Using hobby servo to rotate 360 degrees?

[Eric R Snow]

Greetings,
I can buy a chip that will convert the signal from an HP encoder,
rotary or linear, to step and direction which can be used with a
stepper driver chip to turn a stepper motor. What I want to do is use
the motor to spin a needle above a dial so the thing emulates a travel
dial indicator. But I need 1000 steps, not 100. I could use a
microstepping drive but these are expensive. I was wondering if an RC
servo with the pot removed could be used for this. And done cheaper.
Any ideas?
Thanks,
Eric R Snow

 

[Garrett Mace]

Greetings,
I can buy a chip that will convert the signal from an HP encoder,
rotary or linear, to step and direction which can be used with a
stepper driver chip to turn a stepper motor. What I want to do is use
the motor to spin a needle above a dial so the thing emulates a travel
dial indicator. But I need 1000 steps, not 100. I could use a
microstepping drive but these are expensive. I was wondering if an RC
servo with the pot removed could be used for this. And done cheaper.
Any ideas?
Thanks,
Eric R Snow

Cheaper, yes; meeting your specifications? No. You won't get 1000 repeatable
steps out of a servo. And without position feedback, it would be very
difficult to actually know how far the needle went; you couldn't say 1000
steps would equal one revolution.

Do you actually need 1000 positions of the needle around the dial? Or would
100 or 200 individual positions be ok, as long as it still took 1000 steps
to make one revolution? For the latter, you could simply use an up/down
counter IC to divide the input steps by 5 or 10.

 

[Dbowey]

Erik posted, in part:
<< I was wondering if an RC servo with the pot removed could be used for this.
And done cheaper. >>

That won't work. Without the pot the motor will spin continuously.

Don

 

[Eric R Snow]

Cheaper, yes; meeting your specifications? No. You won't get 1000 repeatable
steps out of a servo. And without position feedback, it would be very
difficult to actually know how far the needle went; you couldn't say 1000
steps would equal one revolution.

Do you actually need 1000 positions of the needle around the dial? Or would
100 or 200 individual positions be ok, as long as it still took 1000 steps
to make one revolution? For the latter, you could simply use an up/down
counter IC to divide the input steps by 5 or 10.

Yeah, I actually need the needle to have 1000 positions. You see, this
is to emulate a mechanical dial indicator. They can be read while
moving while rapidly changing numbers can't ( at least by me) be read.
So, I figured if I go to all the trouble of making some kind of
electronic dial indicator it would be neat if it was also accurate
then the digital display could be disposed of. Though I have 'em
(digital displays) all over the shop I prefer analog to digital when
just glancing at it. Gecko makes a small servo drive and US digital
sells a small, 1000 line encoder that could be used for feedback.

 

[Eric R Snow]

Erik posted, in part:
<< I was wondering if an RC servo with the pot removed could be used for this.
And done cheaper. >>

That won't work. Without the pot the motor will spin continuously.

Don

What if a 1000 line encoder is connected to the servo? I forgot to put
in my first post that's the plan.
ERS

 

[Garrett Mace]

Yeah, I actually need the needle to have 1000 positions. You see, this
is to emulate a mechanical dial indicator. They can be read while
moving while rapidly changing numbers can't ( at least by me) be read.
So, I figured if I go to all the trouble of making some kind of
electronic dial indicator it would be neat if it was also accurate
then the digital display could be disposed of. Though I have 'em
(digital displays) all over the shop I prefer analog to digital when
just glancing at it. Gecko makes a small servo drive and US digital
sells a small, 1000 line encoder that could be used for feedback.

Allegro sells cheap microstepping drive chips with onboard translators, but
I don't think they have a 5th or 10th microstep capability. So if you could
find a motor with a weird number of steps, like 500 or 250 or 125, you could
do it that way. But be warned that microstepping is used mainly to smooth
operation, and the steps divisions are not usually linear. So to use a
stepper motor, you really need to use something that cogs at each position.
I highly doubt you could really make a mechanical linkage and be able to
accurately eyeball 0.36 degree increments, but that's what you say you
need....

So, here's the plan: get your stepper chip and hook it up to the encoder.
Fine, you spin the encoder once, and the motor spins 10 times if it's a 3.6
degree motors, and 5 times if it's a 1.8 degree motor. Now, go to
http://www.sdp-si.com and buy some nice spur or worm gears in a 10:1 or 5:1
ratio. Their plastic gears are pretty good, but for maximum accuracy go for
a couple of their metal ones. Now you have a dial that makes 1000 steps in a
revolution. The plastic gears won't set you back more than a buck or two
each, the metal ones...a bit more.

 

[Eric R Snow]

Allegro sells cheap microstepping drive chips with onboard translators, but
I don't think they have a 5th or 10th microstep capability. So if you could
find a motor with a weird number of steps, like 500 or 250 or 125, you could
do it that way. But be warned that microstepping is used mainly to smooth
operation, and the steps divisions are not usually linear. So to use a
stepper motor, you really need to use something that cogs at each position.
I highly doubt you could really make a mechanical linkage and be able to
accurately eyeball 0.36 degree increments, but that's what you say you
need....

So, here's the plan: get your stepper chip and hook it up to the encoder.
Fine, you spin the encoder once, and the motor spins 10 times if it's a 3.6
degree motors, and 5 times if it's a 1.8 degree motor. Now, go to
http://www.sdp-si.com and buy some nice spur or worm gears in a 10:1 or 5:1
ratio. Their plastic gears are pretty good, but for maximum accuracy go for
a couple of their metal ones. Now you have a dial that makes 1000 steps in a
revolution. The plastic gears won't set you back more than a buck or two
each, the metal ones...a bit more.

I forgot about the microsteps being non-linear. On a 3 inch diameter
dial face 100 divisions gives a spacing of .090" or about 3/32. 200
marks, so there is one mark for every .0005" and the spacing is down
to about 3/64" which is close to 1mm. Dividing that space by 5 gives a
spacing of about .009". No marks, just interpolating. The average
human eye can easily see the difference between .007". I can see that
with my glasses on. It's surprising how much detail we really can see.
Makes me wonder how small the rods and cones are in our eyes.
ERS

 

[Garrett Mace]

I forgot about the microsteps being non-linear. On a 3 inch diameter
dial face 100 divisions gives a spacing of .090" or about 3/32. 200
marks, so there is one mark for every .0005" and the spacing is down
to about 3/64" which is close to 1mm. Dividing that space by 5 gives a
spacing of about .009". No marks, just interpolating. The average
human eye can easily see the difference between .007". I can see that
with my glasses on. It's surprising how much detail we really can see.
Makes me wonder how small the rods and cones are in our eyes.
ERS

Yes, but what did you think of the gearing idea?

You might also want to investigate how fast this stepper motor is going to
have to spin and react, there is definitely an upper limit.

 

[Eric R Snow]

Yes, but what did you think of the gearing idea?

You might also want to investigate how fast this stepper motor is going to
have to spin and react, there is definitely an upper limit.

That's why I wanted to use a servo. Steppers can lose steps if
commanded to step too fast. Gearing is fine but makes the stepper spin
that much faster. I built a small X-Y stage for engraving using
steppers and software from, well, I forget. Anyway, gearing down can
get really fine motion and was used on the X-Y stage. So your idea is
good but for the speed. Most the time it would work but really
cranking might cause lost steps.
ERS

 

[Garrett Mace]

That's why I wanted to use a servo. Steppers can lose steps if

commanded to step too fast. Gearing is fine but makes the stepper spin
that much faster. I built a small X-Y stage for engraving using
steppers and software from, well, I forget. Anyway, gearing down can
get really fine motion and was used on the X-Y stage. So your idea is
good but for the speed. Most the time it would work but really
cranking might cause lost steps.
ERS

You could use a microcontroller to capture the encoder pulses and keep a
running count of the encoder position verus the dial position, and if the
speed of the incoming pulses is faster than what you discover the stepper
motor can handle, the microcontroller will drive the motor only at the
stepper's maximum speed until it catches up to the encoder. Of course this
would only work with intermittent high speed moves, not continuous high
speed motion. Maybe if a high speed is detected, you could kick in a
divide-by-10 or so, and light an LED that lets you know the dial is
indicating ten revolutions? And then when the speed goes down again, the
needle can advance back to the true-angle representation.

Overall it's looking to be just a bit more complicated than a digital
readout....

 

[Eric R Snow]

You could use a microcontroller to capture the encoder pulses and keep a
running count of the encoder position verus the dial position, and if the
speed of the incoming pulses is faster than what you discover the stepper
motor can handle, the microcontroller will drive the motor only at the
stepper's maximum speed until it catches up to the encoder. Of course this
would only work with intermittent high speed moves, not continuous high
speed motion. Maybe if a high speed is detected, you could kick in a
divide-by-10 or so, and light an LED that lets you know the dial is
indicating ten revolutions? And then when the speed goes down again, the
needle can advance back to the true-angle representation.

Overall it's looking to be just a bit more complicated than a digital
readout....

Yeah.Oh well...

 

[Si Ballenger]

Greetings,
I can buy a chip that will convert the signal from an HP encoder,
rotary or linear, to step and direction which can be used with a
stepper driver chip to turn a stepper motor. What I want to do is use
the motor to spin a needle above a dial so the thing emulates a travel
dial indicator. But I need 1000 steps, not 100. I could use a
microstepping drive but these are expensive. I was wondering if an RC
servo with the pot removed could be used for this. And done cheaper.
Any ideas?
Thanks,
Eric R Snow

You might could use a servo controller like the mini ssc II and 4
servos. I use the servo controller for 180 deg revolution, but I
think it can be set for servo 90 deg rotation with 254 positions.
Stack the 4 servos on top of each other with each turning 90 deg
so the top servo horn will rotate a full 360 deg when each servo
is rotated 90 deg. This would give ~1016 position resolution in
one rotation. Probably not practical, but an interesting gizmo
experiment. I'm looking at using 2 servos set at 180 deg rotation
and stacked to pan my webcam 360 deg.

 

[Garrett Mace]

You might could use a servo controller like the mini ssc II and 4

servos. I use the servo controller for 180 deg revolution, but I
think it can be set for servo 90 deg rotation with 254 positions.
Stack the 4 servos on top of each other with each turning 90 deg
so the top servo horn will rotate a full 360 deg when each servo
is rotated 90 deg. This would give ~1016 position resolution in
one rotation. Probably not practical, but an interesting gizmo
experiment. I'm looking at using 2 servos set at 180 deg rotation
and stacked to pan my webcam 360 deg.

Why not use a 1:2 gear linkage in this case too; choose a more heavy-duty
servo. On my servo-driven webcam I find that 256 steps across (nearly) 180
degrees is more than fine-grained enough; usually, useful movements are in
the 5 to 15 steps range. Moving one step you can barely even see a
difference in the image. 256 steps across 360 degrees wouldn't be too bad.

 

[Si Ballenger]

Why not use a 1:2 gear linkage in this case too; choose a more heavy-duty
servo. On my servo-driven webcam I find that 256 steps across (nearly) 180
degrees is more than fine-grained enough; usually, useful movements are in
the 5 to 15 steps range. Moving one step you can barely even see a
difference in the image. 256 steps across 360 degrees wouldn't be too bad.

Gearing (or pullies made from plastic bottle/jar caps/tops) can
make the 360 from a single servo. The origional poster needs 1000
discrete positions in the 360 rotation. The alternative for
single servo operation would be to use the 2:1 gearing and
make/buy a servo control circuit/chip with 10 bit resolution to
get the 1000 steps.

 

[bj]

Greetings,
I can buy a chip that will convert the signal from an HP encoder,
rotary or linear, to step and direction which can be used with a
stepper driver chip to turn a stepper motor. What I want to do is use
the motor to spin a needle above a dial so the thing emulates a travel
dial indicator. But I need 1000 steps, not 100. I could use a
microstepping drive but these are expensive. I was wondering if an RC
servo with the pot removed could be used for this. And done cheaper.
Any ideas?
Thanks,
Eric R Snow

Hi
I'm sort of new to to this electronics thing but I have watched digital
displays moving so fast that you can't make sense of them when they
are changing rapidly.I know what you're getting at with the trusted
needle
that is visible all all the time. I don't know what you are trying to
measure
but most things are reduceable to a voltage and there is no shortage of
analogue voltmeters.
My other thought is to to use a three column LED display (using three
different colours ) your eyes would soon learn to interpret that
information
and I dont't think it would take that many diodes to get down to the .36
degrees in your original scenario
cheers BJ