Lead screw selection question

[TLARengineering]

After much lurking and researching, I believe I'm ready to order my lead screws for my CNC Router build.

I'm building a modified (read as larger) "book" CNC and plan to use 1/2-10 ACME lead screws for linear travel. (I'll post pics of my build progress once I get back to my home computer.)

I've read about the virtues of single start versus multiple start screws (my take: faster travel with greater torque on multiple start screws) so I think I've settled on 1/2-10 5 start screws. Should I get this for all three axis'? Is it even "necessary" to use the 5 start screw or should I just stick with the cheaper single start?

My intent for the machine is the creation of signs/plaques...wood only, no metal. And, if I really get into this, I'll use the machine to create the next one too.

So, if I could get some advice from some of you who have been there before, I'd really appreciate it.

Thanks,

-Chris

[CarveOne]

One start is going to be painfully slow on anything bigger than 18" x 18" working area. Definitely go with 2 start or 5 start. I prefer 5 start myself.

If you use 1 start it is guaranteed that you will buy new screws, matching adapters, and anti-backlash nuts after you use the machine for one or two projects. I used 1 start on my first 25" x 37" machine and later bought the 5 start hardware. I won't use 1 start anymore. It is much better than using all-thread though.

Use the same screws on your Z axis if you can. Especially if you have a left over piece that is long enough anyway.

CarveOne

[TLARengineering]

CarveOne,

That's just the sort of validation I was looking for. I'd hate to spend all the dough on the lead screws and anti-backlash nuts if it wasn't "necessary".

Now I'll go figure out what length I need and order my screws (McMaster-Carr) and my nuts (probably DumpsterCNC).

Thanks again,

-Chris

[CarveOne]

Try to tailor your machine size to take best advantage of 6' or 3' rod lengths and also minimize how much wasted rod you have left over while meeting the size machine you think you can live with. Left over rod is not a bad thing though. Anything 10" and longer can be used for another Z axis or on a smaller machine later. There's always someone who will take it off your hands if you don't need it.

If you use a heavy 3-1/4 hp router and heavy Z axis slide assembly, you may want to use 2 start on the Z axis to help prevent the Z from dropping downward when step motor power is removed. In severe cases a braking system can be used. This is not likely to be a problem for most DIY builds but keep it in mind.

The DumpsterCNC couplers and nuts have worked fine for me.

CarveOne

[jsheerin]

I agree on the 5 start - that's what I have.

However one of your initial statements was somewhat incorrect. When you go from 1/2-10 single start to 5 start, you have less force for a given torque input to the screw (5 times less) in the simplest case. But because of the way torque curves for stepper motors work and the fact that higher lead screws are higher in efficiency, you would probably end up with more force with the higher lead screws at higher speeds (everywhere above about 40ipm in the example I just ran through my calculator using 425oz*in motors and above 110ipm using 282oz*in motors).

[TLARengineering]

Originally Posted by jsheerin I agree on the 5 start - that's what I have. However one of your initial statements was somewhat incorrect. When you go from 1/2-10 single start to 5 start, you have less force for a given torque input to the screw (5 times less) in the simplest case. But because of the way torque curves for stepper motors work and the fact that higher lead screws are higher in efficiency, you would probably end up with more force with the higher lead screws at higher speeds (everywhere above about 40ipm in the example I just ran through my calculator using 425oz*in motors and above 110ipm using 282oz*in motors).

I think my brains just leaked out of my ears! LOL Okay, it appears that I need to go read some more as I only understood about half of what you're saying. (or perhaps it's the cold medication getting in the way of my brain)



-Chris

[jsheerin]

Well, if you really want to, but the bit to take away is you definitely want a multi-start screw.

Basically, your motors produce some amount of torque. When you change the pitch of the screw, you change how much pushing force that is converted into. A screw that moves the nut farther for one rotation generates less force for a given torque from the motor. The complicating bits (as I mentioned above) are the efficiency of the screw / nut combo and how the torque of a stepper motor varies with rpm. Generally screws are more efficient when they have more starts and a faster lead (move farther for one revolution). So you throw away less of the motor's torque to friction. For the motor, generally stepper motors have a holding torque when they're not turning. That's the number you usually see quoted for a motor. Once the motor starts turning, the torque goes down. The faster it goes the more the torque goes down. But different motors have different characteristics for that. So when you combine all that together, yes, it gets a bit complicated and the easiest way to figure it out is to calculate what actually happens (which is what I did and quoted some numbers from - I already had the numbers handy since it's all stuff I looked at for my router).

[lgalla]

Jsheerin
When using 5 start or 5to1 gearing with a stepper does it result in lower torque than the motors rating? If I look at a speed/torque chart,steppers appear to have little torque above 1000rpm
Thanks
Larry

[jsheerin]

Larry,
Using a particular screw doesn't change the torque the motor applies to the screw. What changes is the force the screw applies to the nut for a given torque applied by the motor. If you gear the motor so that the screw turns faster, then you have decreased the torque applied to the screw (even though the motor still applies the same torque to the input of the gearbox). So if you compared a 5 start screw to a single start screw with 1:5 gearing (making the screw turn faster via gearing), in a very basic comparison ignoring reality, you'd get the same result (the same force at the nut at the same linear speed of the nut). However when you take efficiency into account, the single start screw would apply less force to the nut (5 start is around an efficiency of 0.77 while single start is more like 0.4. Additionally the efficiency of the gear drive would reduce the overall efficiency even more of the single start - if it was 0.9 for the gear reduction, then the over efficiency would be 0.36). Additionally, plastic nuts have problems if the pressure * surface velocity (the screw surface moving past the surface of the threads of the nut) gets too high. I think you melt the nut. So if you spin the single start faster, you're going to run into trouble with that much sooner than with the 5 start. You'll also have more trouble with the screw whipping with the single start as the critical speed stays the same for both screws but now you're trying to spin the single start faster. So the 5 start screw is the way to go, no question.

[lgalla]

jsheerin,
Thanks for the explanation,very informative.
Larry

[jsheerin]

I suppose I should point out that the downside to the 5 start versus single start is that you reduce your resolution by 5 times. For a step pulse out of the computer, your machine will move 5 times farther. However that isn't really a problem on the typical wood router - you're still around 0.0025" / full step and with micro stepping you can get slightly better.

[kentonsj]

In case you were wondering about the formula jsheerin used, or at least an approximation:

(step ratio/step resolution angle)*(360deg/rev)*(rev/in final drive) = steps per inch.

For 5 start 1/2-10 the reduction is 2 turns per inch so the formula for full steps is:

(1/1.8)*360*2= 400 steps per inch

1 inch/400 steps = 0.0025 inches.

1/2-10 acme 1 start is 10 turn per inch, which gives a resolution of 0.0005 inches, exactly 5 times better resolution, what you would expect.