threads per inch in commercial threaded rod

[woodguy2]

I'm hoping to make incremental positioning stop blocks as part of a table saw cross cut sled using a short section of threaded rod meshed into a longer piece of threaded rod, and was surprised to discover that the samples of 6-32 and M6 stainless steel threaded rods I got from from Mcmaster have thread counts that add up to enough error that such a scheme doesn't work for accurate positioning beyond a few inches. I see that there are different class fit to the threads on some more expensive variants, but the stainless metric piece was class 6g, and still not 25.4 threads per inch over a long span (as best as I could measure).
There was a higher end (for woodworking) cross-cut sled that used this scheme a few years back that touted a "precision lead screw" for this, but it couldn't have cost them more than a few dollars each given the price and quality of the whole tool. There are plastic racks sold by Incra, used in their stuff, but they occupy more space than I have available in my design. What would it take to produce (or find?) 6-32 and M6 rod (1 pitch ) in say, 24" lengths, that would be within about .005" of even spacing for that length?

 

[neanderthal mach]

6-32 tpi or 6 mm 24" long and within .005" over that length? That ain't going to happen. For smaller diameter but still larger than you want actual machine feed screws, there somewhere around + - .002" - .004" per foot unless you move into extra accuracy big buck feed screws. Then there's screw straightness. That's almost always part of the much higher expense for real feed screws because it's not easy to strighten them within the accuracy required. I'll assume your screw is going to be used horizontally? How are you going to prevent it from sagging? Expecting all thread at a hardware store to do what your wanting isn't going to happen. Check out the websites such as Nook, SKF etc for the details about proper feed screws and nuts.etc for the information your missing about what's possible and what isn't. Off the shelf all thread rod doesn't have that level of accuracy because it doesn't need it.

 

[technocrat]

All the homemade cnc people use trapezoidal lead screw from ebay or dedicated suppliers. It's cheap and will be more that accurate enough for your purposes. Choice of sizes, steel or stainless plus brass or plastic nuts. Multiple lead options are also available for fast positioning.

 

[EPAIII]

6-32, 24 inches long?????

That is a wet noodle. If you are going to use something like that as a feed screw, you will need to have it under tension so it does not buckle. But that tension could easily stretch it, adding another source of error.

I would strongly suggest that if you need a feed screw that long, make it significantly larger than a #6 screw size. One half inch diameter comes to mind as a first stab at it. One quarter inch would be the absolute minimum that I would consider.

I am not familiar with the trapezoidal lead screw that technocrat speaks about. I do know that there are methods that have been used to compensate for screw errors. Some reading on the subject of precision lead screws (for lathes) may give you some ideas.

PS: If you think the accuracy of threaded rod is problematic, try comparing two or more of your tape measures (standard measuring tool for wood working) against each other. I did and was shocked. A sheet of plywood that has been stored indoors is probably more accurate at eight feet then they will be. Six or eight thousandths over two feet is not going to even be noticable in most wood working projects. That's less than the width of the lines on your tape measures.

 

[EPAIII]

I looked up "trapezoidal lead screw". Another name for them is Acme. Acme threads are very commonly used for lead or feed screws. But be aware that they also come in a variety of grades with different specs. And they often will have a left hand thread.

 

[Scottl]

Hopefully this answer won't be a problem re forum rules about machinery brands that can be discussed but IMO the most practical course for small, relatively inexpensive yet accurate lead screws would be to buy a suitable length 1/4-20 left hand leadscrew and brass nuts from Sherline. They are available as repair parts and come in various lengths for cross slide, milling machine X Y and Z, and the two different length lathes.

They also have a scheme to adjust backlash using a second nut as an adjustment.

I've used some of these machines in lab environments and while they are very small the positioning is far more accurate than ordinary threaded rod would achieve.

Also, IMO this thread should be moved to the woodworking subforum.

 

[woodguy2]

Yes this is more appropriate for woodworking section of this forum.

The rod would not be used as a lead screw. It would be embedded in a recess, and would not turn. That's why 6-32 could work, and it's small size is a benefit for my limited space application. And since it's for woodworking application, 32 is a magic number (unless you go metric, in which case 1 is the magic number).

The idea is the flip stop is forced into a true increment of 1/32" or 1 mm. An accurate tape measure (can I actually find those?) is fixed to the same base. So the commercial rod won't work well because it starts to force the stop into increments between those sought. One advantage for a woodworker is repeatability. So the commercial rod would actually work for most woodworkers, who'd be happy enough with absolute measurements within the nearest 1/32" or 1 mm, as long as those would be consistent.

The Jointech Smart Miter, defunct for some years now, used a system like what I'm envisioning. They actually allowed for rotation of the lead screw (10-32) for micro-adjustment. Theirs was integrated into a custom aluminum extrusion. I'm putting a micro-adjust mechanism on the arm of the stop. I'm looking to create an improved kit version of the Smart Miter for sale, -- hardware, some custom CNC created by me, with digital plan. Interestingly, a few months after I started what's turned into an extended R&D project on this, Woodpeckers announced they're coming out with a version of the Smart Miter for sale starting in September. Theirs loses some of the Smart Miter's most important functionality, the ability to make negative angles for matching miters on profiled picture frame stock, where you can't simply turn the stock upside down to cut the other side. Theirs does have larger cross cut capacity. My plan does both.

Thanks for the ebay lead screw suggestion. China can provide me with 8 mm D 1mm-pitch screw for $17 delivered for a 700 MM length. But I'm not sure that 1mm pitch means 1 mm thread spacing, given the multi-start nature of cnc lead screws. It costs $11 to ship a piece of rod from Mcmaster from an hour's drive away......

I think I might be able to use Incra's plastic racks (their improvement over the lead screw-as indexer design), cut in half for my application, and keep the whole thing domestically produced.

 

[sfriedberg]

woodguy2, I did something similar about 35 years ago. Bedded two pieces of threaded rod in a shallow channel as a fixed "rack" and stuck a third, short piece on the part I wanted to register along the rack. Yes, the long-range pitch accuracy of threaded rod isn't great. For many woodworking applications, the accumulated error isn't enough to matter. If it is, do like Jointech and make one of the two pieces adjustable so you can fine tune your cuts.
For multistart screws, lead will be the travel per rotation, but pitch will be the crest-to-crest distance, which is what you want. If you're not already aware, you cannot mix and match start count in this application. I.e., a normal M6 screw with 1mm pitch will not mesh nicely with a multistart screw with 1mm pitch.

 

[richard newman]

Ok, someone has to ask this - do you have a way to measure your wood parts to less than .005"?

 

[gbent]

Ok, someone has to ask this - do you have a way to measure your wood parts to less than .005"?

And if you do, what will they measure tomorrow?

For better grades of threaded rod visit Green Bay Manufacturing.

 

[neanderthal mach]

Not only measuring the parts, but actually positioning them to even do the cutting. Ok I get the idea of using what would require higher accuracy threads as some sort of incremental positioner to pre set the stop block. What are you going to use as a very accurate scale to ensure your at the length your shooting for during each movement? Unless I'm missing something obvious, why not cut all the mickey mouse trial and error bull shit out of the equation and just use something off the shelf that's already designed for linear measurement?

Something like the (retch) cheap Chinese linear scales. https://www.amazon.ca/Accurate-Digital-Readout-Milling-Machines/dp/B07DNYT9LB But to CMA, I'm not and never would recommend that junk even for wood working tolerances, that link is for illustration only. There accuracy claims are also no doubt completely bogus at that price. The same for it's long term durability as well I'd bet. If I wanted something at least ok, I'd buy the same with either the Mitutoyo or Starrett brand on it. But your not going to do this with anything found at a big box store unless it has the McMaster Carr, MSC, name on the building. Accuracy and just as important it's repeatability has very fixed costs, the more you want the more it's going to cost.

 

[DanielG]

Rolled screws have a lot of pitch error, but are very consistent. Is your actual thread pitch more or less than nominal. If it's less than nominal, you could tension the screw and dial in your thread pitch to be dead on.

 

[stephen thomas]

Ok, someone has to ask this - do you have a way to measure your wood parts to less than .005"?

I can and do. especially for set ups.
.005 is about the tolerance of a joint that is acceptable and one that is too sloppy.
Or acceptable snug, and so tight it will split something now or later.

Besides, i've seen you with your optivisor and stepping systems.
You know i've always looked up to you as a paragon of obssesiveness to aspire to.



smt
PS, to the OP, my experience is that un-plated B7 (high strength) tends to have a better thread form and slightly better PD contol. I never checked to see how the linear pitch compared. But all that stuff is rolled thread, so there is little pitch control beyond the rolling dies themselves. Similar but different to if you ran a long rod through a geometric die head and expected the pitch to remain uniform for any significant length.