Threaded rod accuracy

My latest printer uses three threaded rods to lift the Z axis (see photo). I'm using stainless steel threaded rods from McMaster-Carr because I thought they might be more accurate than random hardware store rod. Two of the rods are pretty good and produce a pretty consistent layer thickness. If I print something close to the 3rd rod, however, there is a substantial variation in layer thickness. In other printers I've built using cheapo threaded rod, this hasn't been a severe problem. Has anyone had any experience using different types of threaded rod? I seem to recall a thread somewhere saying that brass threaded rod is more accurate. Is the thread rolling process maybe more accurate in softer metals (which is why my harder stainless rod runs like crap)?

I'm ultimately going to bite the bullet and swap in ACME lead screws, but I'm curious if anyone has looked into the differences between different types of threaded rod (for those on a budget).


Edited 1 time(s). Last edit at 09/07/2014 10:36PM by LoboCNC.

Read this: [www.protoparadigm.com]

I'd check to make sure that the odd threaded rod isn't bent, has smooth threads that don't cause binding, and is securely coupled to the stepper so that it can't slip. Flipping the rod around may show that the threads are slightly misformed, or trading two rods to see if the problem follows the rod would be additional tests.

Brass isn't automatically going to fix the problem. It may turn easier then a plain steel threaded rod, which will reduce binding which might be a source of your issues. But it isn't guaranteed.

Just a wild guess here; could this be due to the triangular structure? I mean the extruder with the motor is definitely not light, and at higher speeds the structure is going to deform slightly.

One way to cross out this possibility would be to print at ultra slow speeds, if the problem is no longer there then viola. If it still hangs around.... it's time to crack our brains.

I've put a dial indicator on the Z-axis table (the part that moves) and have pretty well isolated the problem to a periodic error in the thread. The lead is 0.050" and I've tested 1/4 turn motions which should each produce 0.0125" of movement in the table. You can see that periodically some 1/4 turn motions the translation is about 0.0135 and others are more like 0.0115. The rods are pretty straight (when rolled on a ground surface) and they are free to float side-to-side, so I don't think side-loading is a problem.

Which brings me back to the brass rod. Two of the stainless rods (fairly hard material) with small periodic errors were most likely rolled on the same thread rolling machine as the one bad rod. What I'm wondering is if the harder material might be more prone to inconsistent manufacturing than a softer material like brass. I did talk to a guy at Noll, which manufactures precision rolled threaded rod (both ACME and V-groove), and he said that rolled V-groove threads can actually be more accurate than rolled ACME threads. (I'm guessing this is modulo a ton of other process variables.) If softer brass rod is appreciably more accurate than steel rod, that could save a lot of people some money using brass v-groove rod over ACME.

Hi,

I tried both brass and Stainless steel M8 threaded rod on my machine for all axises. With about 20 turns per inch I really had to push the steppers to get my machine to move at decent pace. At max rpms the steppers are not producing alot of toque and will skip steps when changing direction. The rods also tend to whip around alot (yes even straight rods) and will cause wavy prints. I have since moved over to 3/8" 0.100 Haydon kerk Leadscrews. The HK screws are about 10 turns per inch wich allowed me to slow down the steppers and speed up the machine. The prints are now smooth and the machine is much happier. I would say go with the ACME rods from the get go and save yourself alot of time.

Gray

For grins, I decided to take my one bad 1/4-20 threaded rod and run it through a 1/4-20 die I had lying around. I wasn't expecting much, but it did make a noticeable improvement in the Z-axis banding. Not perfect (I'm still going to switch to ACME rods) but definitely worth doing. The thread die I used just happened to have some bite into the stainless threaded rod, but it wouldn't remove any material from a cheaper rod which had a looser fit thread. However, you can get adjustable dies (for like $4 on amazon) that should allow you to shave off minute amounts from even looser fitting screws. If you are lamenting the Z-axis banding in your printer, it might be worth giving this a try.

One thing I do on all the thread rod used for Z drive is chase the part with a die. There is always some grit and grime. The best threaded for drive I've used is 10.9 hardened steel rod. McMaster doesn't have it in M5. I get it from my fastener distributor but it's about six times the cost of the McMaster rod.

Quote
herrgray
With about 20 turns per inch I really had to push the steppers to get my machine to move at decent pace. At max rpms the steppers are not producing alot of toque and will skip steps when changing direction.

This is why it's not recommended to use standard threaded rod for your high speed axis that are making rapid changes. It's not as much of a problem for a Z-axis on a typical cartesian printer. It might make things slower at the begining or end of the print when the bed or extruder carriage is moving up or down, but after that the only movement is typically a fraction of a mm and only in a single direction.

Quote
vegasloki
The best threaded for drive I've used is 10.9 hardened steel rod. McMaster doesn't have it in M5. I get it from my fastener distributor but it's about six times the cost of the McMaster rod.

So you've had better luck with harder steel rod (so much for my theory about softer materials producing a more accurate thread). Is this because they are straighter, smoother, more accurate, or all of the above?

Quote
LoboCNC

So you've had better luck with harder steel rod (so much for my theory about softer materials producing a more accurate thread). Is this because they are straighter, smoother, more accurate, or all of the above?

I've found it less susceptible to bending or warping. The low strength mild steel while it works and is a good value, it flexes when held up in the middle in a full stick. Also the same nuts have less play even after the threads are chased. With a die and a relatively clean stick I can feel just a bit of resistance and almost no play. I can't say if it's the spec of the material or just that this supplier make a more quality product. It came from Germany via a metric specialty importer then to my distributor. I'd reckon that it's a more quality process all the way around as the application for something like a 10.9 grade fastener is more demanding than low cost steel. I haven't seen any 12.9 in M5 but I'd be curious to try it. The drawback is that by the time one sources and gets the material they could have just as easily and for not too much more sourced TR or ACME lead screws. For the MM2s I use TR10-2 lead screws from Makers Tool Works for under US$40 a pair with the nuts.

Quote
cdru
This is why it's not recommended to use standard threaded rod for your high speed axis that are making rapid changes. It's not as much of a problem for a Z-axis on a typical cartesian printer. It might make things slower at the begining or end of the print when the bed or extruder carriage is moving up or down, but after that the only movement is typically a fraction of a mm and only in a single direction.

On the back burner I've got a mod for the MM2 that uses the same ball screw as the Aluminatus from Trinity used. It's a PBC part, IIRC a 2 start TR12 screw assembly. I haven't got the part # on me right now but it was farily expensive, over US$200 but that includes the entire assembly, carriage,motion, etc. Apparently that is how Ezra was able to get the X and Y so fast and still retain precision.

Quote
vegasloki
I've found it less susceptible to bending or warping. The low strength mild steel while it works and is a good value, it flexes when held up in the middle in a full stick. Also the same nuts have less play even after the threads are chased.

Low strength steel is definitely more susceptible to being (or arriving) permanently bent, however, the stiffness of plain carbon steel is pretty much independent of the strength. That is, low strength steel will flex no more than high strength steel (as long as you don't cause a permanent bend). As you noted, the cheaper threaded rod usually has a Class 1A fit which is looser than Class 2A fit. Because a Class 1A threaded rod will actually have a slightly smaller diameter, it will flex more -- the overall stiffness of the rod is proportional to the cube of the diameter so a small difference in the diameter can make a noticeable difference in the stiffness. All Class 2A rods should have the same stiffness regardless of the strength.

Also interesting to note that ACME or trapezoidal threads, for the same nominal OD, have a smaller root diameter than V-groove threads, and hence, will be a little less stiff.

Final recap - I replaced my v-groove threaded rods with ACME threaded rods and the banding was worse than ever! This finally made me realize that the problem wasn't the threaded rod, but rather, a design error in how I was supporting the threaded rod. (The details are too arcane to bother describing.) Now, if I squint really hard, I can see just the tiniest bit of banding when printing 0.1mm layers, and only when I print in black, which visually, seems less forgiving. There's also no real difference in print quality between my v-groove and ACME rods. But the ACME rods do look more pro.

Quick off-topic: What sort of printer is that, and are those just wheels holding up the effector rods?

Quote
epicepee
Quick off-topic: What sort of printer is that, and are those just wheels holding up the effector rods?

The printer is one I designed. Here's a thread with more about it:

Lobo3d printer

and a couple of video links:

http://youtu.be/YN8xI0NG7kU
http://youtu.be/NoDgxs9XlQo

It's quite a bit different from most standard designs. I should post some photos of prints now that I've got it reasonably well tweaked up.

Ive noticed with the threaded rods, that is they are used at higher speeds (same speeds of Belt Systems) that they are very loud.
find a video of the Aluminatus Printer and listen for when the printer stops at a corner or tries to stop at a point, you will notice the difference in sound compared to belts.

hear the noise from this video, [www.youtube.com]

Quote
pushthatbolder
Ive noticed with the threaded rods, that is they are used at higher speeds (same speeds of Belt Systems) that they are very loud.
find a video of the Aluminatus Printer and listen for when the printer stops at a corner or tries to stop at a point, you will notice the difference in sound compared to belts.

My printer only uses threaded rods on the Z axis (see OP). The noise in the Aluminatus printer is because lead-screw drives are much stiffer, and vibration from the stepper motor radiates into the structure much more. With a belt drive, the belt damps out much of the vibration, but you do lose a little accuracy and stiffness. My printer uses belts for the X and Y axes.

Edited 1 time(s). Last edit at 09/19/2014 11:28AM by LoboCNC.