for engraving pcb which is a good thread size for the lead screw?

[NickLatech]

I know that the smaller the thread size the more resolution i should be able to get. What is a good thread size for a lead screw?

Nick

[ViperTX]

Originally Posted by NickLatech I know that the smaller the thread size the more resolution i should be able to get. What is a good thread size for a lead screw? Nick

Well, let's look at steppers (without microstepping, etc)...say that you have a stepper that has a 1.8 degree movement per step, so one complete revolution of the stepper motor would require 200 steps if you were coupled 1:1 to your lead screw which has 24 threads per inch....one complete stepper revolution (200 steps) would result in the leadscrew moving whatever was attached to it, 1/24 of an inch (0.04166) or each step would be 0.000208 inches. So, let's say that for some reason you lost a step ocassionally....well the first step lost would shift your circuit by 0.000208. If you had pads for a Dual In Line socket which has a center to center pin spacing of 0.100 inches and each pin is 0.036 inches wide and you were using a 0.042 inch diameter hole in a 0.062 diameter pad you could afford a few lost steps before you run into problems.....draw yourself a picture it will help.

[Jan]

how about posting the formula or link for computing the above.

[ger21]

Originally Posted by ViperTX Well, let's look at steppers (without microstepping, etc)...say that you have a stepper that has a 1.8 degree movement per step, so one complete revolution of the stepper motor would require 200 steps if you were coupled 1:1 to your lead screw which has 24 threads per inch....one complete stepper revolution (200 steps) would result in the leadscrew moving whatever was attached to it, 1/24 of an inch (0.04166) or each step would be 0.000208 inches. So, let's say that for some reason you lost a step ocassionally....well the first step lost would shift your circuit by 0.000208. If you had pads for a Dual In Line socket which has a center to center pin spacing of 0.100 inches and each pin is 0.036 inches wide and you were using a 0.042 inch diameter hole in a 0.062 diameter pad you could afford a few lost steps before you run into problems.....draw yourself a picture it will help.

This is making a big assumption that a homebuilt machine can actually move in accurate .0002 increments. I don't believe allowing for a few missed steps here and there is a good idea. Chances are, if you miss a few steps, you're going to miss a lot of steps. A properly setup stepper system should not miss any steps. If you're missing steps, you should correct the problem. Don't count on your leadscrew to mask the problem.

As for leadscrew choice, 1/2-10 Acme is a pretty universal leadscrew. It works very well on many homebuilt machines. And still gives you a theoretical resolution of .0005. It's cheap, too.

[ViperTX]

Twas an example. I suspect that most everyone knows or has experienced a stepper system that does not move as accurately as you expected. What causes these problems...the stepper encounters a condition where it is not able to move a step these conditions can be that the required torque to move a step exceeds what the stepper can provide, you are attempting to step at a frequency that the stepper, the hardware or software cannot attain.

Let's assume that you've requested your stepper to ramp up from a dead stop to some maximum speed of say 16 inches per second (so with our 1/2 10 ACME screw and the 1.8 degree stepper that would be 16 inches per second X 10 threads per inch per revolution X 200 steps per revolution = 32000 steps per second). Given this a parallel port without interference from Windows could handle this, no problem for properly written software, the ACME at 10 revolutions per second is 600 RPM, which shouldn't be a problem, if the stepper has the torque to overcome the friction that the nut is placing on the ACME due to it's load and the normal friction between the nut and ACME screw then we don't have a problem.

We encounter difficulties when we ramp up to full speed and then ramp down to home into a movement of only 0.100 where we are going to change direction. So, software attempts to do this is several ways (ramp up speed, ramp down speed, compensation factors....those of the little things that we feed the software when we see something that is not correct, or to compensate for the diameter of the cutting tool or the know inaccuracy of the leadscrew, etc.

In reality we use many things to compensate (or mask) anomalies...I could go on, but I'm hungry.

Hope this helps!

[ger21]

I'm guessing you meant 16 inches per minute? Which is only 160rpm, right?

[ViperTX]

Nope, actually 16 inches per second...remember it's an example...you have to look at other factors like material removal rates, depth and size of cut, hardness of material, etc.

[ger21]

Well then it's 9600 rpm for 16ips. Or are they 2 different examples?

[ViperTX]

Originally Posted by ger21 Well then it's 9600 rpm for 16ips. Or are they 2 different examples?

Oops you are correct we have a problem...cool you followed the math and the thought process.

So 16 inches per second X 10 Revolutions per inch x 60 seconds per minute = 9600 Revolutions per minute. This is beyond the theoretical limit for this size of ACME screw with a 20 lb load by about 1000 RPM. As you all have seen write down everything, draw pictures and check the mathematics and the assumptions. Anyway, you can go to http://www.nookindustries.com and look at the ACME screws and at the bottom there are links that will allow you to determine critical speed of an ACME screw, you can even calculate the required torque needed to move loads. I hope this helps!

[ger21]

Originally Posted by ViperTX This is beyond the theoretical limit for this size of ACME screw with a 20 lb load by about 1000 RPM.

And at least 3x the useable speed of a stepper motor.

[Jan]

and thats why i suggested to post the formula.

[ger21]

Jan. Most steppers are 200 steps per revolution. So, if you have, for example, a 10 turns per inch leadscrew, the motor turns 10 times to move 1 inch. 10 * 200 = 2000 steps per inch. 1/2000 = .0005 per step