This just for my curiosity, I'm not about to go CNC.

I assume that the lead screw is controlled independently of the spindle,
with its own motor, not through a spindle-linked gearbox. Or, is it
linked & the gear selection computer controlled?

If a separate motor, there is a huge range of speeds required: feed from
..0025 to .250 is 100:1 & that's multiplied by the spindle speed range,
say 200 to 2000 is 10:1. 1000:1 total. How do they do that?

If not a separate motor ... never mind.

Thanks,
Bob
--
Nota for President

Bob Engelhardt wrote:

This just for my curiosity, I'm not about to go CNC.

I assume that the lead screw is controlled independently of the spindle,
with its own motor, not through a spindle-linked gearbox. Or, is it
linked & the gear selection computer controlled?

If a separate motor, there is a huge range of speeds required: feed from
.0025 to .250 is 100:1 & that's multiplied by the spindle speed range,
say 200 to 2000 is 10:1. 1000:1 total. How do they do that?

If not a separate motor ... never mind.

Thanks,
Bob
--
Nota for President

All axes, including the spindle on most modern CNC lathes are precision
servos. This allows nearly anything to be done, including indexing the
spindle precisely and using powered tooling to do such things as cutting
flats, drilling cross holes, etc. Try looking at some of the "Swiss CNC"
porn on Youtube if you want to do some drooling.

No gearbox on any CNC I've seen.

Servo systems should be able to hold within a couple digits of encoder
resolution at no to low speeds.
Encoders with 10,000 pulses per rev are cheap.
The E3 is less than $60,
http://www.usdigital.com/products/en...al/rotary/kit/

Assuming a precision ball screw, of .250 lead, that's 25 millionths
per pulse, a well written control should be able to hold 75 millionths
at moderate rotational speeds.

Rapids are another story.

Bob Engelhardt wrote:
This just for my curiosity, I'm not about to go CNC.

I assume that the lead screw is controlled independently of the spindle,
with its own motor, not through a spindle-linked gearbox. Or, is it
linked & the gear selection computer controlled?

If a separate motor, there is a huge range of speeds required: feed from
.0025 to .250 is 100:1 & that's multiplied by the spindle speed range,
say 200 to 2000 is 10:1. 1000:1 total. How do they do that?
100:1 speed control is about the limit for most stepper systems,
but it no problem at all for a servo. I did some tests when I
built a servo drive for my Bridgeport, and the motion started to
break up into stick/slip moves at about 0.01 IPM. With a top
speed of about 100 IPM, that's a 10,000:1 range. I'm sure
expensive industrial servos can do even better. (In fact, with
computer and software upgrades, I would expect my servos do
better now, too.)

Jon

"Bob Engelhardt" wrote in message
. ..
This just for my curiosity, I'm not about to go CNC.

I assume that the lead screw is controlled independently of the spindle,
with its own motor, not through a spindle-linked gearbox. Or, is it
linked & the gear selection computer controlled?

If a separate motor, there is a huge range of speeds required: feed from
.0025 to .250 is 100:1 & that's multiplied by the spindle speed range, say
200 to 2000 is 10:1. 1000:1 total. How do they do that?

If not a separate motor ... never mind.

Thanks,
Bob
--
Nota for President

I'm in agreement with everything stated so far. I don't think since the
advent of Numerical Control, any manufacturer used a mechanical gear drive
linked to the spindle for feed.

Automatics are an entirely different beast, but you asked about CNC.

On Thu, 28 Aug 2008 21:04:02 -0500, "Pete C."
wrote:


Bob Engelhardt wrote:

This just for my curiosity, I'm not about to go CNC.

I assume that the lead screw is controlled independently of the spindle,
with its own motor, not through a spindle-linked gearbox. Or, is it
linked & the gear selection computer controlled?

If a separate motor, there is a huge range of speeds required: feed from
.0025 to .250 is 100:1 & that's multiplied by the spindle speed range,
say 200 to 2000 is 10:1. 1000:1 total. How do they do that?

If not a separate motor ... never mind.

Thanks,
Bob
--
Nota for President

All axes, including the spindle on most modern CNC lathes are precision
servos. This allows nearly anything to be done, including indexing the
spindle precisely and using powered tooling to do such things as cutting
flats, drilling cross holes, etc. Try looking at some of the "Swiss CNC"
porn on Youtube if you want to do some drooling.

http://www.omni-turn.com/Pages/Examp...niSamples.html

Most of these are done in a single machine run (cycle)

Gunner



"Confiscating wealth from those who have earned it, inherited it,
or got lucky is never going to help 'the poor.' Poverty isn't
caused by some people having more money than others, just as obesity
isn't caused by McDonald's serving super-sized orders of French fries
Poverty, like obesity, is caused by the life choices that dictate
results." - John Tucci,

On Thu, 28 Aug 2008 22:10:58 -0500, Jon Elson
wrote:

Bob Engelhardt wrote:
This just for my curiosity, I'm not about to go CNC.

I assume that the lead screw is controlled independently of the spindle,
with its own motor, not through a spindle-linked gearbox. Or, is it
linked & the gear selection computer controlled?

If a separate motor, there is a huge range of speeds required: feed from
.0025 to .250 is 100:1 & that's multiplied by the spindle speed range,
say 200 to 2000 is 10:1. 1000:1 total. How do they do that?
100:1 speed control is about the limit for most stepper systems,
but it no problem at all for a servo. I did some tests when I
built a servo drive for my Bridgeport, and the motion started to
break up into stick/slip moves at about 0.01 IPM. With a top
speed of about 100 IPM, that's a 10,000:1 range. I'm sure
expensive industrial servos can do even better. (In fact, with
computer and software upgrades, I would expect my servos do
better now, too.)

Jon

Many modern mills have 1800 IPM rapids or more.

And thats moving a table assembly that may weigh 800 lbs.

Gunner

"Confiscating wealth from those who have earned it, inherited it,
or got lucky is never going to help 'the poor.' Poverty isn't
caused by some people having more money than others, just as obesity
isn't caused by McDonald's serving super-sized orders of French fries
Poverty, like obesity, is caused by the life choices that dictate
results." - John Tucci,

Bob Engelhardt wrote:

This just for my curiosity, I'm not about to go CNC.

I assume that the lead screw is controlled independently of the spindle,
with its own motor, not through a spindle-linked gearbox. Or, is it
linked & the gear selection computer controlled?

If a separate motor, there is a huge range of speeds required: feed from
.0025 to .250 is 100:1 & that's multiplied by the spindle speed range,
say 200 to 2000 is 10:1. 1000:1 total. How do they do that?

If not a separate motor ... never mind.

Thanks,
Bob


The spindle is most times electronically linked to the feed screw
drives. This is needed for threading but not general turning. The
encoder on the spindle has three outputs, A B and Z. the Z is the
reference pulse and comes on once in every rotation. This pulse is used
to indicate the angular position of the spindle. When threading, this
pulse starts the feed axis in the same position for each cut of the
thread. The thread pitch is programmed into the control so that it feeds
as a function of the spindle rotation.


John

Jon Elson wrote:
... I did some tests when I built a servo
drive for my Bridgeport, and the motion started to break up into
stick/slip moves at about 0.01 IPM. With a top speed of about 100 IPM,
that's a 10,000:1 range. ...

Wow, that's impressive. I had been thinking about an independent drive
for my lathe feed screw & was about to give up because of the wide speed
range it needs, but this revives it.

Now, about servos: I understand that feedback is the essence of servos
and a controller is the heart, while various motors can be used. Unlike
a stepper, where the motor is critical. Specifically, could I use a
Baldor 2500 rpm, 1.65hp PMDC brush-type motor? It's from a treadmill,
but it is a Baldor, not some Chiwandian brand X. 1.65hp at 2500 rpm is
42 in-lbs of torque, if I did the numbers right. I'd need about that
for the lead screw.

Where I get stuck is the controller (& associated feedback). I looked
at the KB Electronics controllers, just as a place to start, & all their
dc controllers have a speed range of 50:1. How do I find a controller
for 1000:1 range?

And, can I really get full torque at very low speeds, say 2 1/2 rpm?

Thanks,
Bob
--
Nota for President

On 2008-09-02, Bob Engelhardt wrote:
Jon Elson wrote:
... I did some tests when I built a servo
drive for my Bridgeport, and the motion started to break up into
stick/slip moves at about 0.01 IPM. With a top speed of about 100 IPM,
that's a 10,000:1 range. ...

Wow, that's impressive. I had been thinking about an independent drive
for my lathe feed screw & was about to give up because of the wide speed
range it needs, but this revives it.

This is why servos are used for this type of work.

Now, about servos: I understand that feedback is the essence of servos
and a controller is the heart, while various motors can be used.

Four things are needed:

1) A servo motor with a tach generator (assuming a DC servo motor
instead of an AC one. The tach generator must be on the same
shaft as the motor for most stable results.

2) A servo amplifier which can combine a command voltage from
the controller computer and the tach generator feedback from
the motor.

3) An encoder (either a rotary encoder on the motor's shaft, on
the leadscrew, or a liner encoder on the tables (more for a mill
than a lathe, I think, but it depends.

4) A controller (the computer, with additional cards for ouputting
control voltages, command signals (e.g. start stop spindle
motor, select forward/reverse, lock unlock axes, etc. For a
machining center, this will probably include commands to
lock/open the chuck(s).


Unlike
a stepper, where the motor is critical. Specifically, could I use a
Baldor 2500 rpm, 1.65hp PMDC brush-type motor? It's from a treadmill,
but it is a Baldor, not some Chiwandian brand X. 1.65hp at 2500 rpm is
42 in-lbs of torque, if I did the numbers right. I'd need about that
for the lead screw.

The servo motors which I have used are rather specialized, and I
don't think that the treadmill motor will be sufficient. The rotor and
the tach generator are just cages of wire, wound to shape, connected to
the commutator, and rigidly attached to the shaft. The magnetic circuit
is a combination of a set of poles inside the cage (which do not move)
and either a permanent magnet field or a DC coil to drive the field.

The reason for this construction is to minimize the rotating
mass, so it can get from one speed to another (including stop) as
quickly as possible.

Where I get stuck is the controller (& associated feedback). I looked
at the KB Electronics controllers, just as a place to start, & all their
dc controllers have a speed range of 50:1.

Those are for drive motors, not servo motors. And the feedback
which *they* use comes from measuring the voltage generated by the
motor's drive winding during moments of no power.

a true servo motor will have a built-in tach generator -- either
in the same housing as the rest (in which case expect to find four brush
caps at 90 degree intervals instead of the usual two), or on a smaller
diameter extension of the motor's body which has its own pair of brushes
and the tach generator.

How do I find a controller
for 1000:1 range?

A servo amplifier. It is not called a "controller" for that
application.

And, can I really get full torque at very low speeds, say 2 1/2 rpm?

You can get full torque from a true servo motor and amplifier
at as little as 1/10 RPM. It depends on the tach generator to tell it
what speed is is currently producing, so the amplifier can provide it
with more or less current as needed to produce a stable speed.

Going to eBay and searching for servo amplifier, I find:

260279721259 (an AC servo amp by Fanuc -- with a price to
scare you away.

280260204020 (a DC servo motor of what looks like good quality.)
As an example of the tach -- this one produces 19 V per
1000 RPM.

350092690530 (if you want something to play with and to learn
from, it has both a mix of servo motors and "drives"(the
servo amplifiers). Looking at the photos, several of
the amplifiers (drives) look like the ones which I use.

260278680985 (these look like rather impressive amplifier specs,
especially given the buy-it-now price. But you would
have to add your own DC power supplies. The amplifiers
which I have have built-in power supplies.

Just some things to look at.

Good Luck,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

DoN. Nichols wrote:
Bob Engelhardt wrote:

Four things are needed:

1) A servo motor with a tach generator ...
2) A servo amplifier ...
3) An encoder ... on the leadscrew, ...
4) A controller (the computer, ...

OK, but my application is so simple (degenerate, really), that it seems
that I could do with less. I just want to control the leadscrew speed,
not the position of the carriage, and the speed would be set manually
(e.g., with a pot). Doesn't my requirements collapse your list to just
#1 & 2?

... Specifically, could I use a
Baldor 2500 rpm, 1.65hp PMDC brush-type motor? ...

The servo motors which I have used are rather specialized, and I
don't think that the treadmill motor will be sufficient. ...
The reason for this construction is to minimize the rotating
mass, so it can get from one speed to another (including stop) as
quickly as possible.

Again, my requirements would not include quick speed changes.

How do I find a controller for 1000:1 range?

A servo amplifier. It is not called a "controller" for that application.

Ah! Knowing the terminology makes a BIG difference. What are the
chances of my discovering that without RCM?


And, can I really get full torque at very low speeds, say 2 1/2 rpm?

You can get full torque from a true servo motor and amplifier
at as little as 1/10 RPM. It depends on the tach generator to tell it
what speed is is currently producing, so the amplifier can provide it
with more or less current as needed to produce a stable speed.

Now that brings up something else I've wondered about. I've heard
before how servo amps control the motor speed with voltage and motor
torque with current. Makes sense conceptually, but how does the amp
control the current independently of the voltage? My intuition is that
you set the voltage for the speed you want & as the motor is loaded
(more torque needed), it just draws more current to supply the torque to
keep the motor running at the voltage-selected speed. That does require
a low impedance source - is that the complication that brings in
separate control of the current?


Going to eBay and searching for servo amplifier, I find:
260279721259 (an AC servo amp by Fanuc -...

$400!

350092690530 (... a mix of servo motors and ... amplifiers ...

$600!

260278680985 (these look like rather impressive amplifier specs,
especially given the buy-it-now price. ...

$47 (shipped). That's better. I just want to drive a lead screw, for
crying out loud.

Good Luck,
DoN.

Thanks for tutorial! I feel like I've taken a significant step up the
learning curve.

Bob

--
Nota for President

On 2008-09-02, Bob Engelhardt wrote:
DoN. Nichols wrote:
Bob Engelhardt wrote:

Four things are needed:

1) A servo motor with a tach generator ...
2) A servo amplifier ...
3) An encoder ... on the leadscrew, ...
4) A controller (the computer, ...

OK, but my application is so simple (degenerate, really), that it seems
that I could do with less. I just want to control the leadscrew speed,
not the position of the carriage, and the speed would be set manually
(e.g., with a pot). Doesn't my requirements collapse your list to just
#1 & 2?

Yes -- as long as you did not need to accurately tie the
leadscrew speed to the spindle speed (such as for threading). If you
just want a nice finish, yes, the servo amplifier and the servo motor
with built-in tach would be sufficient.

... Specifically, could I use a
Baldor 2500 rpm, 1.65hp PMDC brush-type motor? ...

The servo motors which I have used are rather specialized, and I
don't think that the treadmill motor will be sufficient. ...
The reason for this construction is to minimize the rotating
mass, so it can get from one speed to another (including stop) as
quickly as possible.

Again, my requirements would not include quick speed changes.

O.K. You could probably get away with less -- but you still
need a good tach generator sharing the motor's shaft to get stable
speed control.

And are you planning to put ths on a ball screw? You get more
reliable speed with that than with half nuts. At really slow speeds you
will have stick-slip behavior with Acme threads.

How do I find a controller for 1000:1 range?

A servo amplifier. It is not called a "controller" for that application.

Ah! Knowing the terminology makes a BIG difference. What are the
chances of my discovering that without RCM?

You could probably find it eventually -- searching on "servo"
first, and reading a lot. :-)

Also -- in some of the eBay auctions they were called "drives".
The ones which I am most familiar with are capable of producing voltages
from -40V to +40V and -20A to +20A. The label says "Aerotech 4020 DC
Servo Controller", so "controller" is used at least sometimes to
describe them. My first pair of these came from a hamfest with a couple
of servo motors. Later I got others from eBay auctions at quite
reasonable prices.

If you wind up getting those, I have a scanned copy of the
manual (several Xeroxed pages, nothing bound) which I can let you have.

Ebay auction 130075041135 has a pair of them - but I think you
won't like the $575.00 buy-it-now price. :-)

Ebay auction 250180874877 has one of the motor with their label,
going for $275.00. It appears to be new. Note the two pairs of brush
caps -- one near the middle, and one near the bottom end.

The shaft at the bottom end would be used for mounting an encoder
if that was the way you want to go.

And, can I really get full torque at very low speeds, say 2 1/2 rpm?

You can get full torque from a true servo motor and amplifier
at as little as 1/10 RPM. It depends on the tach generator to tell it
what speed is is currently producing, so the amplifier can provide it
with more or less current as needed to produce a stable speed.

Now that brings up something else I've wondered about. I've heard
before how servo amps control the motor speed with voltage and motor
torque with current. Makes sense conceptually, but how does the amp
control the current independently of the voltage? My intuition is that
you set the voltage for the speed you want & as the motor is loaded
(more torque needed), it just draws more current to supply the torque to
keep the motor running at the voltage-selected speed. That does require
a low impedance source - is that the complication that brings in
separate control of the current?

Voltage will track current -- at a constant speed and load.
While feeding a constant voltage, if you increase the load the current
increases and the speed drops a little. While feeding a constant
current, the speed will vary widely with load.

So -- in comes the servo amplifier. You could look at it as
controlling the output voltage or the output current, but it does not
really matter. Here is what happens:

1) You feed a speed command voltage (between 0 and 10V DC
typically) to the amplifier.

2) The motor feeds a voltage generated by the tachometer generator
to the amplifier.

3) The amplifier compares the command voltage and the tach feedback
voltage, and if the tach voltage is lower than the command
voltage, it increases the voltage/current to the motor. If the
tach voltage is higher than the command voltage, the amplifier
reduces the voltage/current to the motor.

(Really -- there are adjustments to both inputs so you can
select what command voltage will give you the full speed from the motor
(typically 10V DC for full speed, but other options are possible -- and
negative command voltages cause the motor to run backwards.)

Now -- the motor brushes and the tach brushes have to be
oriented the same way when connecting to the servo amplifier. If you
reverse one or the other, you won't get stable speed at all. Instead it
will shoot to maximum speed forward or reverse. This is a good reason
to test the wiring *before* you bolt it to the leadscrew. :-)

Now -- reason for wanting the motor to be as low in inertia as
possible -- a change in load or command speed will cause the motor to
get a surge of voltage to catch it up -- and this can cause it to
overspeed, then drop below, then overspeed a bit less, eventually
reaching the desired speed.. There are adjustments, for damping so the
ramp-up and ramp-down are properly adjusted to the total inertia of the
whole load, but the less inertia in the motor itself, the easier it is
to stabilize the speed.


Going to eBay and searching for servo amplifier, I find:
260279721259 (an AC servo amp by Fanuc -...

$400!

Note also that this is an AC servo motor, so it requires a more
expensive AC servo motor -- not the DC ones which I have been talking
about.

350092690530 (... a mix of servo motors and ... amplifiers ...

$600!

But plenty to play with to set up for what you want to do plus
lots of other projects. It looks like about $50.00 of that lot is all
that you need for a single axis -- think of the rest as spares or
experimental items.

260278680985 (these look like rather impressive amplifier specs,
especially given the buy-it-now price. ...

$47 (shipped). That's better. I just want to drive a lead screw, for
crying out loud.

You ideally want to drive a ballscrew for the slow speed end of
things, and for the ability to use smaller servo motors.

That amplifier looks like a nice one -- but remember that you
will also have to make or buy a power supply to feed it. Up to 80V
depending on what your motor will accept, and up to 15 A continuous, and
30 A peak. The Aerotech which I mentioned above has the advantage of
built-in power supply, simplifying the setup.

Remember to avoid a "brushless" servo motor. It won't work with
this (or any DC servo amp).

Good Luck,
DoN.

Thanks for tutorial! I feel like I've taken a significant step up the
learning curve.

The best thing to do is pick up a small motor to experiment with
and you can use it with the servo amp which you have liked the price of
above.

Good Luck,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

On Tue, 02 Sep 2008 13:50:56 -0400, Bob Engelhardt
wrote:

Now that brings up something else I've wondered about. I've heard
before how servo amps control the motor speed with voltage and motor
torque with current. Makes sense conceptually, but how does the amp
control the current independently of the voltage? My intuition is that
you set the voltage for the speed you want & as the motor is loaded
(more torque needed), it just draws more current to supply the torque to
keep the motor running at the voltage-selected speed.

Generally so, but it depends on the application. If you want to
control speed and deliver whatever torque is needed, you control
voltage to get desired speed and let current be whatever it is. If
you want to control torque regardless of speed (from stall to max
speed) you control current to get desired torque and let voltage (and
speed) be whatever they turn out to be. You can also control
voltage/speed until current reaches a preset level (torque limit), at
which point currrent limit takes over and voltage /speed becomes
whatever results in the set torque.

That does require
a low impedance source - is that the complication that brings in
separate control of the current?

Variables a speed/voltage, current/torque and the load's
speed/torque relationship. You usually have no control over the
load's speed/torque relationship, it is what it is. So, you can
control voltage/speed or current/torque but not both at the same time
unless you can control the load's speed/torque relationship.

DoN. Nichols wrote:
On 2008-09-02, Bob Engelhardt wrote:
DoN. Nichols wrote:
Bob Engelhardt wrote:

... If you
just want a nice finish, yes, the servo amplifier and the servo motor
with built-in tach would be sufficient.

What I really want is to be able to adjust the feed with the turn of a
dial. Fast feed for roughing: too slow? - turn it up. Slow feed for
finish: still too rough? - turn it down some more. I have this with
VFD's on my drill press & lathe spindle and I've gotten spoiled by it.

... but you still
need a good tach generator sharing the motor's shaft to get stable
speed control.

Understood. It brings up yet another question: whether I use a tach or
optical sensor. Is it dictated by the amp? Or do amps generally accept
either?

And are you planning to put ths on a ball screw? You get more
reliable speed with that than with half nuts. At really slow speeds you
will have stick-slip behavior with Acme threads.

I wasn't planning on a ball screw. Hadn't even thought about it. If
that's what it takes, I'll probably dump the idea: that would be too
much work & expense for the convenience of dial-it-in feed rate.

There are separate turning & threading lead screws. Well, the turning
feed uses a shaft with a key slot in it - not really a "lead screw". I
guess that avoids the stick-slip problem (?).


snip a bunch of good stuff

Thanks again - I'm off to search eBay on "servo (amp*,drive,controller)"

Bob
--
Nota for President

Don Foreman wrote:
good stuff & bottom line was:
So, you can
control voltage/speed or current/torque but not both at the same time
...

Good! I just couldn't get my head around how both could be controlled
at the same time.

Thanks,
Bob

--
Nota for President

On Tue, 02 Sep 2008 23:54:02 -0400, Bob Engelhardt
wrote:

DoN. Nichols wrote:
On 2008-09-02, Bob Engelhardt wrote:
DoN. Nichols wrote:
Bob Engelhardt wrote:

... If you
just want a nice finish, yes, the servo amplifier and the servo motor
with built-in tach would be sufficient.

What I really want is to be able to adjust the feed with the turn of a
dial. Fast feed for roughing: too slow? - turn it up. Slow feed for
finish: still too rough? - turn it down some more. I have this with
VFD's on my drill press & lathe spindle and I've gotten spoiled by it.


Ive got it on the feed motor of my Hardinge HLV-H

Its soon to have a twin....$1500

Giggle!!!!!!


Gunner

"I am for doing good to the poor, but I differ in opinion of the
means. I think the best way of doing good to the poor, is not
making them easy in poverty, but leading or driving them out of
it. In my youth I travelled much, and I observed in different
countries, that the more public provisions were made for the
poor the less they provided for themselves, and of course became
poorer. And, on the contrary, the less was done for them, the
more they did for themselves, and became richer." -- Benjamin
Franklin, /The Encouragement of Idleness/, 1766

Bob Engelhardt wrote:
...
... the turning
feed uses a shaft with a key slot in it - not really a "lead screw". I
guess that avoids the stick-slip problem (?).
....

Oh, no - wait. There's a sliding gear on the shaft that's driven by a
key in that slot & that gear & key slide along the shaft as the carriage
moves. I guess there could be a lot of stick-slip!! Ahhh ... damn - I
really wanted to be able to slow feed. Well, I'll see what happens.

I bought a servo amp on eBay last night - $17.50. "Too good to be true"
comes to mind, but I found some data for it & I think it'll work. eBay
180282383302 & data:
http://www.a-m-c.com/download/datasheet/25a20i.pdf

Bob

--
Nota for President

On Wed, 03 Sep 2008 09:27:22 -0400, Bob Engelhardt
wrote:

Bob Engelhardt wrote:
...
... the turning
feed uses a shaft with a key slot in it - not really a "lead screw". I
guess that avoids the stick-slip problem (?).
...

Oh, no - wait. There's a sliding gear on the shaft that's driven by a
key in that slot & that gear & key slide along the shaft as the carriage
moves. I guess there could be a lot of stick-slip!! Ahhh ... damn - I
really wanted to be able to slow feed. Well, I'll see what happens.

I bought a servo amp on eBay last night - $17.50. "Too good to be true"
comes to mind, but I found some data for it & I think it'll work. eBay
180282383302 & data:
http://www.a-m-c.com/download/datasheet/25a20i.pdf

Bob
If it works it is a steal! Do not bu**er it up!
Now find a servo motor and build a PSU.
Only problem you may have is Analogue input not Step and direction,
but it depends how you think you will control it.

--

Richard

Email address is valid but remove burrs before sending!

On 2008-09-03, Bob Engelhardt wrote:
DoN. Nichols wrote:
On 2008-09-02, Bob Engelhardt wrote:
DoN. Nichols wrote:
Bob Engelhardt wrote:

... If you
just want a nice finish, yes, the servo amplifier and the servo motor
with built-in tach would be sufficient.

What I really want is to be able to adjust the feed with the turn of a
dial. Fast feed for roughing: too slow? - turn it up. Slow feed for
finish: still too rough? - turn it down some more. I have this with
VFD's on my drill press & lathe spindle and I've gotten spoiled by it.

O.K. No threading on this one, so that should work.

I would suggest either a logarithmic taper pot (so you can make
very fine adjustments at slow speeds and coarser but quicker ones at
high speeds. The alternative would be a 10-turn linear pot which should
give you sufficient fine control at the low end -- but you would need
lots of turns at the high end to make a noticeable change in speed.

... but you still
need a good tach generator sharing the motor's shaft to get stable
speed control.

Understood. It brings up yet another question: whether I use a tach or
optical sensor. Is it dictated by the amp? Or do amps generally accept
either?

A tach generator is what your really want for very slow speeds.

You could use an encoder with a circuit to produce a voltage
proportional to the pulse rate, but that would probably make your speed
bumpy at the low end.

And are you planning to put this on a ball screw? You get more
reliable speed with that than with half nuts. At really slow speeds you
will have stick-slip behavior with Acme threads.

I wasn't planning on a ball screw. Hadn't even thought about it. If
that's what it takes, I'll probably dump the idea: that would be too
much work & expense for the convenience of dial-it-in feed rate.

There are separate turning & threading lead screws. Well, the turning
feed uses a shaft with a key slot in it - not really a "lead screw". I
guess that avoids the stick-slip problem (?).

That should do. You've also got extra gearing in the apron to
give you a torque advantage. I was assuming that you were also planning
to thread with this as well as turn.

snip a bunch of good stuff

Thanks again - I'm off to search eBay on "servo (amp*,drive,controller)"

You'll get lots of audio amplifiers -- and automobile mounted
stereo systems with that. Try this instead:

+servo +(amp*,drive,controller)

The '+' means that the entry has to be present.

Your original one gave 3,546 hits.

My first mod only took it down to 3,545 hits. Hardly worth the
effort. :-)

Hmm ... add "+DC" to the list so you don't come up with
AC/Brushless servos and amplifiers as well.

That takes it down to 461 hits.

Add "-car" and we get 424 hits

Add ' -"hard drive" ' and it is 365 hits.

Add ' -turntable -"tape drive" ' and it is 321 hits

This seems to be mostly reasonable things -- some model airplane
stuff, and some audio stuff still sneaking through. :-)

Be careful to avoid the ones marked:

"New SERVO 1150 RPM 38 VDC AMETEK motor, wind generator"

because they have no tach generator.

Hmm ... add ' -"lost plane finder" '


This one is a nice Baldor DC servo motor -- 260279397093 -- but
out of your proposed budget I think.

Good Luck,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

On 2008-09-03, Bob Engelhardt wrote:
Bob Engelhardt wrote:
...
... the turning
feed uses a shaft with a key slot in it - not really a "lead screw". I
guess that avoids the stick-slip problem (?).
...

Oh, no - wait. There's a sliding gear on the shaft that's driven by a
key in that slot & that gear & key slide along the shaft as the carriage
moves. I guess there could be a lot of stick-slip!!

But you should have a well-lubricated worm gear in there which
will help a lot.

Ahhh ... damn - I
really wanted to be able to slow feed. Well, I'll see what happens.

I bought a servo amp on eBay last night - $17.50. "Too good to be true"
comes to mind, but I found some data for it & I think it'll work. eBay
180282383302 & data:
http://www.a-m-c.com/download/datasheet/25a20i.pdf

That looks like a nice one -- especially for the price. It will
handle motors up to 180 V and up to 12.5 A constant load, or 25 A peak.

I also see that it has adjustable current limits. On DC servo
motors you will proably find a "maximum current" spec. Don't exceed
this, as it can partially demagnetize the permanent magnet field
current.

You will still have to build or find an appropriate power supply
for the voltage which your servo motor wants.

Good Luck,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

On 2008-09-03, Richard Edwards wrote:
On Wed, 03 Sep 2008 09:27:22 -0400, Bob Engelhardt
wrote:

Bob Engelhardt wrote:

[ ... ]

I bought a servo amp on eBay last night - $17.50. "Too good to be true"
comes to mind, but I found some data for it & I think it'll work. eBay
180282383302 & data:
http://www.a-m-c.com/download/datasheet/25a20i.pdf

Bob
If it works it is a steal! Do not bu**er it up!
Now find a servo motor and build a PSU.
Only problem you may have is Analogue input not Step and direction,
but it depends how you think you will control it.

Since he wants to command it from a pot for a variable speed
feed, this is just what he needs.

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

DoN. Nichols wrote:
On 2008-09-03, Bob Engelhardt wrote:
... a sliding gear on the shaft ... I guess there could be a lot of stick-slip!!

But you should have a well-lubricated worm gear in there which
will help a lot.

You give me hope - thanks.


I bought a servo amp on eBay last night -...

I also see that it has adjustable current limits. On DC servo
motors you will proably find a "maximum current" spec. Don't exceed
this, as it can partially demagnetize the permanent magnet field
current.

I suppose that's a real problem with servos - where the amp can really
goose the motor when it's falling behind.

You will still have to build or find an appropriate power supply
for the voltage which your servo motor wants.

Maybe yet another app for the venerable MOT G. I'd only need
buck/boost on the ac line, a bridge, & cap. Yes?

Thanks,
Bob
--
Nota for President

DoN. Nichols wrote:
....
I would suggest either a logarithmic taper pot (so you can make
very fine adjustments at slow speeds and coarser but quicker ones at
high speeds. ...

Ah ... good point!

A tach generator is what your really want for very slow speeds.

You could use an encoder with a circuit to produce a voltage
proportional to the pulse rate, but that would probably make your speed
bumpy at the low end.

OK - definitely don't want bumpy at the low end (finish cuts).

Thanks,
Bob
--
Nota for President

On 2008-09-04, Bob Engelhardt wrote:
DoN. Nichols wrote:
On 2008-09-03, Bob Engelhardt wrote:

[ ... ]

I bought a servo amp on eBay last night -...

I also see that it has adjustable current limits. On DC servo
motors you will proably find a "maximum current" spec. Don't exceed
this, as it can partially demagnetize the permanent magnet field
current.

I suppose that's a real problem with servos - where the amp can really
goose the motor when it's falling behind.

Yes -- unless it is one of the exotic magnets such as are found
in modern disk drives. The problem is that servo motors on eBay are
more likely to be ferrite magnets than anything else -- and those can be
demagnetized with excess current.

So -- set the current to minimum current, clamp the shaft of the
servo motor, turn the command voltage up until you get whatever that
current limit is (while measuring the current through the motor), and
slowly turn the limit up until you reach say 90% of the current limit on
the motor's label. This should be safe enough under all conditions.

You may find yourself needing to play with the damping
adjustments as well -- if you find the motor speed jerking above and
below the set speed.

The best way to adjust the damping (with the full load on the
motor) is to feed it a square wave as a command voltage, a high enough
frequency so the carriage only goes an inch or less, and observe the
output from the tach with an oscilloscope (if you don't have one, check
around for a friend with one). Adjust the damping so the leading edge
of the square wave has just a little visible rounding at the top. If it
has a very slow ramp up to speed (speed is what the tach output
measures), it is over-damped and will trail behind your desired
position. If the speed overshoots the set speed, then comes down below
it, then back above it somewhat less -- it is under-damped and you will
want to increase the damping.

O.K. There are two levels of damping in it -- a lower one when
S9 is open, and a higher one when S9 is closed. There are also
provisions for more precise tuning by adding a capacitor (C72) through a
hole in the case of the amplifier. There is apparently a manufacturer's
web page on tuning according to the PDF you (and I) downloaded. The
_Inhibit_In_ is used to shut down the amplifier under bad conditions,
such as when the carriage hits a limit switch at either end of travel.
It should never get there, but if it does, you want to shut the carriage
down *quickly*. I would suggest that you put the tailstock end limit
switch on the tailstock, since that can move from time to time, and a
stationary limit switch will not always be in the right place.

Add a relay and a panic button to shut it down even before it
hits a limit switch (such as when you discover that you are about to try
to turn the jaws of the chuck. :-)

You will still have to build or find an appropriate power supply
for the voltage which your servo motor wants.

Maybe yet another app for the venerable MOT G. I'd only need
buck/boost on the ac line, a bridge, & cap. Yes?

Hmm ... ideally, you want a transformer which isolates between
the primary and secondary since there is no isolation in the servo
amplifier. You can probably get away without this, but it is safer if
you have it -- and it will be mandatory if you eventually wind up
connecting it to a computer.

As for what voltage to make it for -- get your servo motor first
and then you will know. :-) If you don't need the full speed of the
servo motor -- go for perhaps 20% under the maximum voltage of the servo
amplifier or the motor -- whichever is lower -- since you seem to be
designing without a regulator. With a regulator, you can work closer to
the limits of the motor and/or amplifier.

Enjoy,
DoN.

--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

Bob Engelhardt wrote:
Jon Elson wrote:

... I did some tests when I built a servo
drive for my Bridgeport, and the motion started to break up into
stick/slip moves at about 0.01 IPM. With a top speed of about 100
IPM, that's a 10,000:1 range. ...


Wow, that's impressive. I had been thinking about an independent drive
for my lathe feed screw & was about to give up because of the wide speed
range it needs, but this revives it.

Now, about servos: I understand that feedback is the essence of servos
and a controller is the heart, while various motors can be used. Unlike
a stepper, where the motor is critical. Specifically, could I use a
Baldor 2500 rpm, 1.65hp PMDC brush-type motor? It's from a treadmill,
but it is a Baldor, not some Chiwandian brand X. 1.65hp at 2500 rpm is
42 in-lbs of torque, if I did the numbers right. I'd need about that
for the lead screw.

You need 1.65 Hp for a leadscrew? Are you taking 1" depth of
cut at 50 IPM? I seriously doubt you need ANYTHING like this.
Check out the TINY Baldor motor used on a Hardinge, for
instance. It is probably 1/30th Hp. (Not used for rapid feeds,
however, you just disengage the half nuts for that.) In almost
all cases it is better to use a smaller motor with a simple belt
reduction that requiring a massive motor. The one place the
large, direct-drive motor is better is where rapid acceleration
is needed, and that sounds like a CNC-only case.
Where I get stuck is the controller (& associated feedback). I looked
at the KB Electronics controllers, just as a place to start, & all their
dc controllers have a speed range of 50:1. How do I find a controller
for 1000:1 range?

You really can't do 1000:1 with manual controls. How will you
set the speed? Even one of those "knob pots" with the digital
dial readout is really only good for manual setting of 100:1 or
so.
And, can I really get full torque at very low speeds, say 2 1/2 rpm?
A true servo, with a positioning loop, can get full torque at
ZERO RPM. Apply full power to the motor at stall, and you get
full rated torque.

But, I "think" you need to rethink this whole thing. Do you
really need to set speed to 0.1% of the setting? How will you
ever know? If the max speed is 100 IPM, then the minimum speed,
or setting resolution will be 0.1 IPM. Do you ever make cuts on
your lathe at 0.1 IPM? Would you know the difference between 10
IPM and 10.1 IPM?

The only way this makes any sense is in a real CNC application,
where the encoder gives true position feedback to the computer,
and the computer commands movement to put the tool where it
needs to go.

Jon

DoN. Nichols wrote:

The servo motors which I have used are rather specialized, and I
don't think that the treadmill motor will be sufficient. The rotor and
the tach generator are just cages of wire, wound to shape, connected to
the commutator, and rigidly attached to the shaft. The magnetic circuit
is a combination of a set of poles inside the cage (which do not move)
and either a permanent magnet field or a DC coil to drive the field.

The reason for this construction is to minimize the rotating
mass, so it can get from one speed to another (including stop) as
quickly as possible.

This is an ironless rotor motor, a very expensive sort used
where extreme acceleration is needed, such as computer tape
drive capstans, which have to go from zero to several thousand
RPM in a millisecond or so. I have an ironless rotor motor with
ceramic shaft on my Bridgeport Z axis, entirely because I pulled
it out of a scrap tape drive. You most certainly do not need
such a motor on a machine tool. A motor with a sufficient
number of windings (slots & commutator segments) will have low
enought torque and velocity ripple to work fine in most cases.
The treadmill motors may be a tiny bit marginal in this case,
but likely will work fine.

Jon

Bob Engelhardt wrote:

Now that brings up something else I've wondered about. I've heard
before how servo amps control the motor speed with voltage and motor
torque with current. Makes sense conceptually, but how does the amp
control the current independently of the voltage? My intuition is that
you set the voltage for the speed you want & as the motor is loaded
(more torque needed), it just draws more current to supply the torque to
keep the motor running at the voltage-selected speed.
Yes, if there was no resistance in the armature and no "armature
reaction" (where the fields in the motor fight against each
other) then the motor speed would remain constant. But, the
resistance is quite real, and the motor will slow down under
load. A shunt-wound or PM motor will slow down the least, but
it will still slow down. DC motor controllers have an
adjustable compensation for this, but it is an approximation.
That does require
a low impedance source - is that the complication that brings in
separate control of the current?
The compensation I mention above acts to decrease source
impedance into the negative region to compensate for the motor's
positive resistance. But, it is an after-the-fact correction.
To hold precise speed, you need to MEASURE the speed, with a
tachometer, or measure the position and calculate movement over
time to perfectly correct the velocity.

Jon

Bob Engelhardt wrote:
What I really want is to be able to adjust the feed with the turn of a
dial. Fast feed for roughing: too slow? - turn it up. Slow feed for
finish: still too rough? - turn it down some more. I have this with
VFD's on my drill press & lathe spindle and I've gotten spoiled by it.

And there's no WAY you need 1000:1 control over this. The 100:1
of a standard DC motor controller with resistance compensation
will be FINE! They actually are better than that, but with a
10-turn pot, you'll never be able to set it any finer, so who
would ever know?
... but you still
need a good tach generator sharing the motor's shaft to get stable
speed control.

Some DC controllers have a tach input, that would perform much
better than the resistance compensation.

Jon

Bob Engelhardt wrote:
Bob Engelhardt wrote:
...

... the turning feed uses a shaft with a key slot in it - not really a
"lead screw". I guess that avoids the stick-slip problem (?).

...

Oh, no - wait. There's a sliding gear on the shaft that's driven by a
key in that slot & that gear & key slide along the shaft as the carriage
moves. I guess there could be a lot of stick-slip!! Ahhh ... damn - I
really wanted to be able to slow feed. Well, I'll see what happens.

I bought a servo amp on eBay last night - $17.50. "Too good to be true"
comes to mind, but I found some data for it & I think it'll work. eBay
180282383302 & data:
http://www.a-m-c.com/download/datasheet/25a20i.pdf
I think this is the same as a Copley 412, their setup and
install data may be more understandable and complete.

You will need a DC tach. The motor can have some velocity
ripple, the tach MUST NOT! So, you do need to use an ironless
rotor tach. The best deal is to find a Maxon motor (These are
tiny, about 1.25" diameter) with silver brushes. They sell the
same design as a tach. A number of surplus dealers have them,
or look on eBay. You will be able to hook a 10-turn pot to
either the -10 or +10 V supply with a switch (for direction) and
signal ground, and the wiper of the pot to +ref in. ground -ref
in. Set it up for tach mode as shown in the chart. I think you
want to ground both +inhibit and -inhibit, and put a run-stop
switch on inhibit that grounds it to run. (You could also set
up the direction switch to be center off for the stop setting.)

Jon

DoN. Nichols wrote:
[about setting current limits & damping parameters]

Great - I've copied this for when I can actually do it.

Add a relay and a panic button to shut it down even before it
hits a limit switch (such as when you discover that you are about to try
to turn the jaws of the chuck. :-)

That's a good idea.

Hmm ... ideally, you want a transformer which isolates between
the primary and secondary since there is no isolation in the servo
amplifier. You can probably get away without this, but it is safer if
you have it -- ...

Yeah, I know. I was in avoidance over a 12.5A, 130v (say) transformer.
Particularly, buying one - I just looked at Surplus Sales: YIKES. I
guess that it's MOT time again.

Thanks,
Bob

--
Nota for President

Jon Elson wrote:
.... The
treadmill motors may be a tiny bit marginal in this case, but likely
will work fine.

Good to hear! It may be marginal, just so it's not "no way!"

Thanks,
Bob
--
Nota for President

Jon Elson wrote:
You need 1.65 Hp for a leadscrew? Are you taking 1" depth of cut at 50
IPM? I seriously doubt you need ANYTHING like this. ...

It's for the torque. I did a rough measure of lead screw torque needed
at 15 in-lb. I posted that & several repliers confirmed that I was in
the ballpark. See the thread at http://tinyurl.com/5jnmst
This motor is 42 in-lbs and, very importantly, it's one that I have.

You really can't do 1000:1 with manual controls. How will you set the
speed? Even one of those "knob pots" with the digital dial readout is
really only good for manual setting of 100:1 or so.
snip
But, I "think" you need to rethink this whole thing. Do you really need
to set speed to 0.1% of the setting? How will you ever know? If the
max speed is 100 IPM, then the minimum speed, or setting resolution will
be 0.1 IPM. Do you ever make cuts on your lathe at 0.1 IPM? Would you
know the difference between 10 IPM and 10.1 IPM?
snip

What I meant by 1000:1 was the range of speed, not the resolution.
E.g., 2 rpm to 2000 rpm. Picking a speed will be by what looks good,
not by a particular rpm. I figured the range like this: the minimum
will be for a finish cut (.003) with lowest spindle speed (30 rpm) (i.e.
large cast iron work). That figures to 2 1/4 rpm. The max lead screw
rpm will be a rough cut (.050) at high spindle speed (2200) (e.g., small
diam aluminum). That speed is 2750 rpm. A range of 2.25:2750 = 1:1222.

Wait - I just did the ipm for the fastest speed & it's 1.8"/sec! That's
probably too fast to be practical. I'll need to do some trials to get a
better number for that. So, my required speed range could be 1:600,
1:300, ... ???

Thanks,
Bob


--
Nota for President

Jon Elson wrote:
Bob Engelhardt wrote:
I bought a servo amp on eBay ...

I think this is the same as a Copley 412, their setup and install data
may be more understandable and complete.

Thanks - even if it's not the same amp, the Copley instructions could
help my basic understanding.

You will need a DC tach. The motor can have some velocity ripple, the
tach MUST NOT! So, you do need to use an ironless rotor tach. The best
deal is to find a Maxon motor (These are tiny, about 1.25" diameter)
with silver brushes. They sell the same design as a tach. A number of
surplus dealers have them, or look on eBay.

Ah - good to know! Is any Maxon motor OK, or is there something in
particular to look for? I suppose if it says "ironless rotor", I'm set.
But, if not ...?

You will be able to hook a
10-turn pot to either the -10 or +10 V supply with a switch (for
direction) and signal ground, and the wiper of the pot to +ref in.
ground -ref in. Set it up for tach mode as shown in the chart. I think
you want to ground both +inhibit and -inhibit, and put a run-stop switch
on inhibit that grounds it to run. (You could also set up the direction
switch to be center off for the stop setting.)

Damn, it's going to be fun playing with this. 'Just have to wait out
UPS delivery from CA (to Mass.)

Thanks,
Bob

--
Nota for President

On Thu, 04 Sep 2008 22:07:21 -0400, Bob Engelhardt
wrote:

Jon Elson wrote:

But, I "think" you need to rethink this whole thing. Do you really need
to set speed to 0.1% of the setting? How will you ever know? If the
max speed is 100 IPM, then the minimum speed, or setting resolution will
be 0.1 IPM. Do you ever make cuts on your lathe at 0.1 IPM? Would you
know the difference between 10 IPM and 10.1 IPM?
snip

What I meant by 1000:1 was the range of speed, not the resolution.
E.g., 2 rpm to 2000 rpm. Picking a speed will be by what looks good,
not by a particular rpm. I figured the range like this: the minimum
will be for a finish cut (.003) with lowest spindle speed (30 rpm) (i.e.
large cast iron work). That figures to 2 1/4 rpm. The max lead screw
rpm will be a rough cut (.050) at high spindle speed (2200) (e.g., small
diam aluminum). That speed is 2750 rpm. A range of 2.25:2750 = 1:1222.

Wait - I just did the ipm for the fastest speed & it's 1.8"/sec! That's
probably too fast to be practical. I'll need to do some trials to get a
better number for that. So, my required speed range could be 1:600,
1:300, ... ???

Though I don't think it's a reason to stop experimenting with the
servo, I'm with Jon here in thinking you're way overestimating the
required speed range. I just checked my lathe, which is a Feeler clone
of a Hardinge HLV-H. The usable feed range is about .15 IPM to 4 IPM,
less than 30:1, and the spindle runs from 15 to 4000 RPM. (Hardinge
specs 5/16 to 7 IPM; 22:1) I've never found the available range to be
a problem. The motor is a 75W PM.

Re the servo, I've used many Copley and AMC amps, and while they have
very similar lines, I've never run into identical products. I do
prefer Copley, mostly because their literature and support is better.
Most any PM brush DC motor should work for you as long as you can fit
the proper feedback device to the motor. Tight speed control without a
motion controller in the system is the only situation I can think of
where tach feedback is appropriate these days, so I don't think you'll
find many motor/tach combos on the surplus market.

--
Ned Simmons

Ned Simmons wrote:
... I'm with Jon here in thinking you're way overestimating the
required speed range. I just checked my lathe ...
The usable feed range is about .15 IPM to 4 IPM,
less than 30:1, ...

Thanks - actual data! My numbers were strictly based on personal
assumptions & my lathe (e.g., its finest feed is .003).

As to your lathe, let's see: say you are finish turning a 10" cast iron
chuck backing plate, 75 fpm gives spindle speed of 75*12/pi*10 = 29 rpm.
Then .15 ipm / 29 rpm = .005 feed (in/rev). About what I assumed.
Then say a 1/2" diam AL rod roughing at 400 fpm, spindle speed of 3000
rpm. 4 ipm / 3000 rpm = .001+ feed. Not what I'd call a roughing feed.
In fact, 4 ipm seems slow for roughing (a minute to rough 4").

I'm definitely keeping this question open: 1000:1 is out, as is 500:1,
based upon max practical carriage speed. But I think that I'd like more
than 30:1.

I've never found the available range to be a problem.

Personal experience is way better than theory! And I'll keep it in mind.
I'll also keep in mind that if it's trivially easier to have a greater
range, then I'll do it. For example, the amp that I bought can handle
12 1/2A at 190v - that's a lot of power & a given (not something that's
going to cost more).

The motor is a 75W PM.

That is a definite advantage.

...
Most any PM brush DC motor should work for you as long as you can fit
the proper feedback device to the motor. ...

Good!

I don't think you'll find many motor/tach combos on the surplus market.

That was my assumption. I have a small collection of DCPM motors & my
plan was to get a tach to use on them, as experiments.


Thanks for the good info,
Bob


--
Nota for President