$30, free shipping. 15a, 20 - 80v with TACH feedback (not encoder).
Not for CNC, but good for running a spindle with excellent low speed
torque. I just bought one for my mill.

The auction has ended, but the seller had 24 left.

http://tinyurl.com/m898jg auction # 170332489842

Bob

Wild_Bill wrote:
FWIW, chopping everything off after the eBay item number will provide a
shorter link that generally always works (I've never seen it fail).
....

Thanks, that's easy enough to do. It is strange that tinyurl didn't work.

Bob

On Sun, 14 Jun 2009 15:28:20 -0400, Bob Engelhardt
wrote:

$30, free shipping. 15a, 20 - 80v with TACH feedback (not encoder).
Not for CNC, but good for running a spindle with excellent low speed
torque.

It'll work just fine for most any position or velocity control app,
including CNC. The encoder feedback returns to the controller. The
tach feedback is optional, and generally not necessary with a modern
controller and high line count encoders.

If there's no motion controller in the system and the goal is tight
speed regulation, then the tach feedback is useful.

--
Ned Simmons

Bob Engelhardt wrote:
$30, free shipping. 15a, 20 - 80v with TACH feedback (not encoder). Not
for CNC, but good for running a spindle with excellent low speed
torque. I just bought one for my mill.

The auction has ended, but the seller had 24 left.

http://tinyurl.com/m898jg auction # 170332489842
Of course these can be used for CNC. You just
need a servo interface that puts
out +/- 10 V analog velocity commands. Pico
Systems (my co.) and Mesa
make interfaces for this, as well as Galil (gulp -
expensive).

Jon

"Bob Engelhardt" wrote in message
...
$30, free shipping. 15a, 20 - 80v with TACH feedback (not encoder). Not
for CNC, but good for running a spindle with excellent low speed torque.
I just bought one for my mill.

The auction has ended, but the seller had 24 left.

http://tinyurl.com/m898jg auction # 170332489842

Bob

I have an older Anilam CNC Bridgeport and Anilam CNC lathe. They both use
motors with tach feedback, the mill uses DRO scales for position feedback to
control and the lathe uses encoders coupled to the ballscrews. I replaced
the original amplifiers with AMC amplifiers similar to those in your link,
they work great.

My Anilam control didn't work on the lathe so I converted it to EMC2. I set
up the control to try to give a velocity feedforward and a proportional
signal, with a little bit of intergral to get rid of small offsets.

So, if someone has motors with tach feedback and would like to do a homebrew
CNC, use these amps with an external encoder going to the control and they
are great for CNC. I run EMC2 (free) and Linux (free) on a $35 eBay
computer ($70 shipped), a Mesa board ($200), a Mesa PWM to analog board
($69), a Mesa I/O Board ($69). Bottom line is I converted to EMC2 for
probably less than it would cost to fix the old control plus I gained
spindle speed control, more I/O capability, more program memory and storage,
and I can browse the internet while on the lathe. :-)

RogerN

Ned Simmons wrote:
It'll work just fine for most any position or velocity control app,
including CNC. The encoder feedback returns to the controller. The
tach feedback is optional, and generally not necessary with a modern
controller and high line count encoders.

If there's no motion controller in the system and the goal is tight
speed regulation, then the tach feedback is useful.


By "controller", do you mean something other than this servo amp?
'Cause the amp doesn't have inputs for encoder feedback, just tach.

Bob

Jon Elson wrote:
Of course these can be used for CNC. You just need a servo interface
that puts
out +/- 10 V analog velocity commands. ...

Then this amp would just be a driver, and its tach feedback/servo
circuitry wouldn't be used?

Bob

RogerN wrote:
I have an older Anilam CNC Bridgeport and Anilam CNC lathe. They both use
motors with tach feedback, the mill uses DRO scales for position feedback to
control and the lathe uses encoders coupled to the ballscrews. I replaced
the original amplifiers with AMC amplifiers similar to those in your link,
they work great.

How does you system use both tach feedback & encoders? It seems like
the encoders would give all the feedback that's needed.

... and I can browse the internet while on the lathe. :-)

A definite advantage G.

Bob

On Mon, 15 Jun 2009 14:10:32 -0400, Bob Engelhardt
wrote:

Ned Simmons wrote:
It'll work just fine for most any position or velocity control app,
including CNC. The encoder feedback returns to the controller. The
tach feedback is optional, and generally not necessary with a modern
controller and high line count encoders.

If there's no motion controller in the system and the goal is tight
speed regulation, then the tach feedback is useful.


By "controller", do you mean something other than this servo amp?
'Cause the amp doesn't have inputs for encoder feedback, just tach.


Yes, that amp by itself is essentially a high performance DC drive. By
motion controller (and what I gather Jon is calling an "interface") I
mean a general purpose programmable device that can monitor the
position and velocities of the various parts of a system and generate
the command voltage (the input to the amps) to coordinate a motion
routine.

--
Ned Simmons

Ned Simmons wrote:
Yes, that amp by itself is essentially a high performance DC drive. By
motion controller (and what I gather Jon is calling an "interface") I
mean a general purpose programmable device that can monitor the
position and velocities of the various parts of a system and generate
the command voltage (the input to the amps) to coordinate a motion
routine.

Thanks - still learning stuff. Bob

"Bob Engelhardt" wrote in message
...
RogerN wrote:
I have an older Anilam CNC Bridgeport and Anilam CNC lathe. They both
use motors with tach feedback, the mill uses DRO scales for position
feedback to control and the lathe uses encoders coupled to the
ballscrews. I replaced the original amplifiers with AMC amplifiers
similar to those in your link, they work great.

How does you system use both tach feedback & encoders? It seems like the
encoders would give all the feedback that's needed.

... and I can browse the internet while on the lathe. :-)

A definite advantage G.

Bob

The tach feedback goes to the drive, the encoder goes to the controller. I
learned the most about my system due to a lot of backlash in it when I first
bought it. Due to shims and being out of adjustment, my CNC mill had around
a half turn of the handle of play. Lucky the problem wasn't in the
ballscrews, it's just that the bearings traveled in their holders for a
little bit before moving the table.

Here's how it responded. If the table was out a half thou per my DRO
scales, the handwheel would move slowly until it was in position, if it were
a thou out, that handwheel would move around twice as fast, almost
unnoticeable while running, I noticed the backlash when watching the
handwheel moving while a half thou out. So, the control knows it's a half
thou out, sends a move command proportional to the error (.0005") and the
motor moves in the correct direction at a controlled speed, thanks to the
tach feedback, until it gets in position. What I like about this system is
that it positions accurately and is stable even with backlash in the screws.
It goes off of table position, not just motor position, and does accurate
work. The performance improved when I got the shims correct to eliminate
the backlash. If it weren't for the tach feedback, the system wouldn't have
been stable with the backlash. This doesn't excuse sloppy equipment, but it
is nice to know the system is stable and accurate even with some backlash.

RogerN

Bob Engelhardt wrote:
Jon Elson wrote:
Of course these can be used for CNC. You just need a servo interface
that puts
out +/- 10 V analog velocity commands. ...

Then this amp would just be a driver, and its tach feedback/servo
circuitry wouldn't be used?

Bob
No. The CNC control looks at position, only, via the encoder signal.
The velocity servo amp looks at velocity via the tach signal. The CNC
control commands a particular velocity by an analog output, and the
servo amp adjusts motor current to exactly match the commanded velocity.
The velocity error amp in the servo compares commanded to actual
velocity and creates an error signal, which is used as a commanded motor
current. (Motor current equals torque.) That current command signal
goes to a current error amp, and the amp adjusts PWM duty cycle to give
the requested motor current.

This is the "classic" velocity servo scheme used since the late 1960's
on all manner of servo-driven machines.

Jon

On 2009-06-15, Bob Engelhardt wrote:
RogerN wrote:
I have an older Anilam CNC Bridgeport and Anilam CNC lathe. They both use
motors with tach feedback, the mill uses DRO scales for position feedback to
control and the lathe uses encoders coupled to the ballscrews. I replaced
the original amplifiers with AMC amplifiers similar to those in your link,
they work great.

How does you system use both tach feedback & encoders? It seems like
the encoders would give all the feedback that's needed.

The controller puts out a voltage proportional to the desired
speed, and the tach generator produces a voltage which is compared to
that voltage and is constantly adjusted by the amplifier to maintain the
speed -- all without the CPU needing to pay attention once it has put
out the voltage.

The computer checks the counter which keeps track of the encoder
pulses and compares that to the predicted position every so often, and
if necessary adjusts the output speed-command voltage. The counter is
part of the controller card, and also does not need constant CPU
attention, so the computer system can be doing look-ahead calculations
-- and visiting the internet if you so desire.

In contrast -- with steppers, the computer has to keep putting
out pulses -- and if the system is fitted with an encoder, to compare
what the encoder says has happened with what the computer *thinks*
should have happened.

Mostly -- every pulse requires an interrupt and a task switch
which is a rather busy thing for the computer to do. In contrast,
telling a D/A converter to output a voltage, and every so often reading
and resetting the encoder counter is a much lower load on the computer.

Using the encoder as a speed feedback requires a lot of work on
the CPU's part -- checking *every* pulse from the encoder. Using the
tach feedback simply involved telling it what to do and then relaxing.

And without the tach feedback, the computer will also need to
constantly be changing the voltage output to the amplifier -- or to
constantly be putting out pulses to an amplifier which makes the servo
motor pretend to be a stepper motor.

Depending on the precision needed, it is possible to cut a
shallow angle by outputting two voltages -- one for the X axis and one of
the Y axis (and perhaps one for the Z axis as well, if you are being
fancy) and then forget about it until it thinks that one of the axes is
getting close to position.

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 ---

On 2009-06-15, Bob Engelhardt wrote:
Jon Elson wrote:
Of course these can be used for CNC. You just need a servo interface
that puts
out +/- 10 V analog velocity commands. ...

Then this amp would just be a driver, and its tach feedback/servo
circuitry wouldn't be used?

No. The interface described above puts out a voltage somewhere
between -10V and +10V -- proportional to the speed (and direction)
required.

The servo amplifier compares this voltage with the tach feedback
and adjusts the voltage to maintain the speed as commanded -- even as
the load varies. And this is all happening without the computer needing
to get involved until it is near time to change the speed of one or more
axes.

The encoder is used to tell the computer what is happening as a
double-check on the servo amplifier's behavior.

I've hooked up a servo amplifier and servo motor with tach
feedback to a precision power supply which can produce up to 20 volts in
steps of 0.001 V. And there is a pot which will allow adjustment
between any 0.001 V value and then next 0.001V higher.

I've set up the servo amp's pots so the motor spins at its
fastest speed at 10V from the power supply. When I turn it down to
0.001V -- I need to put a piece of tape on the motor's shaft to tell
that it is moving -- and if you time it between full turns, you will
discover that it is maintaining the same speed -- even if you manage to
put a significant load on the motor's shaft.

All of this is with the DC servo motors and amplifiers. I've
never had an AC servo motor running, so I'm not sure how slow it can be
made to go.

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 ---

Jon & DoN - thanks. Nice explanations. I'm not about to go CNC, but
it's interesting to know about.

Always happy to learn,
Bob