After further experementation, Heat set inserts for the spring part will be the following Mcmaster Carr part number:
Thread 6-32 Height 0.150" Small outer diameter 0.199" Large outer Diameter 0.209" Quantity 100 Part number 93365A130 $11.81
with the following soldering iron insert installation heat set tool
92160A119 Cost $14.42
During installation, heated plastic flows into the insert’s knurls and ridges. When the plastic solidifies, the insert resists torque and pull-out. Install or remove using a soldering iron with an installation or extraction tip (sold separately). Inserts are made from brass, which is nonmagnetic, corrosion resistant, and electrically conductive. Thread class is 2B for inch sizes and 6H for metric sizes."
Acme nuts arrived today. I’m actually a tiny bit too small in the latest revision of the spring. you can clean it up and make it work, as I did in the pictures below, but I will upload a slightly enlarged version here later on.
Yet another spring rebuild. This one does not change any mounting holes. I increased the bore slightly, the hex nut seems to fit with minimal slop, but still able to freely move up and down. I also removed a segment of the upper spring to make the whole assembly more flexible. I printed part of the upper section and it seems to fit, will print the whole thing tomorrow. In the mean time, there’s the file if you want it.
I’m not expecting the heat set inserts until Saturday. based on how those go, i’ll either really be done with the spring, or moving things around again. This weekend i’ll print the whole package if it’s ready to go. spring5_a.stl (582 KB)
Spring 5_a printed fine, and is dead on for the nut socket size. If I make it any smaller it’s inside the layer variance for my printer. Also printed the newer backplate It also prints ok.
I experemented with a few slightly too large heat set inserts, and I was able to place them using a solderin iron.
I am going to have to change the mounting holes yet again though, which I was not looking forward to, But i’ll hold off on that until i get the final inserts. The spring itself may be fine as is, the backplate and both the motor and idler clamp ends are going to need work. the lower idler hole may work as is.
Splitting the upper spring seems to have worked, it seems to move better now. If this doesn’t do the trick, next step will be increasing the loop size slightly.
I also changed the design so it will only use 2 Acme nuts. the lower nut socket “floor” is plenty thick for the metal spring specified. I may not even toss a washer in there.
Cleaned the mounting holes slightly, also tightened up the lower nut socket somewhat as the lower acme nut should be snug. If you printed 5_a, that one should still work. This one will make it easier to insert the heat set inserts. spring5_b.stl (584 KB)
The anti backlash nut moves freely in it’s socket. The bottom one is snug. Tomorrow I’ll print the rest of the set and see how everything fits. If I’m happy with it, I may actually get to install it. We’re close now either way!
Redid the backplate a little bit again to make the holes line up better. Seriousl, I hate making holes in Autodesk inventor. I miss my autocad osnaps! Theres got to be an easier way to make them more precise, but I haven’t found it yet.
In other news, I have 2 backplates, 2 springs and 2 lower x clamps printed. Idler and motor mount final will print tomorrow. I expect i’ll have to move a few holes on both of them as well. backplate3_a.stl (189 KB)
Changes to the Idler and Motor mount again. I got 90% through printing the idler when the power dropped due to the storm a few hours ago, reprinting it now but it appears that this will clear the nut correctly. i did have to move the idler bolt out evenfurther, which will mean a longer belt will be required for most people. Which isn’t ideal but i don’t know how else to make everything fit.
Motor side just has the relocated backplate screw holes. Still need to print it, but it’s the part i’ve altered least of any of them so I’m sure it will fit fine.
At this point, All files are theoretically final. Once I print the last pieces and test-fit, I’ll upload them to thingiverse as well. I’ll also upload step and dwg’s for anyone who wants to make compatable parts, etc. spring5_b.stl (584 KB) lowerx_a.stl (103 KB) xmotor_3_a.stl (168 KB) xidler_4_b.stl (186 KB)
And the last 2. Apperently there is a 6 attachment limit per post.
Anyways, here is all the design files. I hereby release them into the wild under the terms of the GNU Creative Commons - Attribution license. If you use these parts in a production 3d printer for resale, name the first one “Tim” or something (or hey, send me a free printer? just kidding, it was worth a shot though right?)
Future enhancements once the final testing and varification occur:
extending the backplate to accept LM8UU or LM10UU bearings.
Project “Linear Z Rail”
Variant springs to accomodate leadscrews, etc
Different tension springs. I am still not convinced my current springs are capable of enough movement.
I actually need to reprint the motor mount before I do the trial fit since I’ve got lifting on the bottom of the clamp surface in one spot. Once that’s finished though I’ll tear the machine apart and see how everything fits.
I need to find stiffer metal springs for the interior of the anti backlash. The ones in there right now work well going up, but coming back down they don’t exert enough pressure to keep the axis aligned. A possible fix would be putting a lower cover over the lower nut to lock it in place or something. Maybe a set screw, I dunno. I’ll figure that out this weekend. The plastic portion of the spring I think is too stiff, but at the same time it seems to be working great. with the hotend turned off, I can feel that it is going up and down in exactly the same spot. before I could feel the back and forth slight wobble.
I also need to clearance around the idler assembly better. I ended up having to cut away a portion of the backplate to clear the nut, which doesn’t impact structural rigidity at all, and it’s hidden by the screws. I am using larger screws than what I recommended above, simply because that’s what I could source locally. I think this will be less of an issue with the proper screws and heat set inserts.
The last issue, and its kind of a big one, is the plastic spring itself sticks out slightly too far, causing the carriage to bind at the bottom. It’s not off by much, and I think part of it is that I don’t have the plastic heat set pieces down all the way flush. I may have to make the spring profile slightly narrower to fit though. Probably a millimeter each side at most, maybe less.
Going up at the moment, according to my dial indicator, I have little, if any backlash. Going down, if I manually put weight on the carriage to counteract the too weak metal spring, the story is similar. It’s not perfect yet, but I am encouraged!
Alright, got everythign back together again and the first post install test piece with the new metal spring is printing now. So far it looks like an improvement, but I haven’t had a chance to look at it close up yet since its on the bed. The new spring is a: Handyman springs SP-9729 Compression Spring 5/8" x 2-34"" x .054 15.9mm x 7cm x 1.4mm) 11.72lb / 5.31kg maximum capacity spring (as found at ace hardware locally) with 15mm worth cut off at one end. That seems to be about perfect in terms of tension.
I ended up pushing the heat set inserts in an extra mm, then shaving the plastic spring feet to fit. I also apperently had the two sides too close together on the clamp ends. Once I backed those off a bit and reinstalled the nut holder spring with the stiffer metal spring, I was able to run up and down the z axis with no binding and no lash at all. Comparison pictures will follow once the print is done. If all seems well, I’ll call the project finished for now, and publish it to thingiverse.
I also got the parts in for my Y rail retrofit finally, so i may also get that in this weekend. At that point I’ll be pretty dang accurate I hope.
Ok, I am officially amazed at the quality difference. Almost all of my wobble is gone. I apperently suck at trying to capture it on camera though, so I’ll have to play around with that a bit and get one thats in close enough so you can see. As a side benifit, the layer spacing is also much improved. I think my remaining wobble is due to my ill aadvised aluminum helicle spring frankencoupler experement, which I will replace with something else shortly. Having seen the results of the rest of the project though. I would definitly say this is a worthwhile upgrade as is. I do think the plastic spring needs some more flex, and I will likely modify it a bit more later on. but aside from that, its working quite well
I noticed an immediate quality improvement. Before the upgrade I had regular banding offset on everything . After I have almost none. The layer height consistency also improved considerably. I’m printing with a .5mm nozzle too.
This picture shows my new linear rail project motor mount, which was printed with AO-taz. There are a few minor blemishes, but that was mainly due to a layer height blip.
Here you can see the corner of the new motor mount for the rail project, on top of an upside down idler which was one of the AO-taz prototype parts that I was using as a spacer. Ignore the first couple of layers on the motor mount, I had to trim away a part that I later removed from the design, but you can see the difference between the top and bottom part, which was printed with essentially AO-101 stock hardware
Here is a prototype Ao-TAZ piece on the right, next to a similar, but not entirely identical part on the left that was printed on the finished AO-TAZ pieces. The one on the right was also printed on basically stock AO-101 hardware.
This is a close up of a tooth segment of a very large gear printed on AO-TAZ hardware, Ignore the white fuzzy plastic bits. that was left over from the Brim that was only partially removed in that picture.
