PDA

View Full Version : Resolution calc help needed for Mach2



Moondog
01-30-2005, 10:44 PM
Hope you guys can help.

Just setting up my X drive in Mach2 and need some help in calculating the steps.

Specs are:

Servo Motor 1750 RPM.
Encoder Quadrature 500 (2000 steps?)
Reduction 27 to 1
Belt Drive: 300mm circumference

1/27 = .037037 RPM per RPM of motor.

Travel = 300 mm * .037037
= 11.11111 mm per rpm.

mm per step = 11.11111 / 2000 (steps)
= .0055555 mm per step.

What would be the max speed I could obtain with Mach2 running at 45,000 hz. Is that 45,000 pulses per second?....

I am getting somewhat confused with this.... I worked it out for the Ball screw drives...but this one...

regards

MrBean
01-31-2005, 06:27 AM
I don't think your calcs are correct. Is the axis belt driven. i.e. the moving part is directly attached to the belt.

My X axis is belt drive 5mm pitch. With a 3.6:1 belt reduction. 200 step stepper running 1/8 microstepping. effectively a 1600 step stepper.

This is the calculation for my setup:

The Drive belt runs on 26T pulleys each end so.

5mm pitch * 26T = 130
1mm / 130 = 0.00769230769
0.00769230769 * 3.6 reduction = 0.02769230768
0.02769230768 * 1600 steps = 44.307692288 steps per mm of travel.

OR 0.02256mm per step.

You don't need the belt circumference for the calculation, but you do need the number of teeth on the pulley the belt is riding on.

Hope this helps.

Regards Terry.....

Moondog
01-31-2005, 03:50 PM
Mr Bean..

It is a 10mm pitch 30 tooth pulley. The belt is attached to the gantry. i.e, the belt moves the gantry. The reduction is 27 to 1.

Using your calcs, would mine be...

10mm pitch * 30T = 300
1mm / 300 = 0.0033333333
0.0033333333 * 27:1 reduction = 0.089999999
0.089999999 * 2000 steps = 179.9999 steps per mm of travel.

OR ?????? per step.

cheers

ynneb
01-31-2005, 05:28 PM
Hey Frans, I never could work mine out mathematically, so I went for the trial and error method.
Draw 3 lines on your table, one at zero and one at 1000 mm and one at 2000mm ( As I know your table is at least this long)

Then zero your axis at the zero line.
Set the steps per unit at a rate that you know would be less than it should be. For example you could set it at 10,000 steps per mm of travel.
Then drive your axis 1000mm buy typing in g0x1000 (replace the x in this code with the letter of the axis, x, y, or z. )
Keep you finger near the emergency stop button and send the axis moving. If the axis passes the 1m mark then activate your emercency stop. This means you will need to increase the steps per unit in your motor tuning set up. If it doesnt reach the 1m mark then you will need to decrease the steps per unit. If for instance the travel is only 500mm then that means you will need to halve your current steps per unit setting.
You will need to do a few tests of different settings until you get it to travel the correct distance.
Once it is about right send your axis down to 2m and make any fine tuning in the steps per unit setting that is required.

I know this method does not sound as good as doing it mathematically but in some ways it is more accurate. Lets say your pulleys are not exacltly made to what you think they are, then all the maths you use will never get it right. The trial and error method gives real world results.

I know it sounds like a lot of work and fussing to do, but just remember you will only have to do it once. Then write down and record the appropiate figures if you ever need to set up mach2 again.

Caution. Because you are using a 27:1 reduction, you may find that you need to use the step multiplier switches on your Rutex cards. If you do switch them then you will have to reset your steps per unit in mach2. Do not just flick those switches and then send you axis travelling. It will result in the axis over traveling and crashing.

EDIT: Because I know you are using the same sized encoders that I use and that I only use a 3:1 reduction and it works ok, I would suggest that you use the step multipliers so that they multiply 4times. Any resolution greater than this would be serious overkill for what you are trying to do, and may result in the axis been driven too slow as a result of Mach2 not been able to keep up. Just a thought.

EDIT: The way to know deside on the amout of step multiplying, is to descide on the acceptable resoltion you require. For cutting timber as you are planing to do, I would think that 10 steps per mm should be more than adequate. That will mean you can cut down to 1/10 of a mm. It is highly possible that your machine does not have those sorts of tollerances built into is anyway, so picking a even higher resolution is just a waste. With this in mind, you work out what step multiplying you need in oreder to give you a figure that will drive that axis at a close number to 10 steps per mm.

I hope all this makes sence?

MrBean
01-31-2005, 06:23 PM
Mr Bean..

It is a 10mm pitch 30 tooth pulley. The belt is attached to the gantry. i.e, the belt moves the gantry. The reduction is 27 to 1.

Using your calcs, would mine be...

10mm pitch * 30T = 300
1mm / 300 = 0.0033333333
0.0033333333 * 27:1 reduction = 0.089999999
0.089999999 * 2000 steps = 179.9999 steps per mm of travel.

OR ?????? per step.

cheers



Well look at that. Your calc comes out exactly the same as doing it my way. Actually it's not my way. I used the formula from the Mach2 manual.

OR ?????? per step.

Just do: 1mm / 179.9999 = 0.00555555 Same as you got earlier.

Bennys suggestion is a good one. I'd start at 179.9999 steps/mm in Mach2 and then do Benny's tests to check if you're on the money, and make any adjustments from there.

Assuming you're running Mach2 engine at 45,000 That would give you a max of 59 IPM for that axis alone. If another axis was moving you'd require the step pulses to be more frequent than 45,000 a second. As Benny suggests, you may need to use the step multiplier on your drivers or go less than 27:1 reduction.

Regards Terry.....

ger21
01-31-2005, 09:22 PM
I just noticed in another thread that you have 880 oz servos. Correct me if I'm wrong, but doesn't 27:1 reduction give you 27 times the torque? Gear that thing 5:1 or 10:1 to get some speed. I'd think you would still have plenty of power. For a 4x8 machine, 59ipm is pretty slow.

Moondog
01-31-2005, 09:53 PM
Ger....

so you figure the max IPM will be 59?... looks like I am being limited by Mach 2...

I figured with a 1750 RPM reduced 27... gives 64.8 rpm's... each RPM gives 300mm travel.. thats 765 IPM?.... that's assuming full RPM even at half RPM I shuld still obtain 380 IPM.... do you think that is correct.?...

ynneb
01-31-2005, 10:07 PM
looks like I am being limited by Mach 2...

probably over zealous gear reduction is letting you down more. :) I wont skyte and say I told you so.....ah bugger it, yes I will :)

Anyway with the step multipliers it wont be mach2 that lets you down. You will get more speed than 59 inch per minute if you use them. Read back on my other post.

He he, I can imagine your heart pounding, your eyes wide open with fear, and the thoughts about all that work you have put into it . Relax mate you have not made a tragic mistake, your machine will work just fine as it is. Trust me.

Moondog
01-31-2005, 10:11 PM
Benny..

I had fiqured that at full RPM I'd get 700+ IPM... which is great.....

so I'll try the step multiplier at 4 to start with...... can always go to 10...

At least I won't lose torque by using the step multiplier....

Moondog
01-31-2005, 10:27 PM
lte me get this right...Mach2 max at 45,000 pulses per second.. that = 2,700,000 per minute... If my X drive is .005555 mm per step. then for say 300 IPM i need..

300 IPMinches = 7620mm per min... / .005555 = 1,371,737 steps per min..... half of mach 2 output... my Y is less than half the X..... so I still have 1,320,000 approx for my Y & Z...

does that sound right?..

ger21
01-31-2005, 11:01 PM
I just read Terry's post that said 59ipm. I didn't really pay that much attention to all the numbers. :D 700ipm sound much better.

Actually, with ~180 steps/mm, and 25.4mm/inch, you get 4572 steps/inch. 45000 steps/sec / 4572= ~9.85inches/sec.
9.85 x 60 = ~590 ipm. That's what Mach2 will limit you to.

I don't know what you're using for a spindle, but with a router you'll probably have to take light passes at that speed, or you'll stall it. To cut that fast, you'll need good quality spiral tooling, too.

ger21
01-31-2005, 11:03 PM
lte me get this right...Mach2 max at 45,000 pulses per second.. that = 2,700,000 per minute... If my X drive is .005555 mm per step. then for say 300 IPM i need..

300 IPMinches = 7620mm per min... / .005555 = 1,371,737 steps per min..... half of mach 2 output... my Y is less than half the X..... so I still have 1,320,000 approx for my Y & Z...

does that sound right?..

I believe you get 45,000 steps for EACH axis, or 270,000 steps/second total if using 6 axis. :)

Moondog
01-31-2005, 11:13 PM
Ger... so by the looks of it.... even with no Step Multiplier reduction I should still get reasonable IPM's?...

Initially I'll only be using a Makita router. Hope to then upgrade to a HSD Spindle.

ger21
02-01-2005, 07:26 AM
If all your calculations are right. I only looked at the 180 steps/mm, I didn't check the rest. We've got 2 9HP HSD's on our machining center at work. I cut 3/4" plywood and mdf in 1 pass at 400ipm. (The machine won't go any faster, or I would :( )

MrBean
02-01-2005, 07:41 AM
I caculate:

0.0055mm per pulse
multiply by 45000 pulses/second (0.0055*45000) = 247.5mm/second

Divide by 10 to get cm/second = 24.75cm/second

Divide by 25.4 to get Inches/second = 0.974 Inches/second

Multiply by 60 to get Inches/minute = 58.5 IPM

Maybe I'm going mad. (chair)

Regards Terry.....

MrBean
02-01-2005, 07:46 AM
lte me get this right...Mach2 max at 45,000 pulses per second.. that = 2,700,000 per minute... If my X drive is .005555 mm per step. then for say 300 IPM i need..

300 IPMinches = 7620mm per min... / .005555 = 1,371,737 steps per min..... half of mach 2 output... my Y is less than half the X..... so I still have 1,320,000 approx for my Y & Z...

does that sound right?..


300 inches is 7620cm, not mm.

Regards Terry.....

ger21
02-01-2005, 07:51 AM
Divide by 10 to get cm/second = 24.75cm/second

Divide by 25.4 to get Inches/second = 0.974 Inches/second

Maybe I'm going mad. (chair)

Regards Terry.....

Terry, divide centimeters by 2.54 to get inches. Multiply your ipm x 10

MrBean
02-01-2005, 07:53 AM
Yep I'm going mad........HAHA.

MrBean
02-01-2005, 07:55 AM
I'm out by a factor of 10. Ooops. (chair)