Maker Pro
Maker Pro

Motor Discussion

donkey

Feb 26, 2011
1,301
Joined
Feb 26, 2011
Messages
1,301
Hey all,
just a quicken, lets say I have designed a method of accuratel positioning a dc motor (well its geared and has an encoder on it) would thaat be a better method than using a stepper motor?
I am after something that has a lot of torque and is very precise, preferably fast too but the first 2 are a must
 

Minder

Apr 24, 2015
3,478
Joined
Apr 24, 2015
Messages
3,478
Stepper motor is the simplest as it requires no feedback, a DC motor for precise positioning requires the encoder with PID feedback to the controller.
M.
 

donkey

Feb 26, 2011
1,301
Joined
Feb 26, 2011
Messages
1,301
yup thats fine, which will provide more torque though?
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
Torque and speed are all over the map when it comes to DC motor drives. I have a pair of DC motors that came off a computer tape drive. These were used as servo motors to spin the reels very fast but stop on a dime when the desired area of the tape was found. I was going to use them with a brand-new 48 V DC lead-acid battery to build an electric go-cart, but my kids all grew up and left home before I could find my "round tuit".

If you are using gears with your DC motor, don't expect both speed and high-torque or even accuracy. Special servo motors I have used in the past had special zero-back-lash, split, spring-loaded, gears. They were very fast and very accurate because they computed ballistic coefficients for aiming a large cannon, mounted on hydraulically actuated gimbals, that fired while in flight. But each servo assembly probably cost the tax payers upwards of $50,000 dollars in 1960s money. You pays for accuracy, repeat-ability, torque, and speed. I like stepper motors, and many I have used incorporate planetary gear reductions, but adding a shaft position encoder can be very tricky and you have to watch out for mechanical resonances excited by the stepper impulses. In many ways, a DC servo motor is a better choice, but only experience will point you in the right direction. Well, it doesn't have to be your personal experience. Learn from others and tell us what you want to DO. Specs (numbers) are a big help.
 

Minder

Apr 24, 2015
3,478
Joined
Apr 24, 2015
Messages
3,478
Hey all,
just a quicken, lets say I have designed a method of accuratel positioning a dc motor (well its geared and has an encoder on it) would thaat be a better method than using a stepper motor?
I am after something that has a lot of torque and is very precise, preferably fast too but the first 2 are a must
If you have already designed it and it works then generally the DC motor with encoder feedback is superior to stepper, but PID loop is required.
Higher end CNC machines use servo's over steppers.
M.
 

donkey

Feb 26, 2011
1,301
Joined
Feb 26, 2011
Messages
1,301
thanks minder.
the idea is for something like a CNC but it needs continuous rotations which servo's don't do, and I wanted to be able to detect 1 degree of rotation which I don't think steppers do.

the only issue with the setup is I have to upgrade the encoder as the one I have only detects 15 steps per second..... kinda want 360ish lol
 

donkey

Feb 26, 2011
1,301
Joined
Feb 26, 2011
Messages
1,301
just an add-on to this I will code this sucker up and do a short video if it works to show part 1 of a grand scheme
 

Minder

Apr 24, 2015
3,478
Joined
Apr 24, 2015
Messages
3,478
the idea is for something like a CNC but it needs continuous rotations which servo's don't do, and I wanted to be able to detect 1 degree of rotation which I don't think steppers do.

I assume you are using the term servo here as in RC application.?
In a CNC a servo is very accurate and positioning with high rpm and continuous rotations.
M.
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
Servo (servomechanism): a device that accurately and quickly responds to a command that represents position, velocity, acceleration, angle, or any other physically measurable output. Donkey, please tell us what you are trying to DO!
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
thanks minder.
the idea is for something like a CNC but it needs continuous rotations which servo's don't do, and I wanted to be able to detect 1 degree of rotation which I don't think steppers do.

the only issue with the setup is I have to upgrade the encoder as the one I have only detects 15 steps per second..... kinda want 360ish lol
Steppers are typically 1.8 degrees, but with gearing you can get any step size you want.

Bob
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Steppers are typically 1.8 degrees, but with gearing you can get any step size you want.

I.e. gear it down by a factor of 9 and then 5 steps is one degree.

Bob
 

Minder

Apr 24, 2015
3,478
Joined
Apr 24, 2015
Messages
3,478
Steppers can also be 1/2 and micro stepped, but at a cost of torque.
M.
 

Doug3004

Sep 5, 2014
119
Joined
Sep 5, 2014
Messages
119
Steppers are typically 1.8 degrees, but with gearing you can get any step size you want.

I.e. gear it down by a factor of 9 and then 5 steps is one degree.
Bob
Steppers can also be 1/2 and micro stepped, but at a cost of torque.
M.
The cheaper Chinese-made CNC-style stepper drivers allow artificial step rates; for larger motors they usually begin at 400 steps/turn and go up to around 25,000 steps per turn. None of the available selections falls on a value that divides cleanly into 360 degrees tho.

I have noticed that (with the China-origin parts at least) 200-step motors are cheap and common but 180 or 360-step motors are rare & expensive... And 400-step optical encoders are common but 360-step optical encoders are also rare & expensive. It's cheaper to use gearing if you want some kind of compass-point step count.

There is no motor that can turn very precisely at low speeds, and still turn at very high speeds.
CNC machining equipment that does this cheats, by either using one motor and a two-speed gearbox, or by using two separate motors+encoders (one built for low-RPM/precise positioning, and the other built for high-RPM use).
One example:
In this video, the machine does high-speed turning then switches to slow machining at the 1:50 time mark.
 

Minder

Apr 24, 2015
3,478
Joined
Apr 24, 2015
Messages
3,478
There is no motor that can turn very precisely at low speeds, and still turn at very high speeds.
CNC machining equipment that does this cheats, by either using one motor and a two-speed gearbox, or by using two separate motors+encoders (one built for low-RPM/precise positioning, and the other built for high-RPM use).
One example:
.
Have done it many times using high quality motion cards such as Galil and quality servo's, BLDC are capable of higher max rpm than DC brushed. Go to the Galil Motion site and check out the applications and instructional video's
The reason the spindle slow down at 1:50 was the spindle has C axis positioning and constant surface feed capability and the XZ axis were performing machining in the XYC 3 axis mode.
The torque curve of most servo motor is maximum at zero rpm with a slight drop at reaching maximum rpm.
M.
 

Minder

Apr 24, 2015
3,478
Joined
Apr 24, 2015
Messages
3,478
Another example of CNC lathe XYC and Live Tooling at 1:15..
M.
 

Doug3004

Sep 5, 2014
119
Joined
Sep 5, 2014
Messages
119
Another example of CNC lathe XYC and Live Tooling at 1:15..
M.
Yea but Haas makes what are smaller and cheaper CNC equipment and even at Haas, a 2-speed spindle gearbox is an option that is offered on most of their machines....
On that very ($68,000) machine, the 2-speed gearbox adds $7,500 onto the price. They must have thought that a lot of people would want it?
http://www.haascnc.com/mt_spec1.asp?id=ST-30&webID=2AXIS_STD_LATHE#gsc.tab=0

It's a bit academic I suppose. Spindle motors on these machines are 20-30 HP, much larger than the axis control motors.

An r/c servo does all that.
I don't like RC servos much, outside of RC toys.
They are relatively fragile and require a constant PWM signal to control--even just to hold still.

For most non-RC uses, a stepper motor is a better choice. It is bigger, heavier and more expensive but it has way more torque (no gears needed and much higher durability and precision) and only needs microcontroller input when you want it to take a step. If it is left enabled, it holds its position on its own.

I think people tend towards using servos because they always come with a bellcrank... stepper motors don't normally have that included, and good ones usually cost a few bucks. :|
 

donkey

Feb 26, 2011
1,301
Joined
Feb 26, 2011
Messages
1,301
ok the reason I am asking is the application requires accuracy, speed is good also but not necessary, but it really needs to hold about 2kg weight on the end of a pole (about a metre long). oh and the pole twists and turns lol
I know I am going to be told this is hard and I agree whole heartedly but if it works then yay me lol. I was just thinking a DC motor with a Hbridge is easier for myself to make and code then stepper motor drivers and coding. also no matter which I use they will both have an encoder put on them to check the motor has turned and that it is where its supposed to be.
 

Merlin3189

Aug 4, 2011
250
Joined
Aug 4, 2011
Messages
250
I know steppers are pretty good these days, but they are still open loop control. Depending on the application of course, I still prefer closing the loop, by adding position encoders whatever sort of motor is used.
 
Top