I am working on a 5th axis for the mill. I think it will need a brake.
I'm not sure if the servo motor will be able to hold position for some
operations. I have looked at magnetic particle brakes and they ain't
cheap. So I'm thinking about making one. Or some type of
electromagnetic brake. The brake, when actuated, must not cause any
motion of the axis, and I'm just not sure what would work best.
On Thursday, January 31, 2013 12:20:28 PM UTC-5, email@example.com wrote:
How much space do you have, and how often will it be locked and unlocked?
What is the rotational range of motion required?
Lots of choices, much fewer depending on the constraints.
A dirt cheap/simple brake would be a few interleaved plates, alternating be
tween turning with one side or the other, and a clamp (solenoid? screw clam
p with gear motor?) clamping the stack together. Would impart no motion, an
d could be in the clamped state with no power applied.
On Thu, 31 Jan 2013 09:20:28 -0800, firstname.lastname@example.org wrote:
I've had several occasions to put brakes on servos, and a couple where
it was important not to disturb the position. It's not as easy as it
sounds, depending on what "any" in "must not cause any
motion of the axis" means. The fussiest one I did involved locking a
table on a direct drive servo on a machine to dril tiny radial holes
in a part. It took a couple tries to get the disturbance under a few
tenths. The final design was a disc brake with the pads mounted on
flexures (flexure = zero free play) and actuated by an air cylinder.
On Jan 31, 12:20 pm, email@example.com wrote:
I have never made one. But looked at Herbach and Rademan and saw they
have clutches from Deltran and Mitsubishi. Doubt if they are robust
enough, but googled Deltran and found that Danaher offers a guide to
Deltran clutches and Brakes. So you might look at Herback and Rademan
and get some part numbers . And then look at the guide to Clutches and
I always look at what is available and see how they did it before
designing my own.
THOSE I'm familiar with. The action might cause a rotational hit since
the clutches are usually on spring-steel strips. But they're strong
I was thinking disc brake style, too. The disc would be mounted to
the lead screw (if there is one) and opposing solenoids would grab it
from either side.
Eric, if added mass wouldn't be a problem, might you increase the
servo size, resulting in a higher holding capacity and bypassing the
need for a brake?
Ive long been passionate about protecting and expanding democracy,
which is really the only viable mechanism to preserve liberty and
DC servos? What amount of motion can you tolerate. I would
expect with steady/slowly-changing loads no more than one or two encoder
steps of motion before it was forced to be right again. (Not sure
where your encoder is relative to the motor, but if you don't have
enough backlash between the axis and the motor for brake to do it, I
would not worry about the ability of the encoder/servo pair to hold
Now -- if there are serious pulse loads, then you might have a
little more error before the servo loop could fix it.
Note that at least some DC servos and AC servos as well have
built-in brakes which the controller has to power to release prior to
There will be pulsed loads. Such as when an end mill is spotfacing
and/or counterboring a curved surface. I don't know yet if I will be
contour milling with the 5th axis. Probably not at first. I just need
to be able index the parts. To do this I will be using an M code to
send a signal to an Arduino. The Arduino will then command the 5th
axis to move. I'm still learning how to use the Arduino and the LCD
touchscreen display I bought for the Arduino. I still need to figure
out how to get the signals from the encoder to the servo amp. I don't
know yet if slip rings will do or if I need to get some type of
wireless setup. But I need to approach this project with what I do
best first, and that's mechanical stuff. I think I have the drive
figured out, now I need the brake.
The part of the encoder which requires connections should not be
moving relative to the motor -- so cables to the motor and cables to the
encoder could be strapped together -- but shield the encoder cables to
avoid noise from the motor from introducing errors.
The encoder (on a motor, at least) is typically a slotted disc
(or a glass/quartz disc with slots printed on it by evaporation of
metal). This is mounted to the back end of the motor shaft. Bolted to
the back of the motor housing is a set of LEDs and photodetectors (along
with matching slots in a stationary part). Wires go from these to the
Arduino or whatever you use.
Just plain cables for the encoder connections.
Now -- if you want the brake which started this, you typically
have a brake disc on the shaft, with a spring mounting (not much give)
and an electromagnet mounted tot he motor to attract the disc to contact
a friction surface.
Adding both encoders and a brake to the back end of the motor
will be a bit tricky. An alternate encoder is one to directly measure
the motion of whatever -- I'm assuming a rotary table or something
similar -- and measuring at the table instead of the motor means that
you don't have as much resolution for the same encoder. For linear
motion, there are linear encoders which mount to the moving part, with a
reading head bolted to the stationary part -- well sort of stationary,
such as the head bolted to the saddle of a mill, and the encoder bar
(which has no wires) bolted to the edge of the table so the table pulls
the bar past the reading head.
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.