Hello Guest it is March 28, 2024, 07:45:51 AM

Author Topic: Steps per calculation  (Read 4438 times)

0 Members and 1 Guest are viewing this topic.

Steps per calculation
« on: May 12, 2015, 09:49:13 AM »
Hi
I'm having issues getting my steps per inch right. I helped a friend do a mill retro fit a year ago with mach 3 and there was a little utility to calibrate it.
the mach4 manual has a formula using your encoder resolution (2500)and the ball screw pitch (.20 per turn) I put in 12500 on the motor tuning but I don't get 1 inch of movement when I mdi 1 inch. Somewhere in the  .85 range, I can't remember exactly what it was this morning.  So would I multiply  .85 (or whatever) by my steps (12500)? If this is a "mathematical relationship" why doesn't it match with the calculation? I'm using teco servos and drives with a Hicon Integra. I hope this doesn't sound too unintelligent I'm pretty new at this stuff, and not very good at expressing myself.
Re: Steps per calculation
« Reply #1 on: May 12, 2015, 10:16:14 AM »
Do you have any belt/pulley reduction to include ? Is it direct coupled ?
Any scaling/gearing set in the drives ?
2500 count encoders should = 10,000 in quadrature.

Russ
Re: Steps per calculation
« Reply #2 on: May 12, 2015, 10:23:42 AM »
X and y are direct coupled. z has a belt but it's 1 to 1. There is a parameter for gearing but I should have it set for 1 to 1. I'll read that section of the drive manual more closely.
sorry for the noon question but what do you mean by quadrature?
« Last Edit: May 12, 2015, 10:26:37 AM by Matt1911 »
Re: Steps per calculation
« Reply #3 on: May 12, 2015, 11:08:41 AM »
Counting both leading and trailing edges of both channels of a quadrature encoder will quadruple (x4) the number of pulses per revolution. As a result, 10,000 pulses per turn can be generated from a 2,500 PPR quadrature encoder.
Most controllers count this way.
Russ
Re: Steps per calculation
« Reply #4 on: May 12, 2015, 12:36:02 PM »

Ok I'm going to have spend some time with google to figure this out, after I clean the gray matter off my keyboard. So are you saying I should use 10000/.2=50000 in my motor tuning?
just if my controller counts quadrature style?
Re: Steps per calculation
« Reply #5 on: May 12, 2015, 12:52:42 PM »
Man, I really dig that smiley .... thanks.
It will be quadrature, for sure.

IF your encoder is in fact 2500 PPR, yes.   Mine are 2000 PPR making it 8000/.2=40000 here.
Make sure your screw is not 5mm as they are quite close to one another and can throw you off a tad but not as much as you are seeing.
Are you sure that your encoder is not 500 PPR ? Then quad would be 2000 / .2= 10000.
Re: Steps per calculation
« Reply #6 on: May 12, 2015, 12:58:36 PM »
Ok I found it in my servo manual it is quadrature. I'm sure they are 2500 so I'll change it to 50000 and see what I get. I'll have to wait till tonight when I get home. I'll report then what I find. I really appreciate you taking time to help me.
Re: Steps per calculation
« Reply #7 on: May 12, 2015, 01:20:45 PM »
Well ... 10000 (2500 ppr) / 12500 (3125 ppr) does = .8
Hope you find it.
Will check back later.
Good luck.

btw ... it's hard to tell which end that explosion is coming from. With my wife, it could be either.  ;D

Regards,
Russ
 :)
Re: Steps per calculation
« Reply #8 on: May 13, 2015, 12:36:11 PM »
Well I changed it to 50000 and that got me within a few thou. I tweaked with it a bit and ended up at 50120 on the x and 50130 on the y. I'm not sure why it would be off, wear maybe. I  checked to see if it was a 5mm the dial said inches. But with the indicator it was like .1989 or something around there, so in between 5mm and .20 anyways it seems to be working.

No My Wife Is Pretty good but if I don't get this thing done and stop staying out in the shop till the wee hours she might start putting on the pressure. Ha ha course when it's done I'll probably be out there all night machining ;D
Re: Steps per calculation
« Reply #9 on: May 13, 2015, 02:16:36 PM »
 I'm not sure why it would be off, wear maybe.

You might do a comprehensive backlash check on each axis.
Be sure to do all calibration moves in the same direction to minimize any errors due to the backlash if any is present.
There is most likely some there, maybe minimal but is almost always a factor.

Glad you're getting closer.  :)

Ahhh ... projects for momma. That keeps them happy.  :)

Russ