Looking for alpha-testers for closed loop stepper controller
Posted by LoboCNC
|
Looking for alpha-testers for closed loop stepper controller October 22, 2013 06:53PM |
I'm working on a new closed-loop stepper controller and am looking for a few alpha testers. The closed loop control uses feedback from an encoder mounted on the stepper motor to essentially turn it into a brushless DC servo motor. This is the same type of control done on my Lobo CNC milling machine controller board (see http://lobocnc.com/motorcontrol.html for a more complete description of how this is done). The net effect is lower power usage, less heat, less noise & vibration, and you can run the motor right up to the torque limit without fear of losing steps. (In many applications, you can even overdrive the motors safely.) In particular, I think this would be a good match for extruders where you want to both minimize the motor mass and minimize the heat generated by the motor.
This new board mounts on the back of the motor and is designed to be used with the new low-cost CUI AMT103 encoders. It can be used in one of two different modes:
In Step & Direction mode, you just hook up your Step & Direction signals, along with motor power, and it will automatically power-up and follow your step signals. In this mode, there is also and Enable input to turn on and off the driver, and a Fault output to let you know if the motor is no longer able to track the step signals.
It also has an RS485/RS422 serial interface that can be connected to RS232 (+/-12v or TTL level) or to a real RS422 port. In this mode, multiple controllers can hang off of a single serial port, and you get much more control over, and feedback from, the motor/controller. The serial interface also supports multi-axis coordination so that you can coordinate motions directly from a PC without any additional electronics. (This is most easily done using a USB to RS422 converter cable.)
The board uses an Allegro A4984 driver chip and will be rated for 2A, 12-24v. It'll be suitable for NEMA 17's or small-to-medium sized NEMA23's. I'll provide the board and aluminum housing - testers will have to provide their own double-shaft motor and encoder. I'm really looking for testers with more motor control experience (although not necessarily with servo motors) who are comfortable with wiring (and know about things like shielding and twisted pairs). Please PM me if you are interested in getting a board to play with. Reply to this post, though, if you want any more general details.
Thanks, Jeff
(I should have the boards and basic data sheet together in a couple of weeks.)
This new board mounts on the back of the motor and is designed to be used with the new low-cost CUI AMT103 encoders. It can be used in one of two different modes:
In Step & Direction mode, you just hook up your Step & Direction signals, along with motor power, and it will automatically power-up and follow your step signals. In this mode, there is also and Enable input to turn on and off the driver, and a Fault output to let you know if the motor is no longer able to track the step signals.
It also has an RS485/RS422 serial interface that can be connected to RS232 (+/-12v or TTL level) or to a real RS422 port. In this mode, multiple controllers can hang off of a single serial port, and you get much more control over, and feedback from, the motor/controller. The serial interface also supports multi-axis coordination so that you can coordinate motions directly from a PC without any additional electronics. (This is most easily done using a USB to RS422 converter cable.)
The board uses an Allegro A4984 driver chip and will be rated for 2A, 12-24v. It'll be suitable for NEMA 17's or small-to-medium sized NEMA23's. I'll provide the board and aluminum housing - testers will have to provide their own double-shaft motor and encoder. I'm really looking for testers with more motor control experience (although not necessarily with servo motors) who are comfortable with wiring (and know about things like shielding and twisted pairs). Please PM me if you are interested in getting a board to play with. Reply to this post, though, if you want any more general details.
Thanks, Jeff
(I should have the boards and basic data sheet together in a couple of weeks.)
|
Re: Looking for alpha-testers for closed loop stepper controller October 23, 2013 07:33PM |
Using an encoder on motor not much help.
What should be used is a linear measurement scale on each axis
to give feedback to exactly where the print head is located.
Precise DRO linear scales are not that expensive.
and have a standard known interface.
What should be used is a linear measurement scale on each axis
to give feedback to exactly where the print head is located.
Precise DRO linear scales are not that expensive.
and have a standard known interface.
|
Re: Looking for alpha-testers for closed loop stepper controller October 23, 2013 07:49PM |
cozmicray Wrote:
-------------------------------------------------------
> Using an encoder on motor not much help.
>
> What should be used is a linear measurement
> scale on each axis
> to give feedback to exactly where the print head
> is located.
>
> Precise DRO linear scales are not that expensive.
> and have a standard known interface.
>
>
Closed loop control using an encoder on the motor is mostly just to improve the motor performance. Correcting for slop in the mechanism using direct sensing on the opposite side of the drive train is an entirely different kettle of fish. It seems like a good idea, but it's what's known as a "non-co-located actuator and sensor" system, which tends to create nightmares with the feedback control. If there are any significant dynamics between the actuator and sensor (ie, backlash, flexibility, intermediate masses flopping around...) you end up needing a pretty good dynamic model of the system and its parameters, which is harder than it sounds. That's why decent milling machines all use zero-play ball screws - it doesn't end up being practical trying to close the loop around DRO feedback.
-------------------------------------------------------
> Using an encoder on motor not much help.
>
> What should be used is a linear measurement
> scale on each axis
> to give feedback to exactly where the print head
> is located.
>
> Precise DRO linear scales are not that expensive.
> and have a standard known interface.
>
>
Closed loop control using an encoder on the motor is mostly just to improve the motor performance. Correcting for slop in the mechanism using direct sensing on the opposite side of the drive train is an entirely different kettle of fish. It seems like a good idea, but it's what's known as a "non-co-located actuator and sensor" system, which tends to create nightmares with the feedback control. If there are any significant dynamics between the actuator and sensor (ie, backlash, flexibility, intermediate masses flopping around...) you end up needing a pretty good dynamic model of the system and its parameters, which is harder than it sounds. That's why decent milling machines all use zero-play ball screws - it doesn't end up being practical trying to close the loop around DRO feedback.
Sorry, only registered users may post in this forum.