I was able to get mine to travel out and back to the original location just fine, but when machining with backlash compensation enabled the machine makes some horrible clunking noises. It could be that bearing end-support you mentioned, so I'll give that a shot next.
- updated - tightened end-supports, both x & y were loose, tightened finger tight, x ~+20 degrees, y ~+40 degrees.


I'd definitely like to give the lift-struts & z-pulley upgrade a shot while I'm getting things squared up. If anyone has the actual belt info (pitch and length) or a part# for a pulley they've used, can you please post it? Pictures of how you've got your struts mounted would also be handy.

If anyone has already upgraded their x-axis pulley to go 1:1, some part#'s or details would be handy. It'd be nice to order up the parts without stopping to pull it apart & count the teeth. It looks to be using a standard 3/8" wide 190 tooth XL belt. The stepper shaft is the standard 1/2", but I can't quite get my calipers in there to verify the shaft size on the screw side of things. I might just replace both pulleys and the belt at once using a lower-backlash solution like the GT or Powergrip belt profiles.
- update - When tightening x-axis support bearing I was able to get a good hold on the screw and work the stepper back & forth. It looks like a great deal of the backlash is due to the belt backlash (6 full steps = 9 degrees = 0.0625mm or 0.00246") so I'll definitely be switching belt & pulleys. It looks like the shaft for the screw side pulley is between 9/16" & 15mm? If anyone has theirs apart and can confirm the exact size I'd appreciate it.

The last big outstanding hardware concern is the spindle. If anyone has some details on the bearings and the rebuild operation, I'd like to start researching a fix or upgrade for the bearings (or whole spindle assembly) before they cook themselves. Thanks.

[smallblock]

but when machining with backlash compensation enabled the machine makes some horrible clunking noises. It could be that bearing end-support you mentioned, so I'll give that a shot next.
[B]-
If anyone has already upgraded their x-axis pulley to go 1:1, some part#'s or details would be handy. It'd be nice to order up the parts without stopping to pull it apart & count the teeth. It looks to be using a standard 3/8" wide 190 tooth XL belt. The stepper shaft is the standard 1/2", but I can't quite get my calipers in there to verify the shaft size on the screw side of things. I might just replace both pulleys and the belt at once using a lower-backlash solution like the GT or Powergrip belt profiles.
- update - When tightening x-axis support bearing I was able to get a good hold on the screw and work the stepper back & forth. It looks like a great deal of the backlash is due to the belt backlash (6 full steps = 9 degrees = 0.0625mm or 0.00246") so I'll definitely be switching belt & pulleys. It looks like the shaft for the screw side pulley is between 9/16" & 15mm? If anyone has theirs apart and can confirm the exact size I'd appreciate it.

Thanks.

On the clunking sound, if all your handles are tight etc. it is probably due to your settings in backlash compensation and shuttle wheel acceleration. Reduce backlash max speed to 50% and set shuttle wheel to about 0.01. In your motion mode, set to exact stop rather than CV that will allow the axis to decelerate to a stop rather than stopping instantly from full speed. On your X axis belt, be sure it is adjusted properly to remove any lash in the belt- it should be tight with no deflection, but not extremely tight so it overloads the stepper bearings. I got a new motor pulley and belt from JT when I changed to 1-1 ratio.

[n1tr0]

I think the XL belt on my machine isn't actually matched to the pulleys; right pitch, wrong tooth profile. It's tensioned ok, with no load there's no slip, but as soon as it's loaded up a bit, the belt shifts on the pulleys. That's cool that JT hooked you up with a pulley and belt, at this point I'd be happy with technical documentation on the machine and a spindle that doesn't suck. I'm not sure how accurate the FLIR is as I haven't taken any direct measurements with a thermocouple, but after a bit of machining at ~4500 rpm it's too hot to touch. I can definitely understand why the initial units had air cooling to the spindle. Preload on the spindle is at the bare minimum to keep the pulley from rattling so I suspect it's either the quality of the bearings or yet another problem with how it was assembled. I'm running a 2mm end mill so I doubt it's the loads.
Attachment 329152Attachment 329154

[n1tr0]

So I'm in the midst of building a new dedicated PC to replace the defective used POS that shipped with my Mill-Turn. I figured it was a good time to get an ESS so I could move beyond Windows XP and put the computer on the network so I'm not having to constantly transfer files via sneaker-net. I know smallblock had some issues trying to do threading with the ESS (and saw some posts on other forums trying to rectify it), but I'm willing to potentially give up threading for the moment to have a functioning mill.
Not to ramble on too much, but I found this little gem of info in the ESS support pages. It's a shame Shoptask only uses this forum to push new products and doesn't offer any similar support :P

We strongly recommended (January 1st, 2016) that you use Mach3 version 3.043.062 (click to download). The .062 release is the lock down version that works well, fixing stability issues experienced by people running version .066.

The latest release was Mach3 version 3.043.066. However, a number of people have had stability issues with their systems when using version .066. The .066 version has had these issues for some users:

MPG issues (we have seen this many times)
Pendent issues (we have seen this many times)
Feed hold issues
Homing issues
Macros not being fully run
VB buttons being ignored
If you are using v.066, we will tell you to install v.062 as the first step before we do anything else in term of support.

I can attest to having seen all of those gremlins (usually destroying bits and parts) in .066.

If anyone else has added one of the inexpensive USB pendants (they say XHC on the hand-wheel), I also found the drivers disc that ships with them is seriously outdated. You can download the latest Mach3 drivers here. They don't seem to have any Mach4 support 'tho, so if you haven't bought one yet, I'd recommend spending a bit more and getting one from vistaCNC instead as they have drivers for Mach3 & 4 for all their pendants.

[instructor37]

Nitro,
A couple of things to consider that JT warned us about early on-
1. Mach 3 is very sensitive to programs running in the background and internet running while Mach 3 is working. You may find issues with lost steps that are hard to trace, that is why he recommends keeping your PC dedicated to Mach 3 only.
2. The ESS as far as I know cannot do threading, so you could install it on the original computer, and just have the original XML files as well, so you could do threading through the parallel port cable. However if you go for a new computer with no parallel port, then threading will not be possible. Here is a link to a long discussion about this on the ESS forum.
Search

[n1tr0]

The original computer is a steaming pile of boot errors so if I can't get threading to work on the new controller, I'll do without, but at least I'll have a better running mill. My friends have CNC mills and routers at home running mach3 without any of the problems I've had (and they're running off tablets and various non-dedicated PC's). It seem's most everyone I know with a home-built or kit-machine is using the ESS. Warp9 does recommend a true multi-core machine (not just hyperthreaded). The good news is that rev 2 of Mach4 is out with some early lathe support, so there's a better chance bugs will be sorted out since Mach3 development has been dead for a couple years now.

I read through Smallblock's posts on the Warp9 site and his problems didn't make a whole lot of sense, messing with potentiometers to get the voltage down into the correct range. The NowForever VFD (like most) should output the required 10V. There shouldn't be any wiring changes required to command the spindle speed with the ESS outputing to the Gecko instead of using the parallel port. A01 and Com provide the +10V and Ground to the Gecko's pins 9 and 7 respectively. Then you use the Gecko's pin8 to provide the analog input for the VFD (I believe it's set to A1N1 by default). Everything should pass back and forth through the ESS fairly transparently, it's just offloading the signal generation/timing for the steppers. I'll sit down and make up a full wiring diagram probably this weekend. I'm not sure I quite followed the problems Smallblock was having that required more effective PID control of the spindle, but that can be offloaded directly to the VFD rather than having Mach3 handle it, then it's just a matter of making sure the latency/jitter isn't a problem. This manual has a very similar setup to the Mill-Turn.

[n1tr0]

[ JT says they actually de-tune the G540 to avoid issues of missed steps. They are capable of running 900 OZ/IN motors, but Shopmaster puts in a current resistor and uses 500 OZ/IN motors.

I just couldn't get over how slow the Mill-Turn is so I did a bit of digging into what's actually shipped with the machine. The Nema34 motors are actually only rated at 1.8 Nm (and 2.0A) which translates to only 255 oz-in. They might be the lowest powered Nema34 steppers ever made.
The Gecko G540 can run any bipolar stepper up to 3.5A. Anaheim Automation (for example) has Nema34's rated at 1700 oz-in at 2.5A when run in the usual bipolar series wiring, but you have to keep in mind that the windings to get that sort of holding torque comes with higher inductance and thus a quicker drop-off in torque at higher rpms... although the torque curve on the 34y307 steppers still shows more torque at 800rpm than the stock steppers have at stall.
Keep in mind the machine only has a 350W power supply.

[02corvette]

So what are you saying. False advertising? I'm just curious have you made anything on yours yet or is it the machine just "another project" added to the list. Are any of your problems with covered under warranty?

[JTJT]

Nitro,
That number does not match our numbers. Check your other motor numbers. If the factory made a mistake, we can arrange to get the proper motors shipped to you.

[n1tr0]

All 3 motors are the same. Supposedly this machine was "rigorously inspected" and "carries the ISO9001 certification", yet the mountain of problems I've discovered since it arrived say otherwise. Had the machine actually come with any real documentation I might have been able to do a proper inspection on delivery and immediately return it, instead there's just new problems to discover every time I think it's ready to go.
Attachment 331632
Fair warning to anyone else wanting to check your motors, You may not be able to put the handles back on because if your machine is like mine, most of the set screws are cross-threaded in place and the pot metal is going to give way before the threads on the screw.
Attachment 331634

[Norton_BMW]

All 3 motors are the same. Supposedly this machine was "rigorously inspected" and "carries the ISO9001 certification", yet the mountain of problems I've discovered since it arrived say otherwise. Had the machine actually come with any real documentation I might have been able to do a proper inspection on delivery and immediately return it, instead there's just new problems to discover every time I think it's ready to go.
Attachment 331632
Fair warning to anyone else wanting to check your motors, You may not be able to put the handles back on because if your machine is like mine, most of the set screws are cross-threaded in place and the pot metal is going to give way before the threads on the screw.
Attachment 331634

All my three motors are the same as well. and two of my three handwheels needed to be helicoiled.
Attachment 332444

[JTJT]

Nitro,
That number does not match our numbers. Check your other motor numbers. If the factory made a mistake, we can arrange to get the proper motors shipped to you.

UPDATE-
We sent a message to the factory about the motors, and they contacted the factory where they get the steppers. The stepper factory says they never supplied that model of motor to our factory. They say it could only have been a shipping error or even a mis-labeling of the motor. In any event, our factory will supply the correct motors for anyone wanting to change out. Regarding those handles, the set screw pictured is obviously the wrong style- to secure the handle to the flat on the shaft , it should be a flat point, not a cone point, as the cone style will simply wear away the tip and constantly loosen up, causing a tendency to over tighten it in an attempt to stop it from loosening. Helicoil is a complex fix, a simpler method is just re-tap to the next size up. We are happy to do the repair for anyone with stripped threads. For anyone out there with the motor issue, please contact us direct with your serial # so we can try to see if there was a specific batch of machines we can trace.

[Norton_BMW]

UPDATE-
We sent a message to the factory about the motors, and they contacted the factory where they get the steppers. The stepper factory says they never supplied that model of motor to our factory. They say it could only have been a shipping error or even a mis-labeling of the motor. In any event, our factory will supply the correct motors for anyone wanting to change out. Regarding those handles, the set screw pictured is obviously the wrong style- to secure the handle to the flat on the shaft , it should be a flat point, not a cone point, as the cone style will simply wear away the tip and constantly loosen up, causing a tendency to over tighten it in an attempt to stop it from loosening. Helicoil is a complex fix, a simpler method is just re-tap to the next size up. We are happy to do the repair for anyone with stripped threads. For anyone out there with the motor issue, please contact us direct with your serial # so we can try to see if there was a specific batch of machines we can trace.

Measured the stepper torque with a Snapon digital torque wrench with peak hold and it suggests that N1tr0's observation is correct that the motors are the model indicated on the label.

[JTJT]

Thanks for the rapid response and test- just as a caution, remember that the G540 Gecko drive drops the current 70% at idle on the motors to prevent overheating, so you need to do the test in operation mode.

[n1tr0]

I finally got around to pulling the sheet metal off my machine so I could troubleshoot the angled wear showing up on my z-belt. That mostly appears to just be some bad assembly from the factory with pulleys spaced at different heights from each other.

While I was at it, I figured this would be a good time to try and tram the spindle with respect to the table to get rid of the saw-tooth cuts. I mounted my dial-gauge in the spindle so it could be rotated to different points giving me about an 8.5" circle. It'd be great if someone could check my math here. Side-to-side I'm measuring a bit over 0.005" on that 8.5" pitch circle, given that the screws are ~35" apart, that's a bit over 0.02" height (or about 0.5mm) difference left to right. The ball-screws have a 5mm lead (5mm/rev) and a 72 tooth pulley, so each tooth represents about 0.07mm in height, meaning I need to adjust the belt on one pulley ~7 teeth.

The sucky part is that it turns out the easy to adjust side-to-side tram was relatively good compared to the front-to-back tram which is out about 0.016" over the same 8.5"...

[n1tr0]

So another day of peeling back the layers of this onion...

@cbud, you should contact JT about your stepper motors. From your pictures it looks like you have the same under-spec'd motors Norton_BMW and I have. I did the torque wrench test (coupled my torque wrench to the output shaft and tried stepping the motor) and the best I could get was ~13ft-lbs or 208oz-in. It's a calibrated mechanical torque wrench so that should be within 1-2ft-lbs. I swapped out the factory motors for some known-good AA 637oz-in motors and the difference is off-the-chart. I used to be able to easily overpower the motors with the hand-crank and now they feel rock-solid.
Attachment 332684

Moving on to tramming the spindle. Pulling it out of the machine, it's pretty obvious how the factory attempted to tram it out. So far rotating it other directions has only made the situation worse, so I'll have to wait until I can get some shim stock in.
Attachment 332688

When I was up adjusting the belt on the z-axis to tram the beam, I got looking at the upper structure and boy does it look out of square! It looks like they just didn't have things aligned when it was welded up. I'm pretty much at a loss on how to fix this.
Attachment 332690Attachment 332692

[Norton_BMW]

I finally got around to pulling the sheet metal off my machine so I could troubleshoot the angled wear showing up on my z-belt. That mostly appears to just be some bad assembly from the factory with pulleys spaced at different heights from each other.

While I was at it, I figured this would be a good time to try and tram the spindle with respect to the table to get rid of the saw-tooth cuts. I mounted my dial-gauge in the spindle so it could be rotated to different points giving me about an 8.5" circle. It'd be great if someone could check my math here. Side-to-side I'm measuring a bit over 0.005" on that 8.5" pitch circle, given that the screws are ~35" apart, that's a bit over 0.02" height (or about 0.5mm) difference left to right. The ball-screws have a 5mm lead (5mm/rev) and a 72 tooth pulley, so each tooth represents about 0.07mm in height, meaning I need to adjust the belt on one pulley ~7 teeth.

The sucky part is that it turns out the easy to adjust side-to-side tram was relatively good compared to the front-to-back tram which is out about 0.016" over the same 8.5"...

When I First got my machine in Early Nov 2015, my Z axis stepper motor was at a severe angle, plus the Z axis belt had a hole drilled in it by the factory causing it to skip. The pulleys were also misaligned. Shopmaster sent a new belt which eliminated the skipping.

Attachment 332726

Attachment 332728
However; much of the severe misalignment of the pulleys ( they were at different offset heights) was caused by the shoulder on the shouldered retaining bolt being too short. This also caused the non angular contact bearings in the affected pulleys to be damaged. I had to replace the bearings and machine a proper shoulder spacer and add washers to make everything line up.

Attachment 332730

Notice that the pulley on the right has the shoulder recessed.

[Norton_BMW]

Regarding the belt:

My Z axis stepper mount was way off square and caused a flopping of the belt as it changed directions. That might be causing your uneven belt wear. The angle iron used in the top frame is not square at all.

I first marked off the tooth spacing timing marks on both large ball screw wheels with a Sharpie to align with the belt so I didn't throw the bridge out of tram - if it ever was in tram from the factory. I stacked spacer washers in the top holes of the mount in trial/error until I got the Z axis stepper pulley to have even tension on the top and bottom of the belt. I made a small wooden V-block to bear against the tensioner pulley wheel and used a woodworking bar clamp to squeeze the adjuster pulley toward the rear angle iron to set the tension while I tightened the bolt on the tensioner puller.

Amazingly enough I didn't need a pin spanner wrench even though those pulleys have pin holes. I was able to just create enough friction force using a standard box wrench with a twisting / shearing motion on the bolt head to stop the pulley from spinning while I tightened / torqued the bolt. The open ended fork end or crescent wrench might not work in doing this. I think I needed the box end of a box wrench.

I'll eventually need to make another Z axis stepper mount with an angle adjustment wedge and screw if I upgrade the BoB, Z driver, and Z stepper size - but that's way down the line.

Cbud;
I had all the same issues with my Z-Axis. Check my other post about the shouldered bolts on the idler pulleys. Your toothed idlers may have shouldered bolts that have shoulders that are not long enough and cause the bearings to be side-loaded and fail.

[Norton_BMW]

@cbud, @n1tr0;
When I checked the torque with my calibrated SnapOn digital Torque wrench, I did two tests. Stall torque and moving torque. I measured the moving torque using a motorcycle friction steering damper.assembly parts. The stall torque reading I got was 203 oz-in and the torque at 100 Rpm was about 240 oz in. The steering dampner was not a constant load because of slight variances in the friction disc, but it was good enough to show how weak the supplied motors were.