Is it extremely important for Reprap controllers to have higher step rates than 10kHz?

I am no expert in motion control or Reprap so not totally sure about this and no one seems to want to explain.

I was saying that a controller I designed was limited to 10kHz as that covered 90% of the possible use cases for stepper motors, but one of the site admin seemed to infer that 100kHz was needed ( [forums.reprap.org] ). Unfortunately I have not received any reply to my queries as to what Reprap requirements needed 100kHz step rates.

I thought maybe I misunderstood what was needed for a Reprap printer so I should ask the rest of the forum users if there was any help available showing what speed a motor needed to turn and what resolution and force would be required by motors.

Anyone have any knowledge about such things?

Quote
mung
I am no expert in motion control or Reprap so not totally sure about this and no one seems to want to explain.

I was saying that a controller I designed was limited to 10kHz as that covered 90% of the possible use cases for stepper motors, but one of the site admin seemed to infer that 100kHz was needed ( [forums.reprap.org] ). Unfortunately I have not received any reply to my queries as to what Reprap requirements needed 100kHz step rates.

I thought maybe I misunderstood what was needed for a Reprap printer so I should ask the rest of the forum users if there was any help available showing what speed a motor needed to turn and what resolution and force would be required by motors.

Anyone have any knowledge about such things?

i pointed out that we already have faster firmware,

assuming the use of gt2 pulleys and 16x microstepping the maximum x y axis speeds you'll get are:

16toothgt2 : 100steps/mm : 100mm/sec max
20toothgt2 : 80steps/mm : 125mm/sec max

these are reasonable x y axis speeds however looking at the extruder axis, if you are running a direct drive extruder you're probably fine as most direct drive extruders will be 90 to 110 steps/mm which means retraction is limited to between 111mm/sec and 90mm/sec however if you are running a geared extruder like a wades extruder or a bulldog extruder where the steps/mm are up in the high 300 to 500 then your dealing with a maximum retraction speed 33mm/sec to 20mm/sec which is a bit of a serious problem, unless you lower the microstepping however the trade off will be extrusion resolution, common retraction speeds people run are between 60mm/sec right up to 160mm/sec

and if you run 32x microstepping you can half all the above speed figures.

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Quote
thejollygrimreaper

Quote
mung
I am no expert in motion control or Reprap so not totally sure about this and no one seems to want to explain.

I was saying that a controller I designed was limited to 10kHz as that covered 90% of the possible use cases for stepper motors, but one of the site admin seemed to infer that 100kHz was needed ( [forums.reprap.org] ). Unfortunately I have not received any reply to my queries as to what Reprap requirements needed 100kHz step rates.

I thought maybe I misunderstood what was needed for a Reprap printer so I should ask the rest of the forum users if there was any help available showing what speed a motor needed to turn and what resolution and force would be required by motors.

Anyone have any knowledge about such things?

i pointed out that we already have faster firmware,

assuming the use of gt2 pulleys and 16x microstepping the maximum x y axis speeds you'll get are:

16toothgt2 : 100steps/mm : 100mm/sec max
20toothgt2 : 80steps/mm : 125mm/sec max

these are reasonable x y axis speeds however looking at the extruder axis, if you are running a direct drive extruder you're probably fine as most direct drive extruders will be 90 to 110 steps/mm which means retraction is limited to between 111mm/sec and 90mm/sec however if you are running a geared extruder like a wades extruder or a bulldog extruder where the steps/mm are up in the high 300 to 500 then your dealing with a maximum retraction speed 33mm/sec to 20mm/sec which is a bit of a serious problem, unless you lower the microstepping however the trade off will be extrusion resolution, common retraction speeds people run are between 60mm/sec right up to 160mm/sec

and if you run 32x microstepping you can half all the above speed figures.

So what effect does microstepping have on torque?
Why does the extruder need to be retracted with microstepping?
surely extrusion speed and resolution will be proportional to speed of movement of the extruder?

To try and make the simplest possible description,
step speed required = the speed of movement required/ the resolution of each step.

If you want 100mm/s at a resolution of 0.01mm ....... 100/0.01 = 10,000

10kHz should be sufficient for most uses.

Quote
mung

Quote
thejollygrimreaper

Quote
mung
I am no expert in motion control or Reprap so not totally sure about this and no one seems to want to explain.

I was saying that a controller I designed was limited to 10kHz as that covered 90% of the possible use cases for stepper motors, but one of the site admin seemed to infer that 100kHz was needed ( [forums.reprap.org] ). Unfortunately I have not received any reply to my queries as to what Reprap requirements needed 100kHz step rates.

I thought maybe I misunderstood what was needed for a Reprap printer so I should ask the rest of the forum users if there was any help available showing what speed a motor needed to turn and what resolution and force would be required by motors.

Anyone have any knowledge about such things?

i pointed out that we already have faster firmware,

assuming the use of gt2 pulleys and 16x microstepping the maximum x y axis speeds you'll get are:

16toothgt2 : 100steps/mm : 100mm/sec max
20toothgt2 : 80steps/mm : 125mm/sec max

these are reasonable x y axis speeds however looking at the extruder axis, if you are running a direct drive extruder you're probably fine as most direct drive extruders will be 90 to 110 steps/mm which means retraction is limited to between 111mm/sec and 90mm/sec however if you are running a geared extruder like a wades extruder or a bulldog extruder where the steps/mm are up in the high 300 to 500 then your dealing with a maximum retraction speed 33mm/sec to 20mm/sec which is a bit of a serious problem, unless you lower the microstepping however the trade off will be extrusion resolution, common retraction speeds people run are between 60mm/sec right up to 160mm/sec

and if you run 32x microstepping you can half all the above speed figures.

So what effect does microstepping have on torque?
Why does the extruder need to be retracted with microstepping?
surely extrusion speed and resolution will be proportional to speed of movement of the extruder?

To try and make the simplest possible description,
step speed required = the speed of movement required/ the resolution of each step.

If you want 100mm/s at a resolution of 0.01mm ....... 100/0.01 = 10,000

10kHz should be sufficient for most uses.

how are you going to handle the heaters? eg bed and extruder? last i checked the raspberry pi didn't have any ADC inputs?

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Quote
mung

Quote
thejollygrimreaper

Quote
mung
I am no expert in motion control or Reprap so not totally sure about this and no one seems to want to explain.

I was saying that a controller I designed was limited to 10kHz as that covered 90% of the possible use cases for stepper motors, but one of the site admin seemed to infer that 100kHz was needed ( [forums.reprap.org] ). Unfortunately I have not received any reply to my queries as to what Reprap requirements needed 100kHz step rates.

I thought maybe I misunderstood what was needed for a Reprap printer so I should ask the rest of the forum users if there was any help available showing what speed a motor needed to turn and what resolution and force would be required by motors.

Anyone have any knowledge about such things?

i pointed out that we already have faster firmware,

assuming the use of gt2 pulleys and 16x microstepping the maximum x y axis speeds you'll get are:

16toothgt2 : 100steps/mm : 100mm/sec max
20toothgt2 : 80steps/mm : 125mm/sec max

these are reasonable x y axis speeds however looking at the extruder axis, if you are running a direct drive extruder you're probably fine as most direct drive extruders will be 90 to 110 steps/mm which means retraction is limited to between 111mm/sec and 90mm/sec however if you are running a geared extruder like a wades extruder or a bulldog extruder where the steps/mm are up in the high 300 to 500 then your dealing with a maximum retraction speed 33mm/sec to 20mm/sec which is a bit of a serious problem, unless you lower the microstepping however the trade off will be extrusion resolution, common retraction speeds people run are between 60mm/sec right up to 160mm/sec

and if you run 32x microstepping you can half all the above speed figures.

So what effect does microstepping have on torque?
Why does the extruder need to be retracted with microstepping?
surely extrusion speed and resolution will be proportional to speed of movement of the extruder?

To try and make the simplest possible description,
step speed required = the speed of movement required/ the resolution of each step.

If you want 100mm/s at a resolution of 0.01mm ....... 100/0.01 = 10,000

10kHz should be sufficient for most uses.

you ask what the micro stepping has on torque.....
First you have to know how the stepper motor works.
When we are talking about extruder motors, Micro-stepping doesn't change anything much to do with torque.
to get the required torque from a stepper motor in an extruder, you use a set of gearing, or a stepper motor with internal gearing.
this has all been tried and tested to be proven right.

a stepper motor has 200 steps per revolution. micro-stepping is turning this into a higher number.
Stepper motor 'Inception' the steps between the steps and so-on

Polulo driver mention

Quote
pololu
Six different step resolutions: full-step, half-step, 1/4-step, 1/8-step, 1/16-step, and 1/32-step

and also

For instance, driving a motor in quarter-step mode will give the 200-step-per-revolution motor 800 micro-steps per revolution by using four different current levels.

this is allowing for 'Micro control'
for 1mm of movement with 1/16 micro-stepping with a M5 lead screw you need 4000 steps
with 1/32 you get this number moved to 8000 steps. allowing you to print at a smaller theoretical resolution.
it is giving greater control of all the 'in between' steps.

and the same with the extruder.

I run my extruder from a $6 stepper motor with 1/16 micro stepping and with a greg/wade gear (i think 1:3 ratio)
and this is how i get the torque you are talking about.
I would have no problem moving to 1/32 micro-stepping.
and it would give me greater theoretical control. which would greater influence my theoretical flow rate.
and the firmware is going to tell it to move a certain distance regardless of microsteps.

and without repeating thejollygrimreaper,
i will point you back to what he said ↑↑↑↑

as long as my motors are receiving enough current, I will not ever have a problem.

Mung, you are being very unruly in the last few weeks.
I see you badly referring to admin and other members.

Many of us have given our thoughts and opinions on what you are suggesting,
stop trying to fight it, take what you can and go back to your drawing table.

I think the rapsi set-up you refer to is probably the more expensive route, with no benefits over the already proven and cheap options.
I am talking about ramps, melzi, and others.

you need to catch-up on some reading

read this - [www.freescale.com]
read this - [www.pololu.com]
read this - [www.onsemi.com]
read this - [www.ti.com]
read this - [www.orientalmotor.com]
read this- [machinedesign.com]
and this - [www.micromo.com]


the last one is important to you
'micro-stepping myths and realities'

Quote

as you increase the number of micro-steps per full step the INCREMENTAL torque per micro-step drops off drastically. Resolution increases but accuracy will actually suffer.

but for me (and i think many others), this doesn't matter.
I am using 1/16 with the right amount of current and its perfect for me. and then it'll use 1/32 which is still fine for me
and as we will never move to 1/256 micro-stepping

In electronics, coding , design, life and research we use Logic.
'Logic (from the Ancient Greek: λογική, logike) is the use and study of valid reasoning. '

if we all totally don't understand you.
Give us valid reasoning
Research properly, and convey your points across in a logical way.

Logic, logic, logic...
Without logic, everything is illogical (1x i and 2x L's)
[www.thefreedictionary.com]




∞ ^peace ∞

Edited 1 time(s). Last edit at 10/19/2014 10:58AM by pushthatbolder.

Quote
pushthatbolder

and this - [www.micromo.com]


the last one is important to you
'micro-stepping myths and realities'

Quote

as you increase the number of micro-steps per full step the INCREMENTAL torque per micro-step drops off drastically. Resolution increases but accuracy will actually suffer.

So you are agreeing with me?

There is no increase in accuracy from increasing microsteps, the EM induced torque increases with each incremental step without movement until it overcomes the force of resistance then jumps to the next full step position with some overshoot.

There is no point in using micro stepping if you have well designed motors running at the corect speed to avoid resonance.

The "incremental torque per microstep" anaylisis does not mean that accuracy will suffer at a higher microstep rate - it only means that you won't see as much improvement in accuracy as you might think - ie, it just means that microstepping a motor one 1/32th microstep does not move the motor that amount. Many years ago, I did some testing with an IMS microstepping driver that was progammable to 256x microstepping. We put an accelerometer on the motor and looked at vibration levels with different levels of microstepping. Above 8x microstepping, there was no noticable reduction in vibration, meaning that the motor never really moved in smaller than 1/8th step increments. I spoke with the head of Lin Engineering at the time (makers of high quality industrial steppers), and he described the great pains they went to in creating rotor tooth profiles to increase the microstepping accuracy. He also pointed out that doing this tended to reduce the peak holding torque. So you can bet that your $9 stepper with a 60 oz-in holding torque has not been optimized for microstepping accuracy.

Also keep in mind that independent of the quality of the motor and the microstepping ratio, you can't guarantee that the actual motor position will match the microstep position to anything less than +/1 one full step.

Way to misquote me and not take everything into consideration.
They have found 1/16 to be a balance of speed/accuracy/resolution/resonance.
And 1/32 has also reduced resonance. And these drivers have become readily available and proven to be good.

We are after resolution.
I want resolution. 50 microns is what i get With 1/16 which probably isnt optimal and should be about 49 microns or whatever im just rounding to 50 micron for the sake of it.

1/16 is readily available and cheap.
1/32 is about the same.

When talking about accuracy, its talking about having 200steps × 32 per revolition and trying to move it to a point of inaccuracy.
It going to want to move.to the next full step.
Thats why theoretical layer heights and far from actual optimal printing heights.

Look at the prusa calculator.

I use the optimal layer height calculator.

And guess what? I get no inaccuracy error. Or atleast non i can see or measure with my tools

Mung, do you have a printer? .
We have chosen to use printers that even after your "loss of torque" due to microstepping.
Still have enough power and torque for our needs.
My stepper motor was $6 ,for a 42mm long nema 17
Even after 1/16 microstepping. I cannot stop the rotor with my hand.
And the same with 1/32 microstepping.

Inaccuracy is cured with optimal printing layer height. And optimised steps taken.
Its a damn $6 stepper motor, what did yoi expect.?

I gave you those link for you too read, did you? So fast?
Did you even read.my.post? Or just gave it a skim?

For the record i dont agree with mung. You need to research and approach from all angles. Logically.
I agree with thejollygrimreaper and Lobocnc.

For the love of everything good.
Respect the opinion of everyone, dont be aggresive.
Thejollygrimreaper and lobocnc know their shit.
And other people who have gave their input also know their shit also.
The mass of people cant be wrong.
So go and learn, never assume you are finished learning.
So go and learn, never assume you are finished learning.
So go and learn.
Rinse and repeat

Edited 2 time(s). Last edit at 10/19/2014 02:30PM by pushthatbolder.

Regarding step rates and achievable feedrates: [reprap.org]

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Regarding stepper motor microstepping above 10x (see 'Stepper FAQ' then 'Why 10 microsteps'): [www.geckodrive.com]
TL/DR - it depends on the tolerance of your motor.

Ian
RepRapPro tech support