Just ordered the last melon (Torus)

[Titaniumboy]

That would be great! Thanks for the help.

[Novakon]

Hello,

Your right, the parts break down on the web side would not help your here. I thought that you were referring to the Torus PRO.

Normally, the ball nut is not removed from the screw and the motor mount not removed from the table. The motor mount has 2 tapered pins for alignment and need to used to pin it into position when you put it back on. I would do this first before assembly. The ball nut may loose some of the ball bearings inside so you will have to closely inspect this. If necessary, I have spare ball screw from the Pulsar that should be the same.

The arrangement of the parts are correct except for "A". this will go between nuts "B" and "C". "A" is the anti-spin device where you bend the tab into the nut slot after you have tightened the assembly. Just snug up the nuts to remove the play and don't over torque them. The bearings need to be placed opposing each other as you said, otherwise the bearing cage will pull away and destroy the bearing also.

Clamping cover "D" will lock the outer bearing sleeves together with the 4 bolts. Use equal pressure between bolts to ensure the bearing don't bind. There will be about 1mm gap between the clamp and the casting when your done. This locks the outer sleeves of the bearings. This is the clamp that pinches the bearings together and you don't want to overtighten or you can damage the ball bearings.

The normal way to remove the table assembly was to:
1: On the right side of the table, remove the ball screw support bearing cap.
2: Remove the motor with coupling.
3: Loosen and remove the X GIBS.
4: Remove/loosen the bearing cover in the motor mount.
5: Slide the table to the right to push the bearing assembly out of the motor mount as an assembly. If the bearing are a tight fit, you may need a little persuasion with a block of wood and hammer.
6: Remove the two locking nuts and bearings from the ball screw.
7: Slide the table to the left to remove it.

The ball screw assembly will still be attached to the saddle and then can be removed from the mount by removing the 4 bolts on the ball nut. The nut should not be removed from the screw or you may loose the ball bearing inside.

The process can be reversed to reassemble.

Hopefully this helps in the assembly for you.

John

[Titaniumboy]

I have successfully installed all of the components for the X-Axis Motor Mount back to the mill during the last week. I'm happy to report that, after a couple of tries, that I have the backlash somewhere between 0.5 and 1.0 thou.

The pic below shows the correct order of the X-axis parts. John was correct in noting that anti-spin washer A goes between the nuts B and C. I had incorrectly shown the order of the two bearings (F and H )and the two spacers (E and G) in the original picture, even though my notes clearly showed the correct way and I had misread those notes.

Inner race spacer G goes on the end of the ballscrew. The end of the ballscrew is then placed through the motor mount before the inner bearing H, outer race spacer E, and outer bearing F are installed. I knew this to be the case and still installed the two bearings and spacer onto the ballscrew at first without having the ballscrew through the motor mount. Thanks goodness for 3-jaw bearing pullers.





I wasn't completely confident in the order that the components needed to be installed, so I took the opportunity to measure all the parts and then drew them up in Solidworks while everything was still in pieces. Good thing that I did because this is when I discovered that I had goofed up in the original picture on where the inner and outer spacers were located with respect to the two bearings. I modeled everything except the anti-spin nut A.

The measurements and subsequent 3D model illustrates the small gap between the Clamping Cover D and the motor mount casting that John mentioned above. The model even highlighted that the outer nut B would have only a very small amount of thread available on the ballscrew. I initially thought I had measured something wrong when this became apparent when drawing the model, but during assembly I found that the outer nut B really did have only about a thread or two to work with. I probably should have shaved a millimeter or two off of the inner spacer G so that the outer nut B would have more threads to work with, but I decided to skip it at this point.

[Titaniumboy]

The picture above shows how my X-axis ballscrew was being lubricated as delivered from the factory. I thought that the oil being merely dripped on the ballscrew outside of the ballscrew nut was not a good idea. And the Y-axis and Z-axis had their lube lines attached to the ballscrew nut. The problem for the factory (and me) was that there is simply no room to attach a lube line to the X-axis ballscrew nut.

I decided that the solution was to do some milling on the ballscrew nut in order to make room for a lube line. You know how they say that the road to hell is paved with good intentions? Or that no good deed goes unpunished? I was about to learn these two lessons the hard way.

I first had to remove the X-axis ballscrew nut from the ballscrew mount. The next step was to remove the ballscrew nut from the ballscrew. So i made the tool below so I could safely remove the ballnut from the ballscrew without losing any of the little balls. The ballscrew removal tool is simply a scrap piece of aluminum turned to a diameter of 0.652" on the right side (ignore the left side as that was for a different project).




The ballscrew removal tool worked wonderfully and I successfully was able to remove the ballscrew from the ballscrew nut. Unfortunately, somewhat later and due only to my clumsiness, I accidentally allowed the ballscrew removal tool to slip out of the ballscrew nut and let loose the all the little balls. Boy, do they bounce good. The silver lining is that I was able to thoroughly clean the inside of the ballscrew nut, and also was able to learn how to reinstall the wayward balls. The secret, of course, is grease. And a good set of magnifying lenses.

I set to the task of milling the ballscrew nut and learned very quickly that the nut is hardened steel. So I was forced to pull out the Dremel and I set to work grinding on the nut. This was a slow process and after a lot of work, this is what I ended up with. There is almost no clearance under my table, so the groove that I ground in the nut is deep enough for a 3/32" diameter tube to fit flush.




The next step was to machine a custom oil fitting out of a M6x1.0 bolt. The bolt head was mostly cut away for clearance from the table, and then milled to length. Then I drilled a 3/32" hole in the center from the bottom, and then another 3/32" hole in the side.




And here is the new oil fitting with the 3/32" aluminum tubing test fitted.




Here is the new oil line to the ballscrew nut. I used some 4mm tubing that John had sent me and I opened up one end with a 3/32" drill for about a 1/4" or so. The 3/32" tubing fits snugly inside the 4mm tubing.




Even with all of that work, the amount of clearance is still very tight.




I spent a stupid amount of time trying to dial in how much oil was being sent through the new oil line. Both of my spare #3 oil metering units sent 3-4 drops of oil to the ballnut with each pull of the manual oiler, which I thought was too little oil. Meanwhile my spare #4 oil metering unit sent 26 drops to the ballnut which is way too much. Since a #4 metering unit is only supposed to be twice the flow of #3 metering units, this didn't make any sense. So I finally broke down and bought a pair of oil metering units - a #3 and a #4 - off of eBay for about $14. The new #3 metering unit flows about 10 drops of oil per pull of the manual oiler and that is perfect. I guess my two spare #3 metering units are partially plugged.

[TOTALLYRC]

Nice write up of what you did.


Mike

[Titaniumboy]

Nice write up of what you did.

Mike

Mike, Thanks for the nice words!


So about three weeks ago I had posted this statement below after a hiatus of three years:

I have successfully installed all of the components for the X-Axis Motor Mount back to the mill during the last week. I'm happy to report that, after a couple of tries, that I have the backlash somewhere between 0.5 and 1.0 thou.

So not being smart enough to leave well enough alone, I went ahead and measured the backlash on the Y-axis and Z-axis also. Y-axis came in with a backlash of 2 thou and the Z-axis came in with a backlash of 4 thou. So my Y is not so great, but the Z is definitely stinking up the joint. Unfortunately I had not measured the Y-axis and Z-axis backlash back when the mill first arrived.

Since I had improved my initial reassembled X-axis backlash from around 3 thou to under 1 thou by merely tightening the bearing retainer plate screws a little, I thought that the same thing might be the cause of my Z-axis backlash. Or maybe just a loose bearing nut perhaps. Both easy fixes in my mind since I had the recent X-axis success under my belt.

So I dug into the Z-axis and proceeded onward to disassembling it. I found the arrangement of the Z-axis a little different than the X-axis with the main differences being the location of the bearing inner race spacer and that there is no locking bearing nut. The picture below shows the bearing assembly after being rotated out of the bearing housing. The bearing retainer plate is the only thing missing from this picture.



I found the bearing nut was pretty mangled. I went ahead and cleaned up the bearing nut. I also cleaned up the bearing retainer plate getting rid of burrs and slightly enlarging the interior diameter so that it could be slid over the bearing nut. Not that it makes a bit of difference because the locking washer won't fit through the bearing retainer plate hole. So you have to install the bearing retainer plate BEFORE you install the locking washer and the bearing nut.

The Solidworks model on the left shows the arrangement of the Z-axis parts more clearly than the picture on the right.



So after cleaning everything up on the Z-axis and reassembling all the bits and bobs, paying special attention to the bearing nut tightness and the bearing plate tightness, I was able to measure the Z-axis backlash at 4 thou. All that work and the backlash was exactly the same as before.

[Titaniumboy]

All that work and the backlash was exactly the same as before.

So I set out to trace down where the four thou of backlash was coming from.

Before starting the search for the backlash, I tried adjusting the Z-axis gib. Loosening gib by a 1/8 turn of the nut produced a higher backlash at five thou. Going back to the original gib location and then tightening the gib by a 1/8 turn of the nut also produced a higher backlash of five thou. So the original gib location was set nicely from the factory.

I first looked into whether there was any lost motion occurring between the stepper motor and the ballscrew. In reality you can only look at the motor coupling versus the bearing nut. I repeated went up and down by one thou and every time the motor coupling looked like it was moving exactly in sync with the bearing nut. Verdict: No Lost Motion here.

Next I looked into whether there might be some looseness in the bearing nut (which compresses the bearing inner races), or the bearing retainer plate (which compresses the bearing outer races). I figured that if there was play in either of these two places, then I should be able to observe the entire ballscrew moving up or down. I set up a dial indicator on the flat bottom of the Z-axis ballscrew as shown below.





The bottom of the ballscrew is not perfectly flat as you go around one complete revolution, but it is pretty close to flat when you are only moving the Z-axis up or down by 6 to 10 thou. Verdict: About 1/2 thou at most of backlash.

The next step of tracing lost motion was to check the ballscrew to ballscrew nut backlash. I set up the dial indicator on top of the ballscrew nut as shown below.





Imagine my surprise when I measured 5 thou of backlash at the ballscrew nut. I'm not sure why I'm measuring five thou of backlash here but only four thou backlash at the head; maybe the indicator wasn't perfectly square to the face of the ballscrew nut?

The ballscrew nut is usually the least likely source of backlash while also being the most expensive to correct. I really thought I was going to find the source of the backlash anywhere but here.

Does anyone see where I might have made a mistake or missed something?

Which one of these options would you choose?

1 - Live with the four thou of Z-axis backlash
2 - Repack the ballscrew nut with oversized bearings
3 - Replace the ballscrew

[kvom]

For 2.5D machining the backlash will be taken up a vast percentage of the time, so as long as you set Z0 with downward motions you can likely ignore it.

[Titaniumboy]

For 2.5D machining the backlash will be taken up a vast percentage of the time, so as long as you set Z0 with downward motions you can likely ignore it.

Very good point. I will definitely be machining with the existing Z-axis backlash. As you point out, the Z backlash is not going to make any difference at all for 2.5D machining. That said, I'm still hoping to get some guidance from John regarding this issue at some point.


While waiting to hear back from John, I decided to dig in a little further and verify that the ballnut really did have four or five thousandth of backlash. After first supporting the head with some wood blocks, I removed the four SHCS holding the ballnut to the L-shape bracket. At this point the head is completely disconnected from the ballscrew and is fully resting on the wood blocks.




Now using Mach3 to move the head in the Z+ direction only results in the ballnut being moved up from the L-shaped bracket.




Here we see the ballnut has moved completely out of the L-shaped bracket.




Getting a dial indicator into position into the cramped confines of the column casting is a bit tricky and not a lot of fun. I can't feel any slop in the ballnut when moving it by hand, and the dial indicator reads zero backlash on the ballnut when it is disconnected from the weight of the head. However, the full backlash appears once again at the ballnut when reattached to the L-shaped bracket and the full weight of the head is being applied.




So the ballnut has zero backlash with zero load being applied, and four or five thousandth backlash with the head attached.

[Titaniumboy]

Success !! I have finally made the mill do some work.





And here is the hokey setup I came up with; a plywood scrap piece held in place by Kant-Twist clamps. I used a mini Sharpie marker in a custom holder I made on the lathe out of delrin.





I used the wizard in Mach3 to create each line of text separately. This little exercise has really driven home the need to finally find a permanent solution for a machine controller because each line of text used around 100 - 120 lines of G-Code, and my demo version of Mach3 has a 500 line limit.

Here are the options as I understand them.

1. Mach3

2. Mach4

3. LinuxCNC without external motion controller

4. LinuxCNC with external motion controller

5. PathPilot with a MESA motion controller

6. UCCNC

7. Centroid Acorn


I'm not enthused about the two Mach options at this point. I suppose Option 3 with the bare LinuxCNC would be the least expensive option at $0, but I am a little intimidated being a Linux virgin.

I am leaning towards Option 4 (LinuxCNC) or Option 5 (PathPilot), with the external motion controller. Because making it harder than the Option 3 bare LinuxCNC is what a masochist would do.

I am used to using real CNC controllers such as FADAL, Hurco and Haas so G41 Cutter Compensation is kind of a deal breaker for me in a machine controller.

Any thoughts? I would definitely need a ton of hand holding if I go with any of the LinuxCNC or PathPilot options.

[CL_MotoTech]

I'd go UCCNC. Setup is a breeze, especially if you have Mach3 working. Though UCCNC hasn't implemented G41/2, which is kind of a shame, but I've gotten used to machining without it. It takes seconds to adjust the tool size in in CAM and post new code. If it's a one off job, then it's no more work than if G41 was available.

The constant problems that people seem to have making PP work on non Tormach machines doesn't seem to be worth it IMO, at least judging by the posts here. I've used LinuxCNC some, but not being able to run my CAD/CAM on my controller is a big turn off for me. I use my controller as a work station just like the PC that's in my office. Acorn looks great too, but since it didn't even exist in its hobby format I was well entrenched before it was an option. That said, I think it's limited to four axis which I currently use, and I hope to add another axis soon.

[kvom]

I use PP, which has G41 that I have used only to dial in a thread mill. I didn't do CAD or CAM at the mill when I had mach3 as I didn't want to do any Win7 updating.

PP work itself has little to do with Linux as your're mostly working in an editor. Very few command-line ops. It did take a while to grok the HAL and INI syntax, for which the online references from LinuxCNC apply. My major problem as I recall was getting the spindle to run. Once I had the correct Mesa firmware I had few other issues.

The main difference for me vs. Mach3 is tool changes. Mach3 allows jogging when stopped at a M6, while PP does not. So I have to Stop the program, and then manually set the next op as the start line. Since I'm not using the Tormach tool holders and collet, I have to set Z0 on each tool change. No change from Mach3 as I didn't use the tool table there either.

I'll bet some helpful soul here could donate their working HAL/ini that would save a lot of work.

[Titaniumboy]

I use PP, which has G41 that I have used only to dial in a thread mill. I didn't do CAD or CAM at the mill when I had mach3 as I didn't want to do any Win7 updating.

PP work itself has little to do with Linux as your're mostly working in an editor. Very few command-line ops. It did take a while to grok the HAL and INI syntax, for which the online references from LinuxCNC apply. My major problem as I recall was getting the spindle to run. Once I had the correct Mesa firmware I had few other issues.

The main difference for me vs. Mach3 is tool changes. Mach3 allows jogging when stopped at a M6, while PP does not. So I have to Stop the program, and then manually set the next op as the start line. Since I'm not using the Tormach tool holders and collet, I have to set Z0 on each tool change. No change from Mach3 as I didn't use the tool table there either.

I'll bet some helpful soul here could donate their working HAL/ini that would save a lot of work.

Thanks for the reply kvom.

You have the NM-200, right? Do you have the built-in Novakon motherboard or a separate computer? I have the Windows 7 motherboard that was installed by Novakon in the electrical cabinet, with the demo Mach3 installed in a solid state hard drive.

My Torus is an odd mashup where the mechanical part is a NM-145 and the electrical part is a Pulsar with the axis motors being steppers and the spindle being a servo.

I think others have suggested the best way to install PathPilot is to get a second solid state hard drive. I am a little confused about what to buy at Tormach. They show Part# 38249 which is a USB stick PathPilot v2.0 Upgrade - Software Backup for $24.49, but I don't see where to buy the original PathPilot software.

I have read of others on the Novakon forum do the PathPilot upgrade over the years, but it always seems like a bit of a struggle. And has anyone been able to install the latest PathPilot onto a a Novakon yet?

[Titaniumboy]

I'd go UCCNC. Setup is a breeze, especially if you have Mach3 working.

I had forgotten that, at one point several years ago, that Novakon had settled on UCCNC UC300ETH with the 5LPT motherboard per post 54. But something must have happened to scuttle that plan.

https://www.cnczone.com/forums/novak...-software.html

I was wondering how it could be that UCCNC was so easy to install. After a bit of research it turns out that UCCNC makes use of the Mach3 XML file to set things up. Very nice.

Though UCCNC hasn't implemented G41/2, which is kind of a shame, but I've gotten used to machining without it. It takes seconds to adjust the tool size in in CAM and post new code. If it's a one off job, then it's no more work than if G41 was available.

Well, looky here on page 59 (pdf warning). UCCNC has been busy. Cutter Compensation is now implemented.

http://www.cncdrive.com/UCCNC/UCCNC_usersmanual.pdf

The constant problems that people seem to have making PP work on non Tormach machines doesn't seem to be worth it IMO, at least judging by the posts here. I've used LinuxCNC some, but not being able to run my CAD/CAM on my controller is a big turn off for me. I use my controller as a work station just like the PC that's in my office. Acorn looks great too, but since it didn't even exist in its hobby format I was well entrenched before it was an option. That said, I think it's limited to four axis which I currently use, and I hope to add another axis soon.

My Windows 7 bare motherboard is so far a dedicated controller. I would be too chicken to run any other software on it, especially now that Windows 7 is now officially a Microsoft orphan.

[Steve303]

Thanks for the reply kvom.

You have the NM-200, right? Do you have the built-in Novakon motherboard or a separate computer? I have the Windows 7 motherboard that was installed by Novakon in the electrical cabinet, with the demo Mach3 installed in a solid state hard drive.

My Torus is an odd mashup where the mechanical part is a NM-145 and the electrical part is a Pulsar with the axis motors being steppers and the spindle being a servo.

I think others have suggested the best way to install PathPilot is to get a second solid state hard drive. I am a little confused about what to buy at Tormach. They show Part# 38249 which is a USB stick PathPilot v2.0 Upgrade - Software Backup for $24.49, but I don't see where to buy the original PathPilot software.

I have read of others on the Novakon forum do the PathPilot upgrade over the years, but it always seems like a bit of a struggle. And has anyone been able to install the latest PathPilot onto a a Novakon yet?

If the BOB in your Torus is the same as the BOB in our Pulsar's you should be able to use the INI, HAL and BIT file for the pulsar with only minor changes for steps and travels. You would need to purchase the USB stick from Tormach, a Mesa card (5i25 PCI or 6i25 PCIE) from Mesa along with a SSD drive for installation. You do not need the original PathPilot software as the USB Stick with v2.0 is a complete fresh install.

Take a look at your BOB and Computer Main Board. Going with PP and a Mesa card allows you to use PathPilot or LinuxCNC. Take a look the ProbeBasic GUI for LinuxCNC. It is very slick. I ordered a new SSD drive to install ProbeBasic on for testing on my Pulsar.

I can't comment on UCCNC as I have never used it. BTW, where has Ray gone?

[CL_MotoTech]

I run Windows 10 but see no reason why you couldn't stick with Windows 7. I stopped Windows 10 updates when it broke Mach 3, and still occasionally use Mach 3, but 99% of my machining is done in UCCNC now. I don't know how Novakon set up their controller but I assume its parallel port and not a dedicated motion controller? With the UC300eth-5lpt there's not reason not to multitask, run various software, etc.. Everybody does it and it's one of the prime advantages of a Windows based controller IMO. I post G-code to a folder on my C drive, load the file in UCCNC, and hit cycle start. No need to have multiple PC's in the shop. And it's nice to take care of email, browse the web, play music, etc..

Importing the Mach3 setup file worked decently for me, but not perfect. I can't recall what the issues were, but I remember having to adjust/enter some settings manually. UCCNC just has different setup options than Mach. That said, UCCNC is way easier to setup than Mach IMO. The setup pages and options are just more intuitive.

Good know that G41 is implemented, I guess I should pay better attention to their release notes.

[kvom]

My NM200 has an external box with all the electronics. The PC connected to the BoB with a short 23? pin cable. For PP I had configured a low cost PC that could accept a Mesa card. This low profile box sits on top of the old box and uses the same cable to connect it to the BoB. To run the mill I power on both boxes. If for some reason the PP box were to fail I could reconnect the cable and the mouse/monitor/keyboard and run Mach3 again. Whether this is possible with the Torus electronics cabinet I can't say.

I haven't upgraded PP from 1.9 as it does everything I need.

[ger21]

After a bit of research it turns out that UCCNC makes use of the Mach3 XML file to set things up. Very nice.

Personally, I do not recommend doing this, because it won't setup everything 100%.
The reason is that Mach3 has settings that UCCNC doesn't have, and UCCNC has settings that Mach3 doesn't have.

Setting up UCCNC is very easy, and is very much like setting up Mach3.


And UCCNC has had G41/G42 in test versions for close to 18 months, maybe longer?

[Titaniumboy]

If the BOB in your Torus is the same as the BOB in our Pulsar's you should be able to use the INI, HAL and BIT file for the pulsar with only minor changes for steps and travels. You would need to purchase the USB stick from Tormach, a Mesa card (5i25 PCI or 6i25 PCIE) from Mesa along with a SSD drive for installation. You do not need the original PathPilot software as the USB Stick with v2.0 is a complete fresh install.

That is great news that the PathPilot USB Stick with V2.0 is a fresh install. That seems like quite a bargain for only $24.99.

My BOB is labeled "Novakon.net Pulsar BOB R0". My computer motherboard, installed by Novakon at the factory, is a Gigabyte GA-H61M-S2PV.

Take a look at your BOB and Computer Main Board. Going with PP and a Mesa card allows you to use PathPilot or LinuxCNC. Take a look the ProbeBasic GUI for LinuxCNC. It is very slick. I ordered a new SSD drive to install ProbeBasic on for testing on my Pulsar.

So it looks like I would need a Mesa 5i25 as the Gigabyte motherboard has PCI slots. Mesa has this listed for $89, which also seems like a very reasonable price.

Having a Mesa 5i25 would allow either installing either LinuxCNC or PathPilot. Any idea how hard it would be to start with a LinuxCNC install and then do a PathPilot upgrade later?

So Probe Basic is much like PathPilot in that both are fancy GUI's for LinuxCNC? Probe Basic V0.2.6 (and V0.2.7 and V0.2.8) was released on March 11, 2020. With such a low version release number, is it ready for prime time? So you have a SSD for PathPilot and a SSD for Probe Basic?

My NM200 has an external box with all the electronics. The PC connected to the BoB with a short 23? pin cable. For PP I had configured a low cost PC that could accept a Mesa card. This low profile box sits on top of the old box and uses the same cable to connect it to the BoB. To run the mill I power on both boxes. If for some reason the PP box were to fail I could reconnect the cable and the mouse/monitor/keyboard and run Mach3 again. Whether this is possible with the Torus electronics cabinet I can't say.

I haven't upgraded PP from 1.9 as it does everything I need.

If I bought a new SSD and installed PathPilot on it, I think I could switch back and forth between Mach3 and PathPilot like you do just by plugging and unplugging SSD's.

Looks like PathPilot is up to Version 2.4.3. I'm not sure I'm going to have any choices to run older versions of PathPilot like you are if I start down this road.

[Titaniumboy]

I run Windows 10 but see no reason why you couldn't stick with Windows 7. I stopped Windows 10 updates when it broke Mach 3, and still occasionally use Mach 3, but 99% of my machining is done in UCCNC now. I don't know how Novakon set up their controller but I assume its parallel port and not a dedicated motion controller? With the UC300eth-5lpt there's not reason not to multitask, run various software, etc.. Everybody does it and it's one of the prime advantages of a Windows based controller IMO. I post G-code to a folder on my C drive, load the file in UCCNC, and hit cycle start. No need to have multiple PC's in the shop. And it's nice to take care of email, browse the web, play music, etc..

Novakon's setup consists of a Gigabyte motherboard feeding signals into a Novakon BOB via parallel port. I'm hoping to setup my touchscreen monitor at the mill, so it isn't going to be very convenient to stand in front of the mill running other software. I can sure see the convenience factor though if you only had one computer in the shop.

Importing the Mach3 setup file worked decently for me, but not perfect. I can't recall what the issues were, but I remember having to adjust/enter some settings manually. UCCNC just has different setup options than Mach. That said, UCCNC is way easier to setup than Mach IMO. The setup pages and options are just more intuitive.

Good know that G41 is implemented, I guess I should pay better attention to their release notes.

The Mach3 demo program came already setup from Novakon, so I'm not sure how hard a brand new Mach3 install would have been. That being said, I did have to play around when the mill first arrived setting soft limits and axis limits and so on.

Personally, I do not recommend doing this, because it won't setup everything 100%.
The reason is that Mach3 has settings that UCCNC doesn't have, and UCCNC has settings that Mach3 doesn't have.

Setting up UCCNC is very easy, and is very much like setting up Mach3.

And UCCNC has had G41/G42 in test versions for close to 18 months, maybe longer?

Thanks for the recommendation to avoid the UCCNC auto setup from Mach3. Thank goodness that it still will be an easy setup.

Is G41/G42 Cutter Comp only in test versions, or is it also in stable releases?