PDA

View Full Version : Multi stepping



ptjw7uk
22-07-2009, 02:39 PM
Hi All,
I am a little puzzled by some information on the web as to multi stepping and it has got me wondering as to the worth of it.
Are the muliti steps always the same ie is an 8th step equal to 0.225 degrees everytime (1/8 of 1.8 degrees) and is it reproducable.

Just wondering as I can see how a half step works but not multi(micro)step works.

peter

irving2008
22-07-2009, 02:53 PM
Hi All,
I am a little puzzled by some information on the web as to multi stepping and it has got me wondering as to the worth of it.
Are the muliti steps always the same ie is an 8th step equal to 0.225 degrees everytime (1/8 of 1.8 degrees) and is it reproducable.

Just wondering as I can see how a half step works but not multi(micro)step works.

peterYes. the reproducability is down to the motor to some extent. But microstepping is done in exactly the same way as 1/2 step... applying differential current to the opposing phase coils so that one coil is pulling the rotor forward and the other pulling it back. The resultant 'vector' is part way between two step locations. The current applied to each coil is a 'stepped sine wave'... here (http://www.stepperworld.com/Tutorials/pgMicrostepping.htm)is a good tutorial on the subject.

HTH

ptjw7uk
22-07-2009, 03:26 PM
Thanks Irving2008 goes some of the way but still not clear how reproducible partial steps are when under load as the bit I read on the web inferred that steps between full and half were hit and miss not that I fully understand but I would think that as one of the coils only has partial power then its actual position will be influenced by the load.
So I would think that greater accuracy could be achieved with gearing albeit with a slower responding unit being as I want to make fairly small parts I think I should have gone with gearing instead of direct drive, only time will tell.

Peter

irving2008
22-07-2009, 03:37 PM
Thanks Irving2008 goes some of the way but still not clear how reproducible partial steps are when under load as the bit I read on the web inferred that steps between full and half were hit and miss not that I fully understand but I would think that as one of the coils only has partial power then its actual position will be influenced by the load.
So I would think that greater accuracy could be achieved with gearing albeit with a slower responding unit being as I want to make fairly small parts I think I should have gone with gearing instead of direct drive, only time will tell.

PeterAs long as you are not overloading the motor the accuracy should be reproducible - how good that will be will depend on the tolerance you can sustain. The tolerance of the motor is about 2 - 5% typically.

As you say, a lot depends on the design and work area of the overall system. With a 200step/rev motor on a 5mm leadscrew, say, 1 full step = 0.025mm. If you wanted 0.01mm accuracy you could either use 1/8 stepping to give .003125mm/step or geardown 2.5:1. The 1/8 step would give positioning of 0, .009375, .021875, .03125, .040625, etc, a maximum error of 0.001875mm, which may be good enough. 1/16 step would halve that error.

The disadvantage as you point out is that gearing down would reduce the ability to do rapids. In the above example at 1000 (full)steps/sec the rapids for the direct drive would be 25mm/sec (1.5m/min), whereas with the geared drive it would be only 10mm/sec (0.6m/min) which makes a big difference on a 1000 x 600 work area. On a 200mm x 100mm PCB router tho its not such an issue.

ptjw7uk
22-07-2009, 03:48 PM
Only problem I now have is that as I want to go small the cutters need to get smaller and so up goes the motor speed.
I think I need to investigate the brushless motors used in models as they seem to give the speed in a small envelope and at a reasonable price although not sure about the high current needed also how long will they run for as I have not seen many models running for hours!

Peter

ptjw7uk
22-07-2009, 09:38 PM
Found an interesting site for spindles
http://www.cncathome.com/spindles.html
Worth a read although he doesnt rate model brushless motors.
Peter

JAFMotion
22-07-2009, 09:44 PM
Hi,

My take on microstepping is that the biggest gain is smoothness of motion and reduced vibration. The increase in resolution is not always the main reason for choosing microstepping. We used to use 1/2 step mode on our machines, now we use 1/8th step. The only change we have made to the controls are to increase the step rate by 4. We now have the same usable resolution but with a much smoother system that has improved acceleration and reliability.

Hope this helps.

Simon

Robin Hewitt
23-07-2009, 12:50 AM
My take on microstepping is that it's fine and dandy so long as there is no load on the motor.

As you go beyond half step it gets springy. Any stiction means it won't start to move until you're far enough ahead to overcome it, then, suddenly, BOING.

I half step to 10um then add an intermediate 5um microstep but it's only there to smooth things out, it ain't gonna happen :heehee:

Gary
23-07-2009, 04:19 PM
Microstepping is not linear (Between full step positions) and should not be relied upon.
the error you will see is only between the full step positions.
I demonstrated this to a customer with a stepper motor set to a high microstepping resolution that had a mirror attached to the shaft and I pointed a laser at the mirror and put a pen mark on the wall on the other side of the room, I then moved the motor 100 microsteps, marked the wall and then moved back 100 microsteps, and the laser was a few cm from the original mark.
Gearing will be better for accuracy, but you have the downside of smoothness, noise, backlash or hysteresis.
If you need accuracy and repeatability with very small increments then closed loop servo would be the way to go (PID driver)