×
Forum Header
step/rev problem
12 Mar 2015 09:45 #56724
by johnl2
step/rev problem was created by johnl2
Hello everyone, and thanks for all of your help in the past. I just purchased this kit:
www.automationtechnologiesinc.com/produc...-stepper-drivers-etc
I am trying to set the drivers up for 5TPI pitch. So .2" should be one turn, right? Well I put 5 in stepconf wizard for pitch and it just doesn't work. I have the dip switches on the drivers set to default. Also, when I change the microstepping, that also changes the turns on the motor, is that right?
Anyway, I decided to experiment a little more and moved the dip switches to the first setting, still with 5 as my pitch in stepconf, and it works now! My question is, if I have it set as default, should't it go by the software? Should I be utilizing software settings WITH driver settings? And my second question is, what's up with changing the microstepping? Why does it affect how many times the motor spins on a given distance command?
Thanks again for all of your help,
John
www.automationtechnologiesinc.com/produc...-stepper-drivers-etc
I am trying to set the drivers up for 5TPI pitch. So .2" should be one turn, right? Well I put 5 in stepconf wizard for pitch and it just doesn't work. I have the dip switches on the drivers set to default. Also, when I change the microstepping, that also changes the turns on the motor, is that right?
Anyway, I decided to experiment a little more and moved the dip switches to the first setting, still with 5 as my pitch in stepconf, and it works now! My question is, if I have it set as default, should't it go by the software? Should I be utilizing software settings WITH driver settings? And my second question is, what's up with changing the microstepping? Why does it affect how many times the motor spins on a given distance command?
Thanks again for all of your help,
John
Please Log in or Create an account to join the conversation.
12 Mar 2015 11:56 #56726
by cmorley
Replied by cmorley on topic step/rev problem
We would need to know the other relevant settings in stepconf.
motor steps
microstepping
pulley teeth
leadscrew pitch
I didn't find any info on the drives microstepping settings on the web site.
Can you tells us what the dip settings set the microsteppings to?
Chris M
motor steps
microstepping
pulley teeth
leadscrew pitch
I didn't find any info on the drives microstepping settings on the web site.
Can you tells us what the dip settings set the microsteppings to?
Chris M
Please Log in or Create an account to join the conversation.
12 Mar 2015 13:12 #56727
by johnl2
Thank you for your reply Chris. Here is the manual for the driver:
www.automationtechnologiesinc.com/wp-con...2014/05/KL-8070D.pdf
Motor steps are 200 (1.8*)
Microsteps are set to 4
Pulley Teeth 1:1
Leadscrew pitch is 5TPI
Replied by johnl2 on topic step/rev problem
We would need to know the other relevant settings in stepconf.
motor steps
microstepping
pulley teeth
leadscrew pitch
I didn't find any info on the drives microstepping settings on the web site.
Can you tells us what the dip settings set the microsteppings to?
Chris M
Thank you for your reply Chris. Here is the manual for the driver:
www.automationtechnologiesinc.com/wp-con...2014/05/KL-8070D.pdf
Motor steps are 200 (1.8*)
Microsteps are set to 4
Pulley Teeth 1:1
Leadscrew pitch is 5TPI
Please Log in or Create an account to join the conversation.
12 Mar 2015 20:21 #56740
by andypugh
A standard stepper motor has 50 poles.
There are two phases which means that there are 4 possible ways to power one winding at a time. (because you can power each winding with two polarities)
So, if you only power one winding at a time you get 200 steps per rev.
Every time you send a step to the driver it changes the winding excitation to the next pattern in the sequence (whether left or right depends on the direction pin).
If you switch to half-stepping then the driver will sometimes power both windings. Now there are 8 possibilities, or 400 steps per rev.
Higher and higher micro-step ratios go to more and more patterns (10% A+ 90% B- for example)
The number that LinuxCNC uses is the number of step pulses is needs to send for one unit of axis travel. So for a 5tpi screw and single stepping that is 1000 steps per inch. At 4x microstepping 4000 steps per inch, and so on.
Stepconf tries to do the calculations for you. It takes pulley ratios and microsteps and such into consideration, but the output to the HAL file is just a single number, the number of steps to move 1 inch or 1 mm.
If you change the microstep ratio in the driver then it takes more or less pulses to move the axis that distance. So the hardware setup and the software setup need to agree.
Replied by andypugh on topic step/rev problem
Why does it affect how many times the motor spins on a given distance command?
A standard stepper motor has 50 poles.
There are two phases which means that there are 4 possible ways to power one winding at a time. (because you can power each winding with two polarities)
So, if you only power one winding at a time you get 200 steps per rev.
Every time you send a step to the driver it changes the winding excitation to the next pattern in the sequence (whether left or right depends on the direction pin).
If you switch to half-stepping then the driver will sometimes power both windings. Now there are 8 possibilities, or 400 steps per rev.
Higher and higher micro-step ratios go to more and more patterns (10% A+ 90% B- for example)
The number that LinuxCNC uses is the number of step pulses is needs to send for one unit of axis travel. So for a 5tpi screw and single stepping that is 1000 steps per inch. At 4x microstepping 4000 steps per inch, and so on.
Stepconf tries to do the calculations for you. It takes pulley ratios and microsteps and such into consideration, but the output to the HAL file is just a single number, the number of steps to move 1 inch or 1 mm.
If you change the microstep ratio in the driver then it takes more or less pulses to move the axis that distance. So the hardware setup and the software setup need to agree.
Please Log in or Create an account to join the conversation.
12 Mar 2015 22:39 #56749
by johnl2
Replied by johnl2 on topic step/rev problem
Thank you very much for your reply Andy, I was hoping you'd come in and answer my Q about that! So then with the way I have it setup, (driver and software working together, driver NOT switched to default), is ok then? So long as I get my output correct?
But this leaves me with one problem. . . what is my microsteps if both are set to something different? It's set to 400 on the driver and I have it set to 4 in emc2.
Thanks,
John
But this leaves me with one problem. . . what is my microsteps if both are set to something different? It's set to 400 on the driver and I have it set to 4 in emc2.
Thanks,
John
Please Log in or Create an account to join the conversation.
12 Mar 2015 23:25 #56750
by andypugh
If the driver says "400" then LinuxCNC probably needs to be set to 2.
Replied by andypugh on topic step/rev problem
But this leaves me with one problem. . . what is my microsteps if both are set to something different? It's set to 400 on the driver and I have it set to 4 in emc2.
If the driver says "400" then LinuxCNC probably needs to be set to 2.
Please Log in or Create an account to join the conversation.
13 Mar 2015 00:30 #56756
by johnl2
Replied by johnl2 on topic step/rev problem
If I set it to 2, then it doesn't make the correct amount of turns for a given command. It's off. This is the only way I've found that it turns one whole turn for .2". I was fiddling with it ALL day yesterday. I was under the impression that if I use the software to tune the driver then the driver should be set to default. That's what the instructions say. But when I have it set to default, I cannot get the correct movement via software. This has been the only way that works. My question is, does it matter? Will something get ruined? And what is my microstepping set at?
Thanks Andy,
John
Thanks Andy,
John
Please Log in or Create an account to join the conversation.
13 Mar 2015 00:33 #56757
by johnl2
Replied by johnl2 on topic step/rev problem
I don't know if there's a way to bypass the software setting and just use the driver, I thought of that but stepconf does not let you continue if you enter "0" into the field.
John
John
Please Log in or Create an account to join the conversation.
13 Mar 2015 00:57 #56758
by andypugh
LinuxCNC does not perform _ANY_ driver configuration. Stepconf allows you to set LinuxCNC to suit the setup of the driver.
I would suggest setting the driver to 800 steps per rev (4 x microstepping) and setting LinuxCNC to 4x and things _should_ match.
LinuxCNC only sends out pulses. It can't damage anything if you get the pilses per rev wrong, so whatever works is fine.
Replied by andypugh on topic step/rev problem
I was under the impression that if I use the software to tune the driver then the driver should be set to default.
LinuxCNC does not perform _ANY_ driver configuration. Stepconf allows you to set LinuxCNC to suit the setup of the driver.
I would suggest setting the driver to 800 steps per rev (4 x microstepping) and setting LinuxCNC to 4x and things _should_ match.
LinuxCNC only sends out pulses. It can't damage anything if you get the pilses per rev wrong, so whatever works is fine.
Please Log in or Create an account to join the conversation.
13 Mar 2015 01:13 #56760
by johnl2
Replied by johnl2 on topic step/rev problem
I must come on here and confess. . . I made a mistake. I had the software set to 4 for microstepping, yes, but I did not realize after being tired last night, I had my pitch set to half of what it's supposed to be, 2.5, when it should have been 5. So that is why the stepping was off by exactly that. Now that I put the stepping at 2, pitch at 5 and stepping at 2 also on the driver, it's working fine. Makes total sense. What does not make sense to me is, why must I have the microstepping set on the driver when there's a default setting for the purpose of using software to program it? I don't get that part.. B/c when I switch it to default, it does not work at all, it's off by like 50%.
Thanks,
John
Thanks,
John
Please Log in or Create an account to join the conversation.
Time to create page: 0.189 seconds