DC motors worthy of being a servo?

[WoodSnarfer]

Trying to learn as much as possible about motors...specifically servos. I'm convinced that my (to be designed) router should use servos.

Figuring an aluminum gantry, with a 3.25 HP router/spindle, and a work area of about 24" by 36"...I guess I'd have to really figure out the speed and torque requirements...but maybe there are some 'general' specs we can talk about regarding DC motors:

1. What makes a DC motor worthy of being called a 'servo'? Is it just the fact that it'll spin at around 3,000RPM, it has adequate HP, and you can stick an encoder on it good enough? Or are there other specs it needs to meet (duty cycle?).

2. My "gut feel" is that for my x-axis, I'd need a hefty motor -- maybe 1/2 HP with a 1/2" or larger output shaft. Say I were to find a cheap motor like this, and it can be altered to accept an encoder, and it spins at 3,000RPM, and it likes 30 to 36VDC. Would that be something to look at, for my axis?

3. I see a lot of 12VDC motors...but I'm assuming that they just won't provide the torque needed. Usually, they seem to be for automotive applications (emissions, or blowers, etc). No good for servo adaptation, right?

4. I also see a lot of 90VDC motors, and a few 'in between' (40V to 60V). If the other specs look right (RPMS, HP), can they be used? Or, will the power supply requirements be too great?

As an example, here's a motor I was looking at tonight. I think the price is kind of high...and I suppose the rear cover would have to be removed and the shaft tapped to put a smaller shaft into it, for the encoder...but in general, would this motor "cut it" as a decent servo?

http://cgi.ebay.com/ws/eBayISAPI.dll...tem=5934321133

It's 50V...looks like it is well made...if I ever tripped across one similar for around $10-$15, would it be worth using?

Thanks!
-Chris

[Noah]

Actually, I think Ametek makes that motor in a servo version. There should be some identical ones on Ebay with encoders selling for about $150.00 each. That might be that guy's secret. Yes I think it would work just fine. One other thing to look for in a motor is timing. If the motor has slightly advanced or retarded timing, it will perform differently in forward than it does in reverse. This is very bad for a servo. Many motors made for certain applications will have their timing advanced a couple of degrees. Obviously this is only an issue with brushed motors. I think pretty much all brushless motors are servo grade.

[arvidb]

Originally Posted by WoodSnarfer 1. What makes a DC motor worthy of being called a 'servo'? Is it just the fact that it'll spin at around 3,000RPM, it has adequate HP, and you can stick an encoder on it good enough? Or are there other specs it needs to meet (duty cycle?).

I think servo motors are often made to have a small rotary inertia to be able to change speed fast. Noah also made some very good comments about the placement of the brushes on brushed motors. The term servo motor I guess should also mean the motor is good quality with low cogging - the rumor has it you should be able to run a 60 V servo motor with a battery (it should not "stick"). I believe this - my 200 V AC servo motors aligns quite forcefully with the field produced from a C cell battery (as AC motors they won't spin, of course).

A motor in itself is not a servo. What turns it into a servo is the drive - a servo is just something that "serves" you; i.e. it goes where you tell it (or as fast as you tell it etc) and it also makes sure it does that by feedback. To do this it needs a drive/controller of some kind.

Another difference is that a "generic" small 12 V motor is made to run at 12 V. A servo motor marked 200 V might burn if you just hooked it up to a 12 V supply! It needs to be hooked up through the drive to limit current.

Originally Posted by WoodSnarfer 4. I also see a lot of 90VDC motors, and a few 'in between' (40V to 60V). If the other specs look right (RPMS, HP), can they be used? Or, will the power supply requirements be too great?

A higher voltage motor is often more efficient (less copper losses). Higher voltage means less current for the same power, and current makes things hot. So the power supply will probably be cheaper to build for a higher voltage motor, but it's also more dangerous. So I guess in the end the suitability of the motor depends on if you can find a fitting drive for it and if you feel comfortable with the voltage required.

Arvid

[WoodSnarfer]

Thanks...your answers and Noah's helped out a lot. Funny you mentioned AC servos -- I was wondering about them, too. I was thinking there might be a way to wire up a 'standard' 120VAC tool motor (like from a 10" table saw) to a DC servo driver, by using relays to DC-switch the motor (and also to re-wire the motor from CW to CCW, depending on the direction signal). For a Y or Z axis drive, that might be silly -- but for a large router with a long X axis, maybe it'll be the answer for the required torque and speed.

Anyone ever try using a 120VAC motor, driven from a DC servo driver via relays?

Of course, I do recognize the requirement for an encoder, and the fact that the rotational inertia on a table saw motor might be too great for this...but I was wondering anyway

Thanks,
Chris

[ynneb]

Woody, I use these as servos. http://cgi.ebay.com/ws/eBayISAPI.dll...tem=3844568448

As you can see they look pretty similar to the picture you have.

I guess if you can fit an encoder on the back then they are the same.

These servos work for me very well.

You could save a lot of money if you can do the conversion yourself. I bought these servos a long time ago for 150 each. I guess if you have success with doing a conversion you will have a lot of others following you.

[Mariss Freimanis]

There is nothing that differentiates a cheap little 3-commutator segment toy motor from a sero motor except quality. The former is completely unsuitable for servo use because of torque ripple while a good quality utility motor may qualify.

Torque ripple is how much does torque change as the motor rotates over a revolution. Servo drives are tuned to a motor's torque response. Ideally they expect the same torque no matter what clock position the motor is at.

With cheap motors this varies a lot. With excellent motors it varies very little. Cheap motors can never be tuned, good ones tune easily.

What makes a good motor? A lot of commutator segments and a skewed (spirally twisted) low-cog armature. Brushes that are not preferentially set for a particular direction.

The range of motors from the cheapest to the most expensive is a spectrum. Pass a certain threshold and a utility motor can be a good servo motor.

Mariss

[WoodSnarfer]

Originally Posted by Mariss Freimani The range of motors from the cheapest to the most expensive is a spectrum. Pass a certain threshold and a utility motor can be a good servo motor. Mariss

Mariss, thanks...yes, I am trying to figure out the 'sweet spot' of that threshold, where I can find a good quality motor and turn it into a servo. Doing an eBay search on 'servo' means I will find items that whoever is selling the motor has built in a price premium because they know what they are selling. A surplus dealer who is just selling a good DC motor will not have that price premium considered, so maybe I can add a $40 encoder and make it into a real servo, at a reasonable overall price.

BTW, my goal is to use Gecko's, so I want to stay within the proper specs as I search out my motors.

Regards,
Chris

[Guy Smiley]

Can a 12 vdc bus wiper motor assembly (with speed reducer gears) be run and controlled by Gecko's? Thanks!

[ViperTX]

Guy Smiley...sure, but is the wiper motor reversible (direction...cw, ccw).

[Guy Smiley]

Originally Posted by ViperTX Guy Smiley...sure, but is the wiper motor reversible (direction...cw, ccw).

Yes it does. Actually, I've wired and mechanically attached an automobile wiper (Honda Civic's) to run the x-axis on my Enco type bench mill. I used 2 DPDT toggle switches to handle direction and speed (fast & slow). I also machined a brass sleeve to connect the motor shaft to the mill's drive shaft. I have been milling aluminium on it for my CNC Router project (documented in the DIY project log thread) but have kept to just max feeding of 0.5mm passes with a 3/8" 4-flute end mill. Works comfortably at that or less feed setting.

Now what I'm thinking of, is if, until I get the Gecko's. Can I get a Xylotex stepper driver board to control relays that will control these bigger wiper motors using a 20A regulated power supply (for the motor) or is there a better way? Using this board, do I need a certain circuitry to accurately control the voltage (cut-off/vent to another load the residual voltage if there's such a thing) from the relay to the motor? Or forget it and just use the Geckos.

I, as most here, am just trying to find an alternative way of doing homebrew CNC with readily available components and surplus wiper motors are sure abundant in auto parts recycling facilities. Thank you for responding.
EC

[Torchhead]

Originally Posted by Guy Smiley Yes it does. Actually, I've wired and mechanically attached an automobile wiper (Honda Civic's) to run the x-axis on my Enco type bench mill. I used 2 DPDT toggle switches to handle direction and speed (fast & slow). I also machined a brass sleeve to connect the motor shaft to the mill's drive shaft. I have been milling aluminium on it for my CNC Router project (documented in the DIY project log thread) but have kept to just max feeding of 0.5mm passes with a 3/8" 4-flute end mill. Works comfortably at that or less feed setting. Now what I'm thinking of, is if, until I get the Gecko's. Can I get a Xylotex stepper driver board to control relays that will control these bigger wiper motors using a 20A regulated power supply (for the motor) or is there a better way? Using this board, do I need a certain circuitry to accurately control the voltage (cut-off/vent to another load the residual voltage if there's such a thing) from the relay to the motor? Or forget it and just use the Geckos. I, as most here, am just trying to find an alternative way of doing homebrew CNC with readily available components and surplus wiper motors are sure abundant in auto parts recycling facilities. Thank you for responding. EC

Xylotex are bipolar microstepping stepper drivers rated at 30VDC (Max) and 2.5A per phase (Max). General DC motors cannot be used on any stepper system. Stepper motors have multiple coils and "poles" so they "step" to the next pole when you pulse the correct coil. The faster you pulse the faster they spin. They are current devices. meaning you control the current as you charge the coil. PM DC motors have a single winding and just spin when you apply DC volts. The more volts the faster they spin. No pulsing or coil switching is used.

The DC motors used in cars are often high current intermittent duty with sleave bushings instead of bearings. If gears are involved there will most certainly be hugh amounts of backlash making them unusable for any type of postional control. Modern servo controllers (Gecko's) close the loop inside the control. They will "dither" one or two encoder counts. If the encoder is not on the primary motor shaft then any system backlash is inserted into the loop equation. It's like driving a car at 60MPH with a really loose steering box and trying to stay within 1 " of the center line!

If you just want to build a machine that moves under computer control than using low precison parts is an option. If you want the machine to be useful beyond showing your friends, then you need to realize that some parts need to be of higher quality.

There are good DC motors that can be used with encoders as servos. They typically are rated for the continuous duty under load you will subject them to and have the torque and RPM that work with good CNC designs. Almost all servos need some form of speed reduction since they spin faster than steppers. Most designs use the toothed belt- pulley approach. The smaller the amount of backlash you can have in a system the better. Using rack and pinion drive a direct coupled servo will produce astounding rapids at really low torque. Leadscrews can help since they act as reduction at the TPI of the screw but can also introduce addtional errors and backlash.

Just hooking up some DC motors and a direction relay to control motor direction would not get you close to anything that would provide positional control.

Do a little more investigation into the Hobby CNC controls and read this list and look at some of the projects and how they were done. There is no substitute for experience (:-)

[Al_The_Man]

Originally Posted by Torchhead PM DC motors have a single winding and just spin when you apply DC volts. The more volts the faster they spin. No pulsing or coil switching is used.

Actually the PM DC motor, especially if a servo grade will have many pole windings and they are switched by means of the commutator.
Servo types will generally have 4 brushes, many pole windings and skewed rotor windings to eliminate low speed cogging.
Al.