160 volt 3-8 amp power supply?

[Fred Abse]

On Sun, 07 Nov 2010 01:15:22 -0500, DJ Delorie wrote:

Well, these have 1000-step quadrature encoders on them

There should be an absolute ("commutation") signal, in addition, usually
Gray code, fairly coarse, to sense rotor position relative to poles.
You'll need that, else you get a random movement on startup, even if
you're clever enough to maintain sync using the quadrature encoder, once
it's locked.

--
"For a successful technology, reality must take precedence
over public relations, for nature cannot be fooled."
(Richard Feynman)

 

[Fred Abse]

On Sun, 07 Nov 2010 01:19:06 -0500, DJ Delorie wrote:

On 11/06/2010 12:09 PM, Fred Abse wrote:
BLDC?

Brushless DC. Wired like a three-phase AC motor, but designed to be
driven by a phase-controlled DC signal with a feedback loop from some
sort of shaft-position-sensor. Magnets in the rotor, coils fixed in the
stator.

That's not brushless DC.

That's a 3-phase permanent magnet synchronous motor.

The "modern" preferred type in robots, CNC axis drives, etc. Very flat
torque, right down to 0RPM.

Needs an absolute reference for rotor position, relative to poles. 4-bit
Gray code, repeated for each pole pair is adequate and usual.

How many poles?

--
"For a successful technology, reality must take precedence
over public relations, for nature cannot be fooled."
(Richard Feynman)

 

[Jamie]

DJ Delorie wrote:

On 11/06/2010 12:09 PM, Fred Abse wrote:

BLDC?


Brushless DC. Wired like a three-phase AC motor, but designed to be
driven by a phase-controlled DC signal with a feedback loop from some
sort of shaft-position-sensor. Magnets in the rotor, coils fixed in the
stator.
And the last motor I looked at had a Gray code output on it that had

4 lines. This gives you 16 positions to work with between each phase
so that the drive can hold the shaft in position easily..

 

[Jamie]

Fred Abse wrote:

On Sun, 07 Nov 2010 01:15:22 -0500, DJ Delorie wrote:


Well, these have 1000-step quadrature encoders on them


There should be an absolute ("commutation") signal, in addition, usually
Gray code, fairly coarse, to sense rotor position relative to poles.
You'll need that, else you get a random movement on startup, even if
you're clever enough to maintain sync using the quadrature encoder, once
it's locked.

They made some like that years ago, there was an index on the shaft and

case so that you could find the pole line mechanically with out placing
some DC in the coils to see where it locked position.

 

On Sun, 07 Nov 2010 01:15:22 -0500, DJ Delorie <dj@delorie.com> wrote:

I need to remember to read news more often...

On 11/06/2010 11:08 AM, default wrote:
The point that they are three phase should have been in the original
post. Servo motors can mean a lot of things including AC, DC, stepper
or amplidyne systems.

That changes the power supply design? I can manage the control side of
this, it's the 1000 Watt power supply I'm worried about, or if I'm way
off in my power estimates.

Three phase from zero RPM?

Well, these have 1000-step quadrature encoders on them, I suppose with
enough finesse you could control the position down to a dead stop. The
guy who picked them was thinking CNC.

My guess was no one was thinking RC servo.

The power supply is a variable frequency three phase AC power supply.
The motor may be specified with a locked rotor amperage - but you
don't have to supply locked rotor current as a rule - just limit the
current to practical values for normal operation - starting and
running torque.

The "control side" is the hard part.

With a 160 V three phase AC motor you can probably get by with
rectifying and filtering the 120 volt mains - steady state would be
about 170 volts - via some H bridge switches into the windings.

International rectifier has loads of chips, pass transistors, and
Application Notes for just what you want to do. Most of the VSD's
I've played with, work with raw 120 or 240 single phase or three phase
and don't step it up or down with a filtered linear or switching
supply before it goes to the motor - that adds to the complexity and
decreases efficiency.

With all the hoopla about unity phase green supplies that may be
changing, but I haven't seen a drive like that makes use of that yet.

Down to zero control implies a synchronous motor with a magnet in the
rotor just like a stepper motor. The encoder doesn't tell you much
about the motor design.

BTW Sept "Power Electronics Technology" shows a chip for controlling
stepper motors with a "micro-stepping" technique - or treating it just
like a synchronous AC sine wave motor. Smoother operation from motors
with fewer steps per rotation and less noise among the advantages of
the technique.

 

[DJ Delorie]

On 11/07/2010 11:25 AM, Fred Abse wrote:

On Sun, 07 Nov 2010 01:15:22 -0500, DJ Delorie wrote:

Well, these have 1000-step quadrature encoders on them

There should be an absolute ("commutation") signal, in addition, usually
Gray code, fairly coarse, to sense rotor position relative to poles.
You'll need that, else you get a random movement on startup, even if
you're clever enough to maintain sync using the quadrature encoder, once
it's locked.

It has three hall effect outputs, one for each of the three coils, and a
one-pulse-per-revolution signal off the encoder for zeroing it.
It's a BLDC motor with all the trimmings. I didn't mention any of that
because I was asking about the power supply, not the control and
feedback circuit.