Dang it, I finally had an answer to a question and he beat me too it. I got lost on the jet copter video.
Chris
No, you now have it set to 1/10th microstepping.
look at the left hand column on the chart.
The 2nd column is the steps/per if just the 200 steps/rev motor were used, they don't multiply also by screw tpi for that number.
You want
5-off
6-on
7-on
8-on
for 1/2 microstepping.
The heli looks bad a$$.
Hoss
Gosh, you've... really got some nice toys here. - Roy Batty -- [URL]http://www.g0704.com[/URL]
Dang it, I finally had an answer to a question and he beat me too it. I got lost on the jet copter video.
Chris
LOL.. here's one for you Chris.
I don't think I understand what the Pulse/rev on the driver means... I set the switches back and the steps in the motor tuning to 2000 an now it dead on for the Y and z axis and off by 1/32 on the x axis (is that normal?)
According to the table on the driver... the switches are set to "400". That doesn't click with me.... but it works now.
OK... wait a minute... something just clicked....
"400" equals "400 pulses per revolution. Since we're using "half steps" then that means it's set to "200" pulses per revolution which is the specs for the stepper
Or maybe not...
ok, 360 degrees divided by 1.8 degree steps equal 200. Half step would make it 400.
Now that X being off 1/32 is odd. Repeatable? Over what distance?
Chris
Check mach to make sure you motor settings are the same amd make sure you press "save axis" button.
Could your mill have that much backlash? Ballscrews or threads? You can run backlash comp, but you need to figure out why it is such a large number. Maybe something is loose under the table.
Recheck wire and colors (Previous duh disclosure)
Chris
while not seeing the mill and how its constructed, one thing to look for. I have seen the shafts inside some stepper motors move in/outward inside the stepper motor can itself. Just be aware of it especially if there are no thrust bearings and the machine is really tight.
I checked the machine and there does seem to be any "looseness" that would account for this. I tired playing with backlash and while it came close to 1 inch in either direction it still fell short. When I went a full 5 units on either side it would be out back to the original amount.
The y and z axis could be out too. I haven't measured them as strictly has I am the x axis.
Could it be that the screw thread is off by a small degree? Maybe being 5.005 tpi instead of 5
Does that happen?
Looks like I found a solution. But is this normal?
I set the steps per in the motor tuning to 2032 from 2000 and it seems to be constantly correct up and down the axis. I need to do further testing, but so far so good.
I'll need to check the other axis as well. Hey could be off as well.
Makes perfect sense if you have metric screws.
400 steps per rev
2032/400 = 5.08 revolutions
5.08*5 mm = 25.4 mm = 1 inch
It's been a while since I bought the parts, but I thought they were imperial. If not would the steps per remain 2032 if I switch the units to mm?
And the answer is
80 steps per for metric
For future generations of newbies like me
Divide the inches step per by 25.4
Multiple velocity by 25.4 and aceleration by 25.4
It was a year or so since I bought the parts. I must have bought metric screws?
Glad you got it figured out, but to answer your question, yes the axes can be off of actual math count for steps.
This is why it is important to calibrate each axis. Mach has a feature that really helps with this.
Lee
2032 is what I use in my machine for my Metric ballscrews. Glad you have figured it out. Now you can cut some metric stuff, too bad you didn't get inch ballscrews!!
Mach takes care of all that stuff via your code. Mach knows it takes 25.4mm to make an inch so it will move the right amount and you dont have to switch a thing.
Chris
Thanks. Getting this all straight in my head takes a little time. This is a project I've been wanting to do for the past 20 years!
Chris, so you have metric screws and run them at 2032? Does that mean your config is set to inches?
Yes, setting the 2032 makes my metric ballscrews look like inches to Mach. It took me a little while to get it straight in my head as well, so don't feel lost. I have just within the last couple of weeks gotten my head around some of this, by no means all of it.
Most of your code will be determined by what software you use. If you are more comfortable in metric, then you will need to backtrack a little and tell Mach you want to set up your system on metric. Code will do this just as easy by using a line at the start of the code. G20 is inches and G21 is Metric. Since we mostly use inches here most of the information out there is on that system.
The way I see it ( and one of the older hands will correct me if I am wrong) is Mach doesn't care what pitch each screw is, but it does need to know the information you put in on your set up. If you use inch on one axis and metric on the others, just make sure you set up the metric axis with the proper amount of steps. 2000 for 5 tpi and 2032 for the 5mm pitch. 4mm would be different just like any other tpi would be. Mach wants to move the axis one inch, how many steps tells mach how to get there. If you are running in metric mach still knows that 1 inch is 25.4 mm. Set you machine and Mach together and then your code software will work it out.
Chris
Thanks Chris,
That did a lot to clear it up in my mind. I think what I need to do now is create some code for something easy ( a plate with a round hole in it) and see what happens.
Cheers.