[Fodder1]

Hi guys and girls,

It's been a very long time between CNC drinks for me, but I am back with a JGRO with linear rails and T5 belts.

I have the JGRO all together and aligned, but I am having real trouble with the setup of the belts in mach3.
I was wondering if anyone can help me?

I am using T5 timing belt from bilbycnc.com and the T5 pull with 12teeth. (Overall length is 21mm, hub diameter is 11m.Dimensions: 23x23x21mm)

I can't seem to get the setting for steps per mm correct and all the "guides" I can find don't really do metric or I just can't make it work, so I was wondering if anyone has some ideas for me!

Chris.

[mactec54]

fodder1

It is very easy to set up the step/per, make sure you have the control set to mm ( Config, then Select Native Units ) Is yours set to what you want to use ( MM )

Then go to the settings page (Alt6)

Just above the RESET Button you will see Axis Calibration, Set Steps per unit, just follow the instructions & you will have it set up in no time at all, there are videos also on You tube how to do this

The set up is the same to do for Metric or for Inch, you are just using a different unit of measurement to set your machine

[Fodder1]

Hi mactec,

Thanks for taking the time to respond!

Yeh I have used that feature of Mach and was not able to get it to work properly, we'll sorta.
Here is what is happing:

I have a test part which is a 100mmx100mm square and my JGRO is 350x700 or there abouts.
But when I run the part the gantry crashes into the limit switch on the other side of the table. But the DRO on Mach is only saying about 10 mm had pasted.

So I ran that calibration and to get it to move 10mm I had to put into that box 0.01 which to my way of thinking is 1mm.

It seems like it is out by a factor of 10. So if i type "1" (one, no quotes) into that box that opens up using the axis calibration tab, and yes I have set it mm, that axis should then try and move 1mm?
Because at the moment I was 1:10 movement.

I was thinking that my initial setting were so way out that I was out by 1:100, after running that wizard I got it back to what looks like 1:10.
Then I started looking at the formulas and was getting even more confused, hence the post.

I am using "standard-ish" steppers 1.8 deg 200 step per rotation, but if my maths is correct I am going to need only 6 steps per mm.

As it stands currently in the "motor tuning" area I have a steps per at 66.66666666666 which was calculated by Mach, but its still wrong.
Can you tell me if I type "1" (one, no quotes) into that axis calibration window it should equal 1mm.
Because if that is the case I should be able to still use it........

Chris.

[LeeWay]

Here is the way to dial it in by hand a bit before you get to the calibration screen.
Place a ruler on the table. Start at the end of it. tell it to move like 5mm. See how far it actually goes. (Done in the MIDI screen.)
If it is close enough to make that move on the table, then do the calibration.
When it is in inches, you just use decimal points, but 1 is 1.00". I am not sure if the metric scale uses MM or CM, but I would think it is MM.

The other way is to do the math for everything you have installed on that axis. That should get you close, then the calibration button will zero you in. On some of my machines, my step count is 5 places below the decimal point, so extremely accurate.

[mactec54]

Fodder1

You are not doing the set up correct

When it asks you how far you want it to move put in say 50mm, after it has moved you measure how far it has moved, you then put that number (how far it moved) in the box, & check again, you will get the correct step per mm if you do it like this

To set it rough to start you can use a tape measure, then you will need to do it with a dial indicator

Here is a video, this is in inch, but you can use mm just the same
https://www.youtube.com/watch?v=XkO5tc-jSxw

[bushwakka]

Chris,
A 12 tooth T5 pulley will travel 60mm in one revolution (5mm per tooth on the belt, 12x5=60)
Assuming that it is directly mounted to the motor shaft you will have 200pulses per 60mm of travel or 3.33pulses per mm.

Cheers
Bushwkka

[mikie]

When i set my unit up i used pure maths. Work out the pulley radius and work out the circumference. This is the distance travelled in one rotation.

2 * PI * R = circumference; or PI * Diameter = circumference

Then apply this as you would for any setup to work out the steps, or specifically as stated by bushwakka. Add pules * microstepping and divide by travel.

[Adverse Effects]

how can you get Mach3's calibration wizard wrong?

you say move <pick a axis> <say a distance> (i normally use 1/2 the max distance the axis can move) then tell Mach3 when it asks how far did it really move

once you have done that for all 3 axis i then redo it again

using about 90% of the max distance the axis can travel

(a error of 0.1mm over 600mm is way more accurate than a 0.1mm error over 10mm travel)

[jwhitten]

Hi mactec,
So I ran that calibration and to get it to move 10mm I had to put into that box 0.01 which to my way of thinking is 1mm.
It seems like it is out by a factor of 10. So if i type "1" (one, no quotes) into that box that opens up using the axis calibration tab, and yes I have set it mm, that axis should then try and move 1mm?
Chris.

I have exactly the same problem, and I went through the same calibration technique. Originally I set it to inches and everything was working fine. But then I decided I would prefer to work in metric so I reset everything to metric. I changed the motor units and the ALT-U units in the settings screen. But now when I load up a g-code file, everything is multiplied by 10. I have gone through the Axis Units setup a number of times to try and correct it, but every time I do it, I always end up 10 times off. When I type in the number of CENTIMETERS I read along the ruler into the set axis window that is supposedly asking about MILLIMETERS, it works fine. Personally, I think Mach3 is talking about Millimeters but expecting Centimeters. The fix appears to be easy though, just divide the motor values by 10 and all is well.

John