clever test for skipped steps?
13 May 2015 05:57 #58614
by pgf
clever test for skipped steps? was created by pgf
in my experience, skipped steps can sort of sneak up on you, so that you might only notice that you have a problem when you notice that your circles aren't round or your squares aren't square.
since steps can start skipping due to mechanical issues that might not be immediately apparent, it would be convenient to have some semiautomatic way of knowing that they're being skipped.
is there a trick, perhaps a special pattern to carve, that can make it obvious that steps are being lost? or is comparing the actual vs. predicted lengths of a cut the only way? the trouble with that method is that if it's a very few steps that are being lost, it might be hard to measure their affect.
paul
since steps can start skipping due to mechanical issues that might not be immediately apparent, it would be convenient to have some semiautomatic way of knowing that they're being skipped.
is there a trick, perhaps a special pattern to carve, that can make it obvious that steps are being lost? or is comparing the actual vs. predicted lengths of a cut the only way? the trouble with that method is that if it's a very few steps that are being lost, it might be hard to measure their affect.
paul
Please Log in or Create an account to join the conversation.
13 May 2015 18:21 #58627
by BigJohnT
Replied by BigJohnT on topic clever test for skipped steps?
Usually a properly configured stepper machine will not loose a step. What usually happens is a stepper stall if you get too much load on the stepper. You should do a stepper test as described in the manual to find the upper limits of your machine. If your machine design allows you to have match marks on the axes you could use them to see if your still in sync.
JT
JT
Please Log in or Create an account to join the conversation.
13 May 2015 19:24 #58632
by pgf
Replied by pgf on topic clever test for skipped steps?
thanks -- searching for the tests also found me the "Stepper Motor Speed Limitations" and some other pertinent pages. i probably saw those years ago, and glazed over, or else they've been written more recently. in any case, i probably have a misconfiguration of, or misguided expectations for, my steppers.
by "match marks", you simply mean setting a mark before running a test, and then seeing if it still matches after? i think i should be able to do that -- good idea.
paul
by "match marks", you simply mean setting a mark before running a test, and then seeing if it still matches after? i think i should be able to do that -- good idea.
paul
Please Log in or Create an account to join the conversation.
13 May 2015 20:13 #58637
by ArcEye
By far the most common cause of lost steps is completely unrealistic velocity and accelerations figures.
The system often does not get stretched, because a move is never long enough to reach the max, but occassional lost steps can often be traced back to this.
Something like a sudden retraction on a lathe and return to the start point at full speed will often show this up.
Another common cause is an underpowered PSU, which will cope with 2 axes moving at one time, but when 3 want to move in a co-ordinated fashion, it simply does not have
the juice to power them all and steps get lost.
regards
Replied by ArcEye on topic clever test for skipped steps?
i probably have a misconfiguration of, or misguided expectations for, my steppers.
By far the most common cause of lost steps is completely unrealistic velocity and accelerations figures.
The system often does not get stretched, because a move is never long enough to reach the max, but occassional lost steps can often be traced back to this.
Something like a sudden retraction on a lathe and return to the start point at full speed will often show this up.
Another common cause is an underpowered PSU, which will cope with 2 axes moving at one time, but when 3 want to move in a co-ordinated fashion, it simply does not have
the juice to power them all and steps get lost.
regards
Please Log in or Create an account to join the conversation.
14 May 2015 06:23 #58657
by BigJohnT
Yes, if you have two surfaces say for the X axis that are close to each other you can move to almost the end of travel then scribe a match mark on the two surfaces. If they are close together you can get very accurate positioning.
JT
Replied by BigJohnT on topic clever test for skipped steps?
by "match marks", you simply mean setting a mark before running a test, and then seeing if it still matches after? i think i should be able to do that -- good idea.
paul
Yes, if you have two surfaces say for the X axis that are close to each other you can move to almost the end of travel then scribe a match mark on the two surfaces. If they are close together you can get very accurate positioning.
JT
Please Log in or Create an account to join the conversation.
14 May 2015 19:13 - 14 May 2015 19:13 #58674
by andypugh
You can judge the alignment of two scribed marks to 0.001" or 0.02mm reasonably easily. If you couldn't then vernier calipers wouldn't work. (All the vernier does is give you alternative marks to align for each sub-division, you still judge the alignment)
Replied by andypugh on topic clever test for skipped steps?
If they are close together you can get very accurate positioning.
You can judge the alignment of two scribed marks to 0.001" or 0.02mm reasonably easily. If you couldn't then vernier calipers wouldn't work. (All the vernier does is give you alternative marks to align for each sub-division, you still judge the alignment)
Last edit: 14 May 2015 19:13 by andypugh.
Please Log in or Create an account to join the conversation.
14 May 2015 19:51 #58682
by pgf
Replied by pgf on topic clever test for skipped steps?
thanks -- i think i can do as you suggest -- my X and Y axes are set up for fairly easy scribing.
so far i've only seen the dropped steps (and a lot of them) while cutting. the scribe trick will let me check the rapids.
my situation is this: as a first step in increasing travel speeds, i replaced my 5/16" threaded rod and couplers with 3/8 acme stuff. i'm now able to rapid at 3X what i was doing before (54 vs 18 ipm). but cutting at anything more than the original 18ipm is unreliable (and i need to go back and verify the 18ipm cuts).
my motors are admittedly undersized, about 65 in-oz. upgrading them is clearly the next step, but that will also require a new driver board, and i want to be sure nothing else is wrong before throw^H^H^H^H^Hspending that money.
paul
so far i've only seen the dropped steps (and a lot of them) while cutting. the scribe trick will let me check the rapids.
my situation is this: as a first step in increasing travel speeds, i replaced my 5/16" threaded rod and couplers with 3/8 acme stuff. i'm now able to rapid at 3X what i was doing before (54 vs 18 ipm). but cutting at anything more than the original 18ipm is unreliable (and i need to go back and verify the 18ipm cuts).
my motors are admittedly undersized, about 65 in-oz. upgrading them is clearly the next step, but that will also require a new driver board, and i want to be sure nothing else is wrong before throw^H^H^H^H^Hspending that money.
paul
Please Log in or Create an account to join the conversation.
14 May 2015 20:30 #58687
by BigJohnT
Voltage is the most important thing to get speed on a stepper machine. Also typically I've seen acceleration to be 10-20 times what max velocity is. To add if you have fallen into the trap of using backlash you must also have headroom in the step generator (see the manual).
JT
Replied by BigJohnT on topic clever test for skipped steps?
thanks -- i think i can do as you suggest -- my X and Y axes are set up for fairly easy scribing.
so far i've only seen the dropped steps (and a lot of them) while cutting. the scribe trick will let me check the rapids.
my situation is this: as a first step in increasing travel speeds, i replaced my 5/16" threaded rod and couplers with 3/8 acme stuff. i'm now able to rapid at 3X what i was doing before (54 vs 18 ipm). but cutting at anything more than the original 18ipm is unreliable (and i need to go back and verify the 18ipm cuts).
my motors are admittedly undersized, about 65 in-oz. upgrading them is clearly the next step, but that will also require a new driver board, and i want to be sure nothing else is wrong before throw^H^H^H^H^Hspending that money.
paul
Voltage is the most important thing to get speed on a stepper machine. Also typically I've seen acceleration to be 10-20 times what max velocity is. To add if you have fallen into the trap of using backlash you must also have headroom in the step generator (see the manual).
JT
Please Log in or Create an account to join the conversation.
Time to create page: 0.450 seconds