Using RC servos instead of stepper motors
Posted by fdavies
|
Using RC servos instead of stepper motors March 06, 2010 08:48AM |
Registered: 15 years ago Posts: 59 |
I have made a prototype of a linear axis that uses an RC servo. [objects.reprap.org]
I think that servos may have the following advantages:
1. Very light.
2. Servos include all necessary drive electronics. This means that you can drive them directly from a variety of commercially available boards. I have used an Arduino.
3. Servos are available in a range of prices, strengths, and robustnesses.
I am going to see if I can set up a life test with my prototype. I will update the reprap community with my progress.
fdavies
I think that servos may have the following advantages:
1. Very light.
2. Servos include all necessary drive electronics. This means that you can drive them directly from a variety of commercially available boards. I have used an Arduino.
3. Servos are available in a range of prices, strengths, and robustnesses.
I am going to see if I can set up a life test with my prototype. I will update the reprap community with my progress.
fdavies
|
Re: Using RC servos instead of stepper motors March 06, 2010 03:19PM |
Registered: 14 years ago Posts: 458 |
RC servos are not very good at absolute positioning, that and the fact that you would need to constantly give them a pulse train of 50/50 PWm for them to stay still is also another thing to consider.
There are however proper closed circuit servos used in robotics, these may be of use, but RC servos are definitely not useful.
There are however proper closed circuit servos used in robotics, these may be of use, but RC servos are definitely not useful.
|
Re: Using RC servos instead of stepper motors March 07, 2010 03:38AM |
Registered: 16 years ago Posts: 1,094 |
also, RC servos have a significant 'dead zone' to prevent hunting, which will totally destroy your precision with small movements. positional accuracy isn't amazing either.
Servos can be used in systems like this, but the feedback systems must be set up properly. An AS5040 or similar would be ideal.
-----------------------------------------------
Wooden Mendel
Teacup Firmware
Servos can be used in systems like this, but the feedback systems must be set up properly. An AS5040 or similar would be ideal.
-----------------------------------------------
Wooden Mendel
Teacup Firmware
|
Re: Using RC servos instead of stepper motors April 01, 2010 05:00PM |
I agree on using anything but stepper motors as they create bumps that I do not like because of their very nature. So then what. I found some nice inexpensive gear driven motors from Dayton. I then reformatted the encoder board and fit a magnet perfectly on the output shaft so that I can accurately read position of each 1/360° rotation and maybe even a little better due to the complex math.
I also had to redo the extruder board. I did not like some of the connectors and I do not like the chips used to drive steppers but rather drive a standard motor. With that I had a lot of trouble with my last board using the toner transfer method (thanks to no boards for the extruder ever in stock). I have since purchased all the equipment to make professional level boards and what a difference it has made. The encoder board (pre new equipment)had a very small imperfection from the toner method in the trace that connects the reset to the outside world. The result was the momentary switch reset the mega168, but nothing else would causing huge difficulties burning the board. Upon looking under a microscope I noticed the broken trace. Built up the whole trace burned the bootloader and low and behold a heartbeat.
The reason I think most do not use the encoder AS5040 is because of its .65mm pitch device that has to not only be precisely etched with all the traces intact (which is a big beef with me and the toner transfer method) it also is that the device is soldered very precisely and over the 6mm hole you placed in the board. It is true the hall effect device does not care which side of the chip the magnet rides on I just like my components on top where I can see them. Personal preference.
At any rate I have now successfully crafted an encoder board and successfully mounted it onto a Dayton permanent magnet gear motor with a ratio of 270:1 which gives me about 25 RPM at full speed. It is a 12 volt device and runs 1.3 amps at full loading. That is the rating of these extruder board output chips A3949 from Allegro and I only need one.
So where am I at with this. I intend to use this motor with all the motors as the smoothness is seen as all that is controlled is speed and direction. To rid the error seen by backlash of the gears, which is there, I intend to only run in one direction with a backlash error zeroing every time it homes back to zero. That is how a milling machine works and they receive a .0005" accuracy. That way the place it is at is precisely known and I have no issues of the jittering edges.
I do not have the source code modified yet to handle this. Anyone interested in collaborating?
This will also solve the jitters seen in laser cutting circles. I made the gears on a laser cutter out of acrylic. The edges would never pass inspection and just barely work.
If someone can tell me how to post pictures I am happy to. Maybe in another forum?
I also had to redo the extruder board. I did not like some of the connectors and I do not like the chips used to drive steppers but rather drive a standard motor. With that I had a lot of trouble with my last board using the toner transfer method (thanks to no boards for the extruder ever in stock). I have since purchased all the equipment to make professional level boards and what a difference it has made. The encoder board (pre new equipment)had a very small imperfection from the toner method in the trace that connects the reset to the outside world. The result was the momentary switch reset the mega168, but nothing else would causing huge difficulties burning the board. Upon looking under a microscope I noticed the broken trace. Built up the whole trace burned the bootloader and low and behold a heartbeat.
The reason I think most do not use the encoder AS5040 is because of its .65mm pitch device that has to not only be precisely etched with all the traces intact (which is a big beef with me and the toner transfer method) it also is that the device is soldered very precisely and over the 6mm hole you placed in the board. It is true the hall effect device does not care which side of the chip the magnet rides on I just like my components on top where I can see them. Personal preference.
At any rate I have now successfully crafted an encoder board and successfully mounted it onto a Dayton permanent magnet gear motor with a ratio of 270:1 which gives me about 25 RPM at full speed. It is a 12 volt device and runs 1.3 amps at full loading. That is the rating of these extruder board output chips A3949 from Allegro and I only need one.
So where am I at with this. I intend to use this motor with all the motors as the smoothness is seen as all that is controlled is speed and direction. To rid the error seen by backlash of the gears, which is there, I intend to only run in one direction with a backlash error zeroing every time it homes back to zero. That is how a milling machine works and they receive a .0005" accuracy. That way the place it is at is precisely known and I have no issues of the jittering edges.
I do not have the source code modified yet to handle this. Anyone interested in collaborating?
This will also solve the jitters seen in laser cutting circles. I made the gears on a laser cutter out of acrylic. The edges would never pass inspection and just barely work.
If someone can tell me how to post pictures I am happy to. Maybe in another forum?
Sorry, only registered users may post in this forum.