Are servos that much more accurate than steppers with ballscrews ( not acme )?
Servo vs Stepper
- Thread starter JoeyInCali
- Start date
Here are some links Joey. It seems it is just as much a can of worms topic as any can be. With advantage and disadvantage about equal to either system. It depends on the final use and controller.
I hope this helps you in your decision.
Neil
http://medicaldesign.com/motors-motion-control/servo_vs_stepper/
http://www.control.com/1026215811/index_html
http://www.torchmate.com/motors/electronics_selection.htm Shows a Graph
http://www.imsrv.com/deskcnc/steppers_vs_servos.htm For Taig etc
I hope this helps you in your decision.
Neil
http://medicaldesign.com/motors-motion-control/servo_vs_stepper/
http://www.control.com/1026215811/index_html
http://www.torchmate.com/motors/electronics_selection.htm Shows a Graph
http://www.imsrv.com/deskcnc/steppers_vs_servos.htm For Taig etc
RBC
Deceased
Joey,
Take a look at what Marris of Gecko Drives has to say here.
http://www.geckodrive.com/support.aspx?n=442621
Marris is considered to be one of the best when it comes to both stepper and servo systems. Currently all our machines run on Xylotex boards, but I will use the new smaller Gecko boards for future machines.
Royce Bunnell
www.obcues.com
Take a look at what Marris of Gecko Drives has to say here.
http://www.geckodrive.com/support.aspx?n=442621
Marris is considered to be one of the best when it comes to both stepper and servo systems. Currently all our machines run on Xylotex boards, but I will use the new smaller Gecko boards for future machines.
Royce Bunnell
www.obcues.com
One of the main differences is feedback.
In basic terms:
A stepper controller tells the stepper to go to X new position. Whether it actually gets to X new position it has no idea. So if you don't size the stepper properly for the mass you are moving or cut you are taking, it will lose track of where it is.
A servo system by design always knows where it is. It has an encoder on it that it needs to function. So you never have to worry about losing position. Of course, you still need to size it properly.
Most all hobby systems have stepper systems. Servo systems were always just too expensive. However, in just the last couple of years, and with ebay, hobbyists can now get into servos for a very reasonable cost.
Just want to add all Taig and Sherline machines, and most anything else on the hobby side are stepper systems. They all work well. It again comes down to sizing the machine properly.
In basic terms:
A stepper controller tells the stepper to go to X new position. Whether it actually gets to X new position it has no idea. So if you don't size the stepper properly for the mass you are moving or cut you are taking, it will lose track of where it is.
A servo system by design always knows where it is. It has an encoder on it that it needs to function. So you never have to worry about losing position. Of course, you still need to size it properly.
Most all hobby systems have stepper systems. Servo systems were always just too expensive. However, in just the last couple of years, and with ebay, hobbyists can now get into servos for a very reasonable cost.
Just want to add all Taig and Sherline machines, and most anything else on the hobby side are stepper systems. They all work well. It again comes down to sizing the machine properly.
Your original question was regarding accuracy. Neither is more accurate than the other when things are running normally.
Only if the machine is under designed or the machine crashes will the stepper system lose track of where it is.
Only if the machine is under designed or the machine crashes will the stepper system lose track of where it is.
Time is money
Steppers can be successfully used for most any cuemaking process. Servos are available in much bigger and faster forms.
I have personally spent a lot more time that I would like to have sitting there waiting for my machine to make 'rapid' moves at 30 inches per minute. This is much faster than needed for cutting, but a snails pace when just moving things around. Some tweak their machines and get maybe 50-100 ipm rapids but the servos can just plain move heavier objects faster. This is why it's important to keep the mass of the moving parts as low as possible (especially with steppers). I may someday replace my steppers on my cnc machine for this reason alone.
Servos can be more accurate than steppers in a closed system with encoders on the axes rather than on the motors, but that ain't cheap.
I don't think that servos with encoders on the motor offer any substantial accuracy benefit over steppers.
You can get more money but you only get so much time.
JoeyInCali said:
Steppers can be successfully used for most any cuemaking process. Servos are available in much bigger and faster forms.
I have personally spent a lot more time that I would like to have sitting there waiting for my machine to make 'rapid' moves at 30 inches per minute. This is much faster than needed for cutting, but a snails pace when just moving things around. Some tweak their machines and get maybe 50-100 ipm rapids but the servos can just plain move heavier objects faster. This is why it's important to keep the mass of the moving parts as low as possible (especially with steppers). I may someday replace my steppers on my cnc machine for this reason alone.
Servos can be more accurate than steppers in a closed system with encoders on the axes rather than on the motors, but that ain't cheap.
I don't think that servos with encoders on the motor offer any substantial accuracy benefit over steppers.
You can get more money but you only get so much time.
THANK YOU for all the response.
RBC said:
Royce, I too used the Xylotex board. I now have a Gecko G540 on my storm. I am very impressed with this box. It has the capability of switching on and off your router motor via a relay that can be wired in. The G540 has an isolated circuit that can handle the current to drive a 24v relay coil. By the way thanks for the link, I will check it out as I am going to make another machine soon as was thinking of going servo motors this time. I was thinking of using Gecko drives with the smooth stepper board. They motors will of course have feedback. From what I can tell, the feedback will actually go back to the amplifiers instead of the cotroller, is this correct?
Jim.
Thnx Jim.Mc2 said:
I spoke to Jeff of Xylotex about having spindle on and off outlet. He said he was working on it.
JoeyInCali said:
Xylotex did sell a relay board awhile back. I used it on my first taper shaper. The only thing is the relay would be active until Mach3 was loaded then would shut off. If my router motor was on when computer booted then it would run. Not a good way to be wired.
If you are looking at building your own system and have not done so befor, I would recomend going stepper motors. This will be the easier of the two.
Jim.
Just a little side comment on the accuracy part and Stepper VS Servo:
There is no need to buy beefy expensive servo controller setups, you could use this device with either a cheap optical rotational encoder on the back of the stepper and you are there :
http://www.safeguardrobotics.com/default.aspx?tab=cncbrain
If you have the money, linear scales and the like could be added but of course more expensive.
"The Brain" seems to be a very good product but it's still in beta testing and the Mach3 plugin is still not made.
I'm considering this for my Gantry setup and possibly expand it with encoders and feedback when the time and money is there.
Kent
There is no need to buy beefy expensive servo controller setups, you could use this device with either a cheap optical rotational encoder on the back of the stepper and you are there :
http://www.safeguardrobotics.com/default.aspx?tab=cncbrain
If you have the money, linear scales and the like could be added but of course more expensive.
"The Brain" seems to be a very good product but it's still in beta testing and the Mach3 plugin is still not made.
I'm considering this for my Gantry setup and possibly expand it with encoders and feedback when the time and money is there.
Kent
RBC
Deceased
Jim,
I am not that familiar with the smooth stepper board, so I can't say one way or another.
Way back, when I first started looking into CNC I read a document put out by Marris at Gecko that talked about the differences between steppers and servos. He talked in depth about the tuning involved in servos and the difficulties in coordinating multiple axis'. Basically, he said that if you can get the job done with steppers, you should do it with steppers, period.
As far as speed goes, I have stepper driven systems that can move at 300ipm rapids without fear of losing steps. And this is with the small Xylotex drives and a 28 volt power supply. If using the Gecko 540 and up to 50 volts, you could get even better. The secret is the lead screw. I use acme screws that are 5 start 10 TPI. That means 2 revolutions per inch of travel. I still have plenty of resolution for the long axis, but I get fast movement and keep the steppers rpm low enough that the motors are strong. Remember, that a steppers rating is HOLDING torque, and that strength falls off rapidly as the rpm goes up. So a long lead screw will get things moving while keeping the rpm low and the motor strong.
Royce Bunnell
www.obcues.com
I am not that familiar with the smooth stepper board, so I can't say one way or another.
Way back, when I first started looking into CNC I read a document put out by Marris at Gecko that talked about the differences between steppers and servos. He talked in depth about the tuning involved in servos and the difficulties in coordinating multiple axis'. Basically, he said that if you can get the job done with steppers, you should do it with steppers, period.
As far as speed goes, I have stepper driven systems that can move at 300ipm rapids without fear of losing steps. And this is with the small Xylotex drives and a 28 volt power supply. If using the Gecko 540 and up to 50 volts, you could get even better. The secret is the lead screw. I use acme screws that are 5 start 10 TPI. That means 2 revolutions per inch of travel. I still have plenty of resolution for the long axis, but I get fast movement and keep the steppers rpm low enough that the motors are strong. Remember, that a steppers rating is HOLDING torque, and that strength falls off rapidly as the rpm goes up. So a long lead screw will get things moving while keeping the rpm low and the motor strong.
Royce Bunnell
www.obcues.com
RBC said:
Royce, I have used both the xylotec and the gecko stepper boards. If I had to purchase another, I would go gecko. It is a little harder to setup. You will need to connect a resistor to two pins of the connector to set current max of the motors. This isn't that hard as gecko gives you the four 9 pin connectors to use. The gecko will give you 10 micro stepping instead of 8. The gecko has a true macro pump built into it. This is easy to setup in Mach3 for the e-stop. When e-stop is pressed the steppers are released and can be moved freely. Not that I would want to. I do plan on adding a 24v relay for my Kress. This should be easy as the gecko can drive up to 1 amp. with two of its outputs. You will only have to supply the external 24v. I had to add a board to the xylotec for that. And then it wasn't safe as when Mach3 wasn't active the relay was fired.
The smoothstepper is an interface that Mach3 has a plugin for. It is a usb interface that you could plug your xylotec board directly into it. The upside of this would be if you wanted to run additional axes or need additional imputs like feedback. It simulates two printer ports. It is getting harder to find printer ports and the printer ports on laptops are a crap shoot. This board is just an interface. You will still need some sort of motor driver. This would allow somebody to mix steppers with dc servos.
By the way Royce could you post some links or part numbers for those lead screws.
Jim.
RBC
Deceased
Jim,
Go to www.Mcmastercar.com, and then enter this part number:
98940A020
(Same as 98940A204)
4140 Alloy Steel Precision Acme Threaded Rod 1/2"-10 Sz, 1/2" Travel/Turn, 6'L, Rh Thread, 5 Starts
In stock at $78.34 Each
Their online catalog is weird. It is always the same url no matter what page you are on.
I also use the anti backlash nuts from www.dumpstercnc.com. They do have allot of drag, but they have zero backlash and last forever.
Royce Bunnell
www.obcues.com
Go to www.Mcmastercar.com, and then enter this part number:
98940A020
(Same as 98940A204)
4140 Alloy Steel Precision Acme Threaded Rod 1/2"-10 Sz, 1/2" Travel/Turn, 6'L, Rh Thread, 5 Starts
In stock at $78.34 Each
Their online catalog is weird. It is always the same url no matter what page you are on.
I also use the anti backlash nuts from www.dumpstercnc.com. They do have allot of drag, but they have zero backlash and last forever.
Royce Bunnell
www.obcues.com
CNC zone
Joey,
You probably already know about this site but I'm posting a link at the bottom just in case. The CNC zone will give you access to all the information you can stand about automated machine operations. Steppers can and do accurately drive far bigger equipment than we have any use for in a cue shop. As a general statement they are cheaper to buy and set up than servo's and speed and strength should not be issues if you spec them out properly.
Steppers are "dumb" as already mentioned if they lose steps when overloaded they don't know it and every operation will have an error in it until the machine is rezeroed. This was a big concern when I first looked at NC but when I explored things more deeply I found that it wasn't an issue for those of us that can happily run our machinery at 90% or less of full load. If pushing the machine to it's limits for maximum production is the goal I am more comfortable with servo's, if I am not running the machine at the ragged edge of crashing to meet production demands steppers suit me fine.
Hu
http://www.cnczone.com/forums/
Joey,
You probably already know about this site but I'm posting a link at the bottom just in case. The CNC zone will give you access to all the information you can stand about automated machine operations. Steppers can and do accurately drive far bigger equipment than we have any use for in a cue shop. As a general statement they are cheaper to buy and set up than servo's and speed and strength should not be issues if you spec them out properly.
Steppers are "dumb" as already mentioned if they lose steps when overloaded they don't know it and every operation will have an error in it until the machine is rezeroed. This was a big concern when I first looked at NC but when I explored things more deeply I found that it wasn't an issue for those of us that can happily run our machinery at 90% or less of full load. If pushing the machine to it's limits for maximum production is the goal I am more comfortable with servo's, if I am not running the machine at the ragged edge of crashing to meet production demands steppers suit me fine.
Hu
http://www.cnczone.com/forums/
Thansk Royce.
It must be this one.
98940A020
(Same as 98940A204)
4140 Alloy Steel Precision Acme Threaded Rod 1/2"-10 Sz, 1/2" Travel/Turn, 6'L, Rh Thread, 5 Starts
It must be this one.
98940A020
(Same as 98940A204)
4140 Alloy Steel Precision Acme Threaded Rod 1/2"-10 Sz, 1/2" Travel/Turn, 6'L, Rh Thread, 5 Starts
RBC
Deceased
My next door neighbors use linear motors.
If you have never seen them, imagine a motor that is flat and not round, and it is stretched out over the length of the axis movement. It is operated like a servo, but the encoder is also magnetic and runs along the length of the axis. There are no lead screws. The encoders are accurate to microns.
These machines move at over 1000 ipm and hold tolerances that are difficult to measure without special measurement devices. Imagine trying to judge how tight to squeeze the micrometer thimble, or what the temperature of the micrometer and the components are because any of that would have a huge affect on the measurement.
For smaller simple movements, they also use stepper motors with tremendous success. They pick up chips that are smaller than .005" across and move them from one chip to another. I mean 88,000 chips fit on one slice that is about 7 inches across.
So, I think steppers are just fine for what we do. Set up properly, they are tremendously accurate. With proper home switches and programming, they are tremendously consistent with no adjustments. Servos may look like the answer, but we might need to consider if we are using steppers to their fullest potential.
Royce Bunnell
www.obcues.com
If you have never seen them, imagine a motor that is flat and not round, and it is stretched out over the length of the axis movement. It is operated like a servo, but the encoder is also magnetic and runs along the length of the axis. There are no lead screws. The encoders are accurate to microns.
These machines move at over 1000 ipm and hold tolerances that are difficult to measure without special measurement devices. Imagine trying to judge how tight to squeeze the micrometer thimble, or what the temperature of the micrometer and the components are because any of that would have a huge affect on the measurement.
For smaller simple movements, they also use stepper motors with tremendous success. They pick up chips that are smaller than .005" across and move them from one chip to another. I mean 88,000 chips fit on one slice that is about 7 inches across.
So, I think steppers are just fine for what we do. Set up properly, they are tremendously accurate. With proper home switches and programming, they are tremendously consistent with no adjustments. Servos may look like the answer, but we might need to consider if we are using steppers to their fullest potential.
Royce Bunnell
www.obcues.com