Advice on Roboclaw Size
Advice on Roboclaw Size
Hello,
I'm taking on my first robot project that uses large motors and needs a larger motor controller. I need some advice on selecting a proper motor controller. I've decided that the Roboclaw has all the features I want but what I'm unsure about is the Amp rating to pick. I'll be using four of these motors to control a bot that will probably end up weighing 75-100 lbs eventually.
http://www.andymark.com/CIM-Motor-p/am-0255.htm
Would the 15A version be sufficient or do I need to go much larger? There's a large gap between the free current (2.7A) rating and the stall current (133A) rating. I assume I don't need to cover the entire 133A but I'm not sure what formula to use in making the selection.
If I choose a motor controller that is under powered will I damage the controller or will the motors just be under powered?
Thanks!
Ian
I'm taking on my first robot project that uses large motors and needs a larger motor controller. I need some advice on selecting a proper motor controller. I've decided that the Roboclaw has all the features I want but what I'm unsure about is the Amp rating to pick. I'll be using four of these motors to control a bot that will probably end up weighing 75-100 lbs eventually.
http://www.andymark.com/CIM-Motor-p/am-0255.htm
Would the 15A version be sufficient or do I need to go much larger? There's a large gap between the free current (2.7A) rating and the stall current (133A) rating. I assume I don't need to cover the entire 133A but I'm not sure what formula to use in making the selection.
If I choose a motor controller that is under powered will I damage the controller or will the motors just be under powered?
Thanks!
Ian
-
Basicmicro Support
- Posts: 1594
- Joined: Thu Feb 26, 2015 9:45 pm
Re: Advice on Roboclaw Size
You have two options.
1. Assume the motors will stall at maximum duty cycle so 133amps * 4 motors is what you would need. Easiest to assume this but would require a massive motor controller.
2. Assume the motors will not stall at or near maximum duty cycle. The maximum current draw will depend on how you have the motors attached to their load(eg if they are driving wheels are they direct drive or gear or belt driven and if so what is the ratio?). Since you say it is a bot I'm assuming wheels or tracks. If so will the wheels/tracks slip if the robot runs into a wall or is it possible they will stall the motors out. These are pretty powerfull motors but if the motors are not geared to their load they will have a more extreme startup current while if they are geared this will greating reduce the startup current.
The freewheel current of a motor is pointless to look at in most cases(other than for efficiency comparisons). The only important number is the stall current. Other numbers(like continous current limit) also are pointless when sizing your controller though they do have relevance on motor wear. Since the stall current is 133a on these motors you can assume under most conditions(like driving your bot) the current draw will be between 25% to 100% of the stall current. On startup if you slam the power full blast you are going to hit the stall current or really close to it. If you have a ramping acceleration you can reduce this startup current peak quite a bit. If you slam into walls or other bots whille running at full power you will hit or come close to the stall current as well. If you slam the bot forward full and backwards full power you can potentiall hit more than stall current though the Roboclaw automatically limits rate of direction change to 100ms for a 100%+ to 100%- power change which helps limit this.
I can say with pretty good confidence there is no chance a 2x15a controller is enough for this setup in any way shape or form. Sorry.
At a bare minimum this setup probably would require a 2x60a board(eg peak current of 120a per channel though it will heat up fast if you are going over 60a per channel). Even then its peak rated current is still less than half the worse case current two of those motors could pull under full stall at 100% duty. Its close enough the current limiting should protect the board and prevent the motors from damaging the controller though.
The other possible cause of damage to a motor controller is ripple current in the DC link caps on the controller(eg the big caps you see in the pictures of the controllers). I would need the inductance of the motor and the expected battery voltage you would be running at to make sure that isnt a problem. The 2x60a can handle motors with very low inductance so I doubt it would be a problem but if you can get the inductance then we can make sure before you buy the motors/controller.
One trick you can do to reduce the maximum stall current is to run the paired motors(eg the two motors on each Roboclaw channel) in series instead of parallel. This will double the resistance of the motor coils instead of cutting them in half(as parrallel connection would do) so the two motors combined will have a stalled current of about 67amps total instead of 266amps. This will reduce the maximum power(by alot I suspect) from the motors as well but it would put you into the ballpark for a 2x30a controller.
This would also double the inductance the motor controller sees which would almost certainly make it high enough to be safe to use on the much smaller controller without adding extra DC Link capacitance(unless these are supper low inductance motors).
1. Assume the motors will stall at maximum duty cycle so 133amps * 4 motors is what you would need. Easiest to assume this but would require a massive motor controller.
2. Assume the motors will not stall at or near maximum duty cycle. The maximum current draw will depend on how you have the motors attached to their load(eg if they are driving wheels are they direct drive or gear or belt driven and if so what is the ratio?). Since you say it is a bot I'm assuming wheels or tracks. If so will the wheels/tracks slip if the robot runs into a wall or is it possible they will stall the motors out. These are pretty powerfull motors but if the motors are not geared to their load they will have a more extreme startup current while if they are geared this will greating reduce the startup current.
The freewheel current of a motor is pointless to look at in most cases(other than for efficiency comparisons). The only important number is the stall current. Other numbers(like continous current limit) also are pointless when sizing your controller though they do have relevance on motor wear. Since the stall current is 133a on these motors you can assume under most conditions(like driving your bot) the current draw will be between 25% to 100% of the stall current. On startup if you slam the power full blast you are going to hit the stall current or really close to it. If you have a ramping acceleration you can reduce this startup current peak quite a bit. If you slam into walls or other bots whille running at full power you will hit or come close to the stall current as well. If you slam the bot forward full and backwards full power you can potentiall hit more than stall current though the Roboclaw automatically limits rate of direction change to 100ms for a 100%+ to 100%- power change which helps limit this.
I can say with pretty good confidence there is no chance a 2x15a controller is enough for this setup in any way shape or form. Sorry.
At a bare minimum this setup probably would require a 2x60a board(eg peak current of 120a per channel though it will heat up fast if you are going over 60a per channel). Even then its peak rated current is still less than half the worse case current two of those motors could pull under full stall at 100% duty. Its close enough the current limiting should protect the board and prevent the motors from damaging the controller though.
The other possible cause of damage to a motor controller is ripple current in the DC link caps on the controller(eg the big caps you see in the pictures of the controllers). I would need the inductance of the motor and the expected battery voltage you would be running at to make sure that isnt a problem. The 2x60a can handle motors with very low inductance so I doubt it would be a problem but if you can get the inductance then we can make sure before you buy the motors/controller.
One trick you can do to reduce the maximum stall current is to run the paired motors(eg the two motors on each Roboclaw channel) in series instead of parallel. This will double the resistance of the motor coils instead of cutting them in half(as parrallel connection would do) so the two motors combined will have a stalled current of about 67amps total instead of 266amps. This will reduce the maximum power(by alot I suspect) from the motors as well but it would put you into the ballpark for a 2x30a controller.
This would also double the inductance the motor controller sees which would almost certainly make it high enough to be safe to use on the much smaller controller without adding extra DC Link capacitance(unless these are supper low inductance motors).
Re: Advice on Roboclaw Size
@acidtech - thank you for the great reply. I apologize for just noticing it was there. Apparently, this forum doesn't automatically send email notifications. I had already decided to go with 2x60A controllers. So, it sounds like you agree that I'll be fine. Unfortunately, one of the controllers seems to have a channel appears to be DOA. I've submitted a tech support ticket to try and get a replacement. 3 of the 4 motors are driving well, though.
Thanks!
Ian
Thanks!
Ian
-
Basicmicro Support
- Posts: 1594
- Joined: Thu Feb 26, 2015 9:45 pm
Re: Advice on Roboclaw Size
I think we already shipped out the replacement. I beleive we talked on the phone this morning.