Re: Question on Power Supply - regulated notes

--

> Some folks say to use unregulated power supplies mostly to deal with
> back voltages (BEMF) from steppers.

(Apologies in advance if I'm playing fast and loose with terms and
technology, but as a general overview this pretty much covers the bases)

Mariss had done a great write-up on this. the regulated PS's have a
chip called a voltage regulator. This. well. regulated voltage.

DC motors spinning without power are generators. So, when you are
decelerating a load, the weight of the machine (some MUCH worse than
others) will try to spin the motor faster, but the driver is
decelerating. The result is the motor is putting power back into the
power supply.

All that power needed for acceleration is not coming back as recovered
energy, kinda like the regenerative braking in cars, the energy used
to accelerate is coming back. This Electro-motive energy is coming
back, hence the name back EMF.

If another motor is driving, that power may just get used, but if
there is not other motor using this it has to go somewhere.

An unregulated PS will store it in the capacitor with little regard
for burps and bumps.

A regulated PS has that chip that tries to dump the energy. That
voltage regulator now is trying to dump energy on it's output.

Mariss posted some time ago that he ran a NEMA23 to 6,000 RPM or sum
such and then measured the BACK EMF, IIRC, it was like 20 to 30%.
But, I don't recall him mentioning any kind of flywheel or pulleys or
such, If it was a test on the motor only, then adding a flywheel could
have really added to those values.

We try to balance ramping speed for this reason. Kinda like as you
decelerate, you allow the power to get to 10 or 20% (no one ever
actually measures this) and ride the curve as you decelerate.

Read the www.geckodrive.com white paper for some insight, but adding a
large capacitor on the output of the regulated PS between the drives
and the PS will help greatly and allow the use of regulated power
supplies.

> This is the MAIN reason you don't
> drive steppers to the max potential.
> They HOPE that not all 3 axis are
> deaccelerating all at the same time.

Some thinking about this allows people to make the PS the proper size.
We do design the PS for full current, then consider if we need that
or not. The efficiency of the Gecko chopper allows you to use 2/3 of
the total of the motors single phase current. Some single step use
double the single phase current all the time.

So, how much do you need ?
I built a machine that runs one axis at a time. period. it moves ,
lock the axis. does and operation, and then that axis stops, the
carriage unlocks and then moves. no two axes can move at that same
time. ergo, I only need enough power for one motor.

On a Bridgeport, or other mill. The X and Y can go flying around, but
the X and Y and Z rarely (if even in the life of the machine) all are
running at max power at any time. Some people figure, run enough
power for two steppers at max.

Routers and large format tables. The Z is often a 2 to 10 inch
stroke, maybe more, but at high speed is a few seconds from end to
end. So no big concern. The carriage is another matter. If you run
leadscrews, it is not as bad as if you run rack and pinion or open
belts. Moving the gantry on rack or belt has the effect of a hug
flywheel and the potential of bunches of back EMF.

Just to touch base, energy is rated in joules. Kinda like CCA (Cold
Cranking AMPS) for the car battery. A capacitor can store just so
much energy, AND is very sensitive to VOLTAGE. Too much voltage and
it will blow like a firecracker. (and make a gooey mess)

Since we KNOW Back EMF exists in this hobby, we know we MUST use
higher voltage caps than out power supply. 50% more is probably about
right. If you have a 65 volt power supply, adding 50% or 33V means a
100V cap. That then offers a 33% cushion between power supply voltage
and max cap rating.

For your (I think I'm still talking to Luke ????) 30 volt max
controller *could* use a 35 volt cap, but I would HIGHLY recommend a
50 volt cap.

Now, if you are in the situation where you have a large router and are
concerned, and are at the point where you need an expert, there are
double E's on here, (EE or Electronics Engineer, ie: Actually school
learnt professionals who don't guess, but KNOW what they are talking
about) and other true guru's and experts who will do the math for you.
They do NOT respond to the basic questions, much like Bill Gates will
never come to your house to show you how to use WORD. They see this
novice PS question too often to use their valuable time. Suffice it
to say, you would pay a couple hundred an hour for some of these guys
time in the real world.

Then KNOW (that word in its trues sense) when a person is at that
stage to need and appreciate their help. If you go back and read old
posts, you'll find those posts that are over your head and may realize
you are listening in to top level stuff, not this introductory stuff.



OK, starting to ramble.... signing off.

Dave