When I process a larger 3D decoration topic, the accuracy in both the x and y direction is not good.
I have made a simple test program with G codes that run fast in one direction and slowly back.
After 25 times back and forth, it interferes between 10 and 20 mm incorrectly.
The pulses for stepper drives look very nice, yet Y is 4.5us and X 9 us ??
DC voltage looks stable.
Are there any others who have that problem?
Can it be solved?
BR
g0 x0 y0 f10000
g0 x500 y20 f10000
g0 x0 y20 f1000
g0 x500 y40 f10000
g0 x0 y40 f1000
g0 x500 y60 f10000
g0 x0 y60 f1000
g0 x500 y80 f10000
g0 x0 y80 f1000
g0 x500 y100 f10000
g0 x0 y100 f1000
g0 x500 y120 f10000
g0 x0 y120 f1000
....
....
g0 x0 y480 f1000
g0 x500 y500 f10000
g0 x0 y500 f1000
g0 x0 y0 f1000
# should be the starting point
Missing XY precision (and maybe Z)
-
RoMaker
- Administrateur du site
- Messages : 5063
- Enregistré le : 02 janv. 2017 20:24
- Localisation : Montpellier
- Contact :
Re: Missing XY precision (and maybe Z)
Hi John,
First time I hear that.
Maybe a problem of interference or overheating of the drivers?
Are the TB6600 hot?
Control cables too close to the power cables?
No hard mechanical points?
A "Hi" and "please" would be appreciated
Romain
First time I hear that.
Maybe a problem of interference or overheating of the drivers?
Are the TB6600 hot?
Control cables too close to the power cables?
No hard mechanical points?
A "Hi" and "please" would be appreciated
Romain
-
johnmonrad
- Messages : 6
- Enregistré le : 29 oct. 2018 21:32
Re: Missing XY precision (and maybe Z)
Hi Romain
We are 2 people in Aarhus / Denmark building this great project and we both have the problem. : - (
I think it's a problem between the TB6600 driver and stepper our motor.
The driver is not hot. The stepper is not hot.
In my RS-CNC, signal and power cables are close to each other, but with oscilloscope, the pulses to the drivers look very nice, without noise of any kind.
It helped a little bit, to mount a 4700uF electrolyte on each driver's power supply.
There is extra ground in each cable, for all three drivers.
At first I had a problem with the Y shaft, where the wheel could move a little. It was solved by making a flat surface, where the pinol screw could be tightened.
BR
We are 2 people in Aarhus / Denmark building this great project and we both have the problem. : - (
I think it's a problem between the TB6600 driver and stepper our motor.
The driver is not hot. The stepper is not hot.
In my RS-CNC, signal and power cables are close to each other, but with oscilloscope, the pulses to the drivers look very nice, without noise of any kind.
It helped a little bit, to mount a 4700uF electrolyte on each driver's power supply.
There is extra ground in each cable, for all three drivers.
At first I had a problem with the Y shaft, where the wheel could move a little. It was solved by making a flat surface, where the pinol screw could be tightened.
BR
Re: Missing XY precision (and maybe Z)
Can you increase the width of the pulses?
The TB6600 modules that I use require a lot more (I think about 40 usec).
I think the optocoupler chips are quite slow (at least in my case).
The TB6600 modules that I use require a lot more (I think about 40 usec).
I think the optocoupler chips are quite slow (at least in my case).
-
johnmonrad
- Messages : 6
- Enregistré le : 29 oct. 2018 21:32
Re: Missing XY precision (and maybe Z)
Hi
I now have increased steppulse to 53,6 uS (#define STEPPER_HIGH_DELAY 50), and it did 'nt help. It even get a little bit worse
BR
I now have increased steppulse to 53,6 uS (#define STEPPER_HIGH_DELAY 50), and it did 'nt help. It even get a little bit worse
BR
Re: Missing XY precision (and maybe Z)
Can you check that the wheel (pulley) does not rotate on the axes.
You can put some marks on each part and check after running the test that the marks did not move.
It could also that the MCU is unable to drive the steppers fast enough. Try working in 1/4 of steps (if you had more).
You can put some marks on each part and check after running the test that the marks did not move.
It could also that the MCU is unable to drive the steppers fast enough. Try working in 1/4 of steps (if you had more).
-
johnmonrad
- Messages : 6
- Enregistré le : 29 oct. 2018 21:32
Re: Missing XY precision (and maybe Z)
Hi
You are right.
In the beginning the big pulley on Y was moving. It is fixed now, and a marker on all pulley's show, that they are ok.
I am running with 1/4 steps, but maybe I should try 1/1.
It will give less work for the CPU and driver.
BR
You are right.
In the beginning the big pulley on Y was moving. It is fixed now, and a marker on all pulley's show, that they are ok.
I am running with 1/4 steps, but maybe I should try 1/1.
It will give less work for the CPU and driver.
BR
Re: Missing XY precision (and maybe Z)
If your microsteps is 1/4 (or less) I expect that your mcu can follow.
So, if pulse width is large enough and without noise and if your pulley are not rotating on the axis, I expect that it is the motor that can't follow the drivers.
Torque provided by stepper motor decreases when rotation speed increases.
If you have some "hard point" on your machine, it coud be that the motor looses steps when running fast.
Check if your axes moves "freely" when the motor belt is not connected.
If this is not the reason, perhaps you have to change your set up in order to reduce max speed and/or acceleration.
So, if pulse width is large enough and without noise and if your pulley are not rotating on the axis, I expect that it is the motor that can't follow the drivers.
Torque provided by stepper motor decreases when rotation speed increases.
If you have some "hard point" on your machine, it coud be that the motor looses steps when running fast.
Check if your axes moves "freely" when the motor belt is not connected.
If this is not the reason, perhaps you have to change your set up in order to reduce max speed and/or acceleration.
-
johnmonrad
- Messages : 6
- Enregistré le : 29 oct. 2018 21:32
Re: Missing XY precision (and maybe Z)
Hi
Without belt it feel good in both directions.
Even when I drive slowly forth and back, the position is failing.
I'm pretty sure it's the combination of my drivers and stepper motor.
The sound, when running, is not allwayes nice.
In between there is a small "click". Not from the rails, but from the stepper motors.
I have ordered an EMA2-050D56 driver to see if it changes anything.
Next try must be the stepper motor (s).
BR
Without belt it feel good in both directions.
Even when I drive slowly forth and back, the position is failing.
I'm pretty sure it's the combination of my drivers and stepper motor.
The sound, when running, is not allwayes nice.
In between there is a small "click". Not from the rails, but from the stepper motors.
I have ordered an EMA2-050D56 driver to see if it changes anything.
Next try must be the stepper motor (s).
BR
-
johnmonrad
- Messages : 6
- Enregistré le : 29 oct. 2018 21:32
Re: Missing XY precision (and maybe Z)
Jubii !!
My RS-CNC now works perfectly in the Y direction: -)
Today I have changed a TB6600 driver with an EMA2-050D56.
It works perfectly, sounds better and is absolutely precise.
So my conclusion is, that I've got some TB6600 drivers, from a bad batch.
Thank you for your input to my problem, so now two more drivers need to be purchased.
BR
My RS-CNC now works perfectly in the Y direction: -)
Today I have changed a TB6600 driver with an EMA2-050D56.
It works perfectly, sounds better and is absolutely precise.
So my conclusion is, that I've got some TB6600 drivers, from a bad batch.
Thank you for your input to my problem, so now two more drivers need to be purchased.
BR