T8 lead screw with anti-backlash for 525 modification

I’ve been updating by MPCNC to the 525 mod with the T8 lead-screw and a anti-backlash nut

The neck and spring on the anti-backlash are thicker than the standard nut and I don’t think there is enough material left in the C-XYZ-T8 to open up the bore to swallow the anti-backlash nut. I wound up just flipping it around. There does not seem to be a lot of wall thickness in the C-XYZ-T8 component and I thought it would be less prone to failure if I put it in compression. I re-modeled the C-Spacer in FreeCad so I could use M3X40 cap screws with self-locking nuts.

One problem that I had was that heads of the M3 bolts semi-interfered with the spring. I found a washer in my junkbox that fit nice over the nut and seated over the cap screws. It seems like this should work.

The T8 lead screw I bought off amazon came with the coupling and bearings with flanges Originally I was just going to with the MPCNC stock bearing assembly but I was having a problem fitting lead-screw through bearings. So I think I’m going go with the flange bearings.

At this point, I’m inclined use the flange bearings that I got off amazon. whats sort of neat about them is that the bearing ID is extended and has two set screws to secure the lead screw. I’m thinking that this will take the thrust load off the z stepper motor…

At this point, I’m thinking my next steps are to:
Redo and extend the z-axis pipe to accommodate the length of the T8 lead screw.

Mod C-Z_lower to accept the flange bearing.

Either Flip C-Z_lower and use the existing C-ToolMount (which is easy but might cause some clearance issues for my router) or mod the C-ToolMount to have a pad to mount the lower bearing (but that’s a lot of work for me to reverse engineer.)

This is still a work in progress.
Any thoughts, would be appreciated.

Nice work, but I’m not sure to see the point in the anti backlash here, since the rod is always pre loaded with gravity due to its own weight.
The lower bearing could be a good idea, but you need to make 100% sure that everything is perfectly straight first, otherwise it will increase the wobble and/or the travel resistance will be inconsistent, which will make things worse.

You can very easily create a support for it using tinkercad, just import the original tool mount stl file in tinkercad and build a box and a few holes to attach your bearing, if you don’t care about how it looks it should be done in less than 5 minutes :slight_smile:

I have some of those anti backlash and found they really are not needed. Worse case with how it is would be that you would find enough backlash after a lot of use to need to replace the brass nut, cheap fast. Those anti backlash nuts will actually make the brass nuts wear faster and for them to work the spring rate has to be greater than the load and I can’t be sure they are. If yours has noticeable backlash you might try another nut.

How you have yours mounted in those pics if I am not mistaken will actually make it worse when you do develop backlash. As Dui says the Z is preloaded with gravity so we aren’t actually concerned with backlash we are worried about slop. Meaning if you had some slop in the z axis nut and you were trying to push into something you might not get the entire z depth you want or it could chatter slightly up and down. Hard to put into words…let me try another way. Gravity is holding the axis down no matter which direction you are driving it (no backlash), if it was a horizontal axis they would be a better idea because nothing is preloading the axis), the concern is the work pushing it back up, most bits we use are UpCut bits meaning they are pulling into the material (same direction as gravity, the system stays preloaded, no slop/backlash). Where the concern comes in is when you plunge really fast or use a down cut bit that could actually fight the preload direction of gravity. So you really want the spring pulling in the direction of gravity or you will be adding to the slop.

As for the flange bearings, all good if you put in working how and where I have mine, supporting the coupler. This is to help alleviate slop and take the load off the stepper. Putting one on the bottom as Dui pointed out will most likely give you a wobbly axis. This would overconstrain the system and show any bend you might have in the lead screw or any imperfections in how concentric the the coupler joint is.

I hope this isn’t discouraging, and I hope I am correct in all of this. I wrote this to put my logic into words so if I have any of this wrong please, anyone, feel free to correct me and we can adjust parts/design accordingly. The whole bit of logic is why I did not just initially make the MPCNC with lead screws, I thought they would be a sloppy mess, a year and a half later when I actually took the time to think about it and do some tests, turns out they are pretty good.

Oh and as for beefing up the screws not such a big deal they only keep the nut from spinning they don’t actually need to hold it in place. I only use 1 screw on mine. Your solution is pretty slick though, We might be able to use that to actually help with assembly. A screw from the top and one into the other end to keep the spacers aligned!

The proper item is a ballscrew, but the cost is very high.

Thanks for all the feedback.

I actually have some ball-screws in inventory, but the 525 doesn’t allow for large diameter screws ;(

I’m not thrilled with my anti-backlash nut arrangement. The more I think about this, I want a system where the spring force isn’t the sole mechanism which eliminates the backlash (which isn’t the case at the moment with the weak spring) . I didn’t measure it, but there is a disappointingly large amount of end play in the nut pair that I have. I think I want to have a weak spring in between the nuts and some way to pull the nuts together with a fine adjustment so I minimize the endplay to whatever works for me. This video I ran across sort of runs across the concept. [youtube]https://www.youtube.com/watch?v=sAOP9kkm-4A[/youtube]

I hear what your saying about using cutters that produce a down thrust… One of the potential use cases I like to do is to drill holes in aluminum and I don’t want a lot of potential end play in the z-axis. I just have a bee in my bonnett and I want a way to be able to minimize the end play in the lead screw without inducing a huge preload on it on the lead screw.

As far as the bearing support a friend from the makerspace I belong to, (who has built a large gantry mill) suggested that I use a thrust bearing on the bottom only. It was an interesting suggestion which I need to think some more on. (He says that it I’ll avert binding on the upper side.) I think what ever I do, I’ll need some way to radially float the bearing and then lock it down when I get it aligned properly. (Any chance you can shoot me a step file of the tool mount so I can add a bearing mount to it. When I convert a STL to a solid in freecad, it seems like it’s not as clean as I would like)
(sorry if you’ve been getting double posts, my mouse doesn’t like your website.)

The top bearing is there to relieve the stepper of the stepper slop, and to save it from trying to support all the weight. the bearing and coupler are a team and are very important. I don’t see how you could use a bearing at the bottom for anything useful.

Any chance you could send me off list a Step (or Iges) of C-Z_lower and C-ToolMount (for the inch version)?

I was looking at pricing on Chinese ball screws. Some of the prices seemed amazing good. (I don’t know if you can get consistent quality at that price level though)

I don’t give out CAD files, sorry.

Instead of modifying parts and buying all kinds of new things, why not use what we all are using and have no backlash issues?

Just use the stl, it doesn’t matter if its not perfectly clean. I’ve done this plenty of times with the tool holder and I’ve got maybe 4 or 5 different versions of it for different tools and tests, it will work fine, don’t worry. If it doesn’t work well in freecad, just try it with tinkercad (entirely free stuff, no need to install anything, the thing is a browser app), if worked sufficiently well for me.

And yes, you can find some pretty decent ball screws from china for affordable prices, you may want to give it a try if your goal is to mill aluminum with a good precision, that seems like the best option.

Well what got me off the main path is don’t want my z-motor and flex coupling to take a thrust load (as well concerns you posted about wear on the nut)

My original plan was to mount a collar on the shaft and have this thrust bearing https://www.amazon.com/CNBTR-Silver-3-Parts-Thrust-Bearing/dp/B06Y5MMSSD/ref=sr_1_6?ie=UTF8&qid=1496529038&sr=8-6&keywords=thrust+bearing+8mm between C-Z-lower and the collar. The bearings unfortunately are on a very slow boat. My plan was to use your bearing arrangement, but the T8 shaft I got (https://www.amazon.com/BIQU-Copper-Coupler-Bearing-Printer/dp/B01H1SBWGE/ref=sr_1_9?ie=UTF8&qid=1496527691&sr=8-9&keywords=T8+300+mm+lead+screw off amazon, is too big to fit through the bearing. I thought the package was a good price and originally I wasn’t planning on using the flange bearings. But now that I seen them they’re sort of cool and I want to give them a try.

One of the things thats nice about the current design, is that it’s really easy to get the z-axis off the carriage. If I make a integral tool mount with a lower bearing assembly, it’s going to become a real pain to assemble. I have no idea how well or bad a lower bearing is going to work, but I thought it be fun to try and see how to works. So… I’m going to extend the pipes slightly beyond the tool mount adapter and mount a modded C-Z-lower for lower flange bearing. If its a fail, I can always get a hacksaw out and revert to the original design.

As far a backlash compensation. After discussing and thinking about this https://www.amazon.com/BIQU-Backlash-Elimination-Threaded-Printer/dp/B06XCCDD51/ref=sr_1_sc_2?s=industrial&ie=UTF8&qid=1496528479&sr=1-2-spell&keywords=T8+antil+backlash+nut nut is probably not the best solution for a MPCNC z-axis.

The issue that with this system, is there is a huge amount of end play without the spring between the two nuts to get this to work because of the high lead. If I start pushing thrust through this thing, the spring would compress and I really wouldn’t have solved anything has far as backlash elimination. I’m going to switch to the single nut for now. The stock nut that came with the lead screw barely has any backlash. If that starts becoming an issue because of wear I have a solution with a second nut, that is inspired by the video I posted.

I found the bearing block that I got on amazon modeled in grabcad if anyone wants it.

I did a mod on C-ZLower to take the pillow block bearing. I’m curious to see how this works on the upper and lower.

Doing it that way you will need to increase the length of the Z rails and leadscrew. A better way would be to completely sink it into the block. I doubt this will have much effect on the zaxis rigidity, but every little bit counts in my book.

I actually had it that way to begin width and I changed my mind If I mirror those and use those on top and bottom I was concerned that that the pocket would be a place for saw dust/chips to accumulate which is why I went that route. It was in my mind to try to have the bearings facing each other. I suppose if I face them away from each other that would make sense to pocket the pillow block. The only down side I could see is that I a tensile load on the bolts and thrust load against the pillow block shoulder.

I think I’m going with the single T8 nut per your suggestion. If I start getting intolerable backlash, I have a plan how to address that.
(missing from the picture the springs on the bolts between the two nuts.)

I’ll wait until you finish. I’m not sure what you are going for so I might giving bad advice.

Np.Hopefully the backlash on the z won’t be an issue.

I’ve been trying to find the files to the tool mount for DW660 to fit your standard tool mount. Is there a master page or somethings that I’m not finding of links to adapters that will fit that mount?

Even better.

I found this thread while trying to understand how to mount the t8 nut to the printed parts.

I see how this system works and I like it. You mind sharing the files for the spacer?



How does the original design intend to mount the t8 nut? Just loosely with some screws?
Thread the printed parts?


Its still sort of a work in progress.

I haven’t installed the nut yet and I’ve changed a few things around.

I needed to add a spacer, which I’m going to morph into a bracket that I can mount my z-axis homing and limit switches to.