Re: Power Supply for 4-axis CNC stepper driver

There seems to be a little bit of confusion on a couple of points:

1) Any motor (stepper, servo, squirrel in a squirrel cage) delivers
power to a load. It's efficiency will always be less than 100%, so
delivering power to a load will incur additional heating in the
motor. The power supply must account for both. Power supply current
must increase with load.

2) Inertial and frictional loads. A motor does not distinguish
between them. The motor load will be the sum of the two.

Let's say you have a 100lb mass and 20lbs of friction. It will take
20lbs of force to keep it moving. If you want to accelerate the load
at 1G, it will take another 100lbs, 120lbs total.

Some time later you decelerate the mass at 1G. The force required?
Minus 80lbs. Think of the 2 components as living expenses (friction)
and a savings account (inertia). During acceleration you "invest"
$100 ($1 = 1lb) and during deceleration you "withdraw" $100. It costs
$20 in "living expenses" to keep things moving at a constant speed,
whatever it may be.

Mariss


--- In CAD_CAM_EDM_DRO@yahoogroups.com, John Dammeyer <johnd@a...>
wrote:

> Hi Jon,
>
> I think you are confusing the electrical performance of DC Servo

Motors with

> Stepper Motors. A stepper motor doesn't draw any more power when
> accelerating than when sitting still. In fact they are almost the

opposite.

> Servos draw virtually no power while sitting and steppers are

sitting with

> full power (depending on step position) running through the

windings.

> That's why some controllers have a current setback where after x

seconds of

> no step pulses they reduce current by up to 75%.
>
> As I said before, once the chopper current control is active, the

current

> through the windings is limited to the set point; 70% of 6 amps in

each

> winding in our example. But we want to keep our DC power supply at

a

> particular voltage +/- 5% so that other motors maintain the same
> performance. That means we have to have a 60V and (12A * 70%) per

motor

> supply. If running full step then a full 12A per motor is

required. But

> since the micro stepping drives rarely suck the full power supply

current we

> can probably get by with a rule of thumb of 60% of winding currents.
>
> So average case for three micro steppers is 60V * (12A * 60%) * 3

== 1300W.

> Remember, this is still for the worst case of all 3 motors

stopping on one

> of the two micro-step locations where both windings are energized

to the

> 70.7% of nameplate winding current. At ten micro steps per step I

believe

> the probability is 0.05 but there are 3 motors so the probability

drops down

> to 0.000125 or 0.0125%. That's why an 800VA to 1000VA transformer

runs the

> system without too much effort or heating.
>
> Just a reminder too why we want to keep our supply voltage within

5% of our

> chosen point. When we tune the system, we usually adjust

acceleration and

> top speed on one motor at a time. If the other two motors are

sitting at

> idle or even in low current mode, then the motor being set up will

have full

> rail voltage which will help top speed for fast traversing - rarely

do we

> mill at that speed.
>
> So now with all three motors set up, imagine the XY axis are

traversing

> toward the start point at full slew rate while the Z axis moves

from home

> down to a point just above the working surface. All three motors

run

> through the high current point and the XY motors skip a couple of

steps

> because the motor rail voltage dropped down to 45V for 1/8 second.

Oops.

> We're out of position.
>
> Not repeatable. Next time the program runs everything works.

Frustrating

> but that sort of problem can be due to power supply issues; just

hard to

> prove. So my suggestion is build the supply for worst case unless

you are

> selling 100+ systems per year and you want the extra profit. After

all if

> you are milling castings that cost you $75 each and you ruin two

because the

> system had a hiccup the $150 could have been put to the power

supply. Or in

> the course of a day if you lose 1 hour milling time due to slower

fast

> traversals and you are billing $50/hr machine time that's $50 per

day lost.

>
> And if you are a hobbyist, drive that beater one more month before

buying a

> new car and save one month payment for the power supply.
>
> John Dammeyer
>
> John Dammeyer
>
> > -----Original Message-----
> > From: CAD_CAM_EDM_DRO@yahoogroups.com
> > [mailto:CAD_CAM_EDM_DRO@yahoogroups.com] On Behalf Of Jon Elson
> > The 14.4 W (plus some iron losses in the motor and switching
> > losses in the driver) is the power dissipation when standing
> > STILL, only. When
> > accelerating,
> > the power requirement (that's not all loss, the motor's shaft is
> > delivering real
> > power) goes up dramatically. It can rise to about 60 * 7 *
> > .66 = 277 W per motor. Except under exceptional conditions,
> > it is pretty hard for a
> > machine
> > to require this kind of power for very long, however. So, a

power

> > supply that can
> > deliver that power (X number of motors) for only an instant
> > will do OK.
> > Bulk,
> > unregulated power supplies are easily capable of delivering
> > short bursts of power above their continuous ratings.
> >
> > So, the worst case load would be somewhere around 1100 W (277
> > * 4 axes), but a power supply with a conservative 600 - 800 W
> > rating should be
> > completely
> > adequate for a milling machine. If you were building a high-

speed

> > router, it
> > might be better to plan for 1000 W rating, as these machines
> > tend to really keep the motors spinning fast.
> >
> > >
> > >
> > Jon
> >
> >