[HDVideo]

A question for the Probotix users.

I recently purchased a pre-owned Asteroid that had excessive slop in the Z axis bearings. New bearings were ordered on eBay, but didn't seem to improve the situation at all. My options at this point are to order another set of bearings, this time from Probotix, and hope for the best or modify the Z axis to accept linear profile rails in place of the unsupported round rod/single bearing arrangement. The modification would consist of adding four holes to the X plate, eight holes to the Z plate and machining four spacers to match the existing X plate/Z plate spacing. Probotix wants $100 for two replacement bearings and I can get two rails cut to length with equal hole spacing on each end and four flanged bearings for $150. The downside is that +Z travel would be reduced by about an inch, but there would be no loss of clearance under the gantry and no loss of Z travel in the - direction.

Has anyone else tried something like this and if so, how did it work out for you?

(My original idea was to machine a new, slightly wider Z plate as in the images below, but the original plate could also be used by machining four holes along each left and right edge.)

 

[4DThinker]

Some of the play in your asteroid, same as in the Nebula they used to sell, is that the beam itself can twist under the heavy cantilevered load of their design. They now have replaced the 3060 beam with a heavier walled 6060 beam to stiffen up the whole gantry/z-axis assembly. So changing out how the Z axis is made can 't hurt, but also may not completely eliminate the play you have.

Their new machines all come with that upgraded gantry beam and taller side plates to increase the Z cutting area.

I proposed a design change for their Z axis assembly awhile back, which would reduce what moved up and down cantilevered, and stop the rail frame from coming down with the bit. They answered that their design is the way it is to be easier/more reliably manufactured.

4D

 

[HDVideo]

Thanks for the comment, 4DThinker. I agree that doing this upgrade will not eliminate all slop in the Z axis, but as it sits now, I can hold the router collet between two fingers and move it forward and back .015" to .020" (measured with a dial indicator) and this was with the Z axis only about 2" below home position. If lowered all the way to frame level, the deflection would be even more. That movement is pivoting at the Z bearings so the bit height is also changing. I never thought this machine would be super rigid, but certainly expected better than this, particularly after reading so many glowing reviews of Probotix machines.

I've called the company twice, asking to speak with Len, but he hasn't been available. I wasn't calling expecting them to fix it, but hoping to get some assurance that if I purchased new bearings from them, significant improvement would be seen. The two technicians I spoke with were not able to offer much encouragement.

The new design from Probotix, announced right after I purchased my Asteroid, is certainly an improvement, but I was disappointed to see they stuck with the unsupported round rod. At least they now use two bearings on each side and were able to spread them apart a little. Better than the prior design, but still a weak point in my opinion.

Ed

 

[4DThinker]

We did replace the Z axis bearings once on our Nebula. The new bearings from probotix did help, and so far that Nebula has done fine for light cuts. Heavy cuts through hardwoods were still triggering limit switch errors on occasion. Their new configurator has a "soft limits" option that basically ignores the limit switches after the machine is homed. This lets it keep cutting. When the spindle is low I can also flex the assembly. The play seems more at the beam itself though rather than at our newish Z bearings.

On their facebook page Probotix has been asked about a gantry upgrade kit for older machines. It would be quite involved, requiring taking the Z axis and gantry apart down to the side frame rails. Cables would have to be lengthened and re-run. Not as easy as simply assembling a new machine. I doubt they'll ever offer one, but I did hear someone sent a machine back to them to be upgraded.

On my 2' x 4' Meteor and the one I use at work the narrower beam means more stiffness at the spindle. I've always thought their 2' x 3' Asteroid would have been stiffer with the gantry spanning the 2' dimension.

4D

 

[HDVideo]

I doubt they'll offer a field installed upgrade to the new design. As you say, it could be owner installed, but would not be as quick and easy as some might expect.

I've been running UCCNC on my milling machine for a year or two, so opted to use that for the Asteroid as well. Both machines run from the same PC by simply loading a different configuration file when UCCNC starts. LinuxCNC is a good program, I've used it in the past, but UCCNC has several advantages, particularly with the 2017 Screenset available from TheCNCWoodworker dot com. To list just a few:

Runs under Windows on newer computers and (in my case) uses an Ethernet port. USB motion controllers are also available.
Up to five parallel ports available for I/O with use of the UC300ETH-5LPT
Soft limits (I think most control programs these days have soft limits)
Auto Zero on tool change using fixed touch plate - 2017 Screenset function (Same method as Probotix Atlas Tool Length Sensor but doesn't cost $300)
Very good UI with easy access to all settings
Very good support from CNCDrive dot com

Maybe the only downside is that UCCNC will only work with one of the USB or Ethernet motion controllers sold by CNCDrive dot com. Not a real issue since their products are solid and an excellent value in my mind. Disclaimer: I don't work for the company or have any affiliation with them at all, just a satisfied customer.

I'm not saying people should abandon LinuxCNC, it is a good program and has a strong following. My only point is there are options available, other than Mach3/4 for those wanting to use Windows and don't have or want to deal with a parallel port.

Ed

 

[probotix]

There are a lot of junk bearings out there. We hunted high and low and several trips to chinese factories to find one that was top notch. That said, those types of bearings will have some amount of variance, even with the high quality ones. Its important to preload them against each other with some clamps to take out as much slop as possible as you are tightening the screws to install them onto the gantry face plate.

The new gantry design is way more rigid, and we do plan to offer an upgrade kit as soon as we can. We're just buried with machine orders right now and haven't had much time to look at it. Things usually slow down in the summer, so hopefully we'll have one for sale in the next couple of months. We have a machine sitting here right now from one of the local customers which we are going to use as our guinea pig.

One of the dilemnas we're facing is what to do about the gantry rails. They could potentially be reused, but installing them centered on the rail is a challenge. We have custom made centering fixtures that we fabricated to aid us in assembly here in the shop. Most of our customers do not have access to a granite slab and dial indicators and such, nor would know how to use them if they did. We want to make the upgrade available, but don't want to create a support headache in the process.

 

[HDVideo]

The new gantry design is way more rigid, and we do plan to offer an upgrade kit as soon as we can. We're just buried with machine orders right now and haven't had much time to look at it. Things usually slow down in the summer, so hopefully we'll have one for sale in the next couple of months. We have a machine sitting here right now from one of the local customers which we are going to use as our guinea pig.

One of the dilemnas we're facing is what to do about the gantry rails. They could potentially be reused, but installing them centered on the rail is a challenge. We have custom made centering fixtures that we fabricated to aid us in assembly here in the shop. Most of our customers do not have access to a granite slab and dial indicators and such, nor would know how to use them if they did. We want to make the upgrade available, but don't want to create a support headache in the process.

If you need a "test dummy" for a field install of the upgrade, let me know. I'd be happy to document the process from the standpoint of an average user. In the interest of full disclosure, I do know what a surface plate is and how to use a dial indicator.

Ed

 

[4DThinker]

The old 3060 gantry beam could be used as a cross beam in the bed frame. Easy enough to cut to the needed length.

BTW @probotix, my K-state shop manager inquired what I would recommend for a 50" x 50" CNC to take some of the load off of our large Multicam. I mentioned that you've occasionally built larger machines. Any chance you're interested in making a 50" wide Nebula with the new taller gantry and 6060 beam? Heck, for that width a 6090 beam would be even better.

4D

 

[probotix]

We've been building 5050s pretty regularly this year. Email us.

 

[4DThinker]

We've been building 5050s pretty regularly this year. Email us.

Email sent.

4D

 

[honesttjohn]

4D,

You think if I got a Meteor with Nema 34's it would be noticeably faster than my Nebula that came with the 23's? I don't think the Nebula's size with the lead screws would handle the 34's very good. Then I would have two machines using the same system. I need a faster machine. Thoughts?

 

[4DThinker]

4D,

You think if I got a Meteor with Nema 34's it would be noticeably faster than my Nebula that came with the 23's? I don't think the Nebula's size with the lead screws would handle the 34's very good. Then I would have two machines using the same system. I need a faster machine. Thoughts?

Probotix still starts them out with a 200ipm feed limit in the X and Y directions. No matter what steppers they use. I've set my X axis to go up to 260ipm, although I left the Y at 200. Longer lead screws tend to start whipping when run too fast. The best improvement I've made was to speed up the Z axis, and when I can reduce the safe Z height I do that. For many cuts the time it takes the Z axis to move up and down is where time is consumed/wasted. I'll also reduce the number of default passes for pocket and profile cuts by one or two to eliminate some cut time.

Aspire isn't the most efficient when calculating pocket toolpaths when they are more complex than simple squares or round holes. I can usually reduce the cut time by half or more if I take a little more time to use profile lines to do a job. I reduce how many times the bit has to stop and lift up, and eliminate any times when the bit is cutting air by going over areas where the wood has already been removed.

4D

 

[probotix]

4D,

You think if I got a Meteor with Nema 34's it would be noticeably faster than my Nebula that came with the 23's? I don't think the Nebula's size with the lead screws would handle the 34's very good. Then I would have two machines using the same system. I need a faster machine. Thoughts?

Bigger motors do not spin faster, actually slower because of inductance. To make these machines run noticeably faster, you'd need to move to a larger diameter screw and turn them with servos, or change to rack-n-pinion drive. Even then, you may not realize much speed improvement because the acceleration and deceleration will limit your ability to reach those higher speeds on such a short length of travel.

 

[honesttjohn]

Beltrmidave,

Need your input here too. You've had experience with both types of machines.

 

[beltramidave]

Beltrmidave,

Need your input here too. You've had experience with both types of machines.

Len knows a lot more about this stuff than I will ever know... That being said, Len, are you still using the lead screws on the 50 x 50 machines? Any plans to go with R&P drives or ballscrews?

Here is a link to my initial setup video running 600ipm rapids.
https://www.dropbox.com/s/afvfduhyub5r9dg/VID_20170404_120006306.mp4?dl=0

 

[honesttjohn]

If Len comes up with a rack and pinion machine the size of a Meteor or bigger with the higher Z axis that uses Nema 34s - I'm in. I'd even be a guinea pig. Being able to use the current Linux setup on both my machines would be great. I don't want to have to learn a new system like Mach 3 or 4. I know Beltramidave can cut a file that takes me almost 5 hours in under 2 hours.

Len -- start a "Pro" line of machines for serious cutters.

 

[4DThinker]

I'd be interested in a beefier/faster 4x4 CNC too. Of course I'd need a larger shop to fit it into.

 

[HDVideo]

There are a lot of junk bearings out there. We hunted high and low and several trips to chinese factories to find one that was top notch. That said, those types of bearings will have some amount of variance, even with the high quality ones. Its important to preload them against each other with some clamps to take out as much slop as possible as you are tightening the screws to install them onto the gantry face plate.

I'm leaning toward ordering a set of Z axis bearings from you rather than doing the linear rail mod. Can you elaborate a little on the steps necessary for preloading the bearings during install? When you say "against each other" do you mean apply pressure to hold the bearings toward or apart from each other? Do you recommend removing the entire carriage assemble from the gantry or will removing the Z limit switch, stepper motor and gantry face plate from the upper and lower X bearing plates suffice?

Sorry for all the questions, but I really want to get this machine operating at a level I know it is capable of achieving. Your business would never have survived this long if the amount of slop I have in my machine was considered normal.

Ed

 

[probotix]

We use a pair of ratcheting bar clamps to squeeze them together with the z fully assembled, but not too tight. If you get it too tight, it will bind at the top and bottom of travel.