[LazyMan]

Mcmaster specifies .009"/ft or better. Many who use it seem to be quite satisfied but .009" to me seems pretty awful especially for dual X driven machines. Does anyone know who manufactures this lead screw? I would like to get more information about this.

[evilc66]

0.009"/ft is most likely for plain nuts, not anti-backlash.

[ger21]

They're probably made by Keystone. My guess is that 99% of these screws are much better than .009"/ft. It's also ±.009, so it could be .018" out in 2 feet, but in reality they're pretty accurate.

[microcarve]

It might also be mentioned that, that's "leadscrew error".
(Nuts don't have anything to do with it...that's a different
problem)

It may gain or lose a few thou depending on where it is on the length
of the screw.

For example, at 6" a few thousandths may have been lost, but then at
9" a couple may have been gained back.

There's a technique called "screw mapping" that's used to map out where
loses and gains are over the length of a screw. Then, software can compensate
for the error. PITA to set up....

Overall, with plastic nuts, the loses/gains balance themselves out very well, and
the error is rarely ever even noticed.

The only way to assure perfection is by use of $$$$ precision ground screws and
special nuts and end blocks/bearings. Those things also come with certificates
with people names on them that guarantees they do what they claim they'll do.

Hardly anyone ever notices the error in acme screws. The very slight error isn't
so important as it may seem on DIY machines. A common fix is to add/subtract
a few steps in software step settings and do a test run til a particular part is
as precise as the part has to be. That can change depending on the part
geometry and even where it's located on the machine bed, so it's usually viewed
as a temporary fix.

It's so rarely even mentioned, and that's because nobody ever notices it....

There are a few videos on youtube that show how screw-rolling is done.
Pretty interesting how they make them....and it's easy to see how amazing
it is that they're as accurate as they are, as quickly as they're churned out.


John

[fizzissist]

McMasterCarr sells average+ to better-than-average stuff. Generally, I like their prices, and love their delivery. But it's on the real generic side. If they say it's .009, it's a CYA, and likely better. But....you pay for convenience and a huge selection of stuff.

Another supplier you could consider:
1/2 X .500 Right Hand Hi-Lead Lead Screws & Nuts for Power Transmission - Roton Products, Inc.

I didn't see an 8-start in a 1/2-8, only a 4-start

[killersoundz]

Yeah I personally don't think 0.009" is anything to worry about at all with a wood CNC router table. I'd say most of the DIY cnc wood router tables I see are lucky to be accurate to 0.01" overall, with rigidity and backlash problems taken into account. Not only that but the accuracy of the setup and alignment of the machine overall by the builder. Personally I would be more than satisfied with 0.01" accuracy overall on my machine, I'm not cutting out engine blocks here...

[underthetire]

I saw that too. I just ordered some from Carr lane I think it was, slightly better, slightly higher price. I'm using it to diagnose a ridged tap problem on a mill. Even the precision ground one wasn't any better from the specs.

[louieatienza]

I use them on my machine and have been pretty happy with the results. Using anti-backlash nuts definitely help. I use only 425 in*oz steppers and they're plenty strong for most of what I do - I use one per axis, and my gantry weighs over 70 pounds.

[LazyMan]

They're probably made by Keystone. My guess is that 99% of these screws are much better than .009"/ft. It's also ±.009, so it could be .018" out in 2 feet, but in reality they're pretty accurate.

The MCMaster site just says .009. I guess its a typo as 2C fit ±.009 .I asked, keystone and their screws are ±.009. To me ±.009 is just sounds brutal even though from others experiences it sounds like they are better than +-.009. Just imagine though, that the worse case scenario is that with a dual x drive with 4 ft travel there could be a .072" difference at the extremes.