xray wrote:
I have a lathe that runs faster than what I would like on the slowest
speed. Two answers seem to be available. Replace the motor with 3-phase
motor and a VFD controller, or replace the motor with a DC motor and PWM
controller.

The DC version, I think I understand electronically, although maybe not
effective powerly (I've heard lower DC power motors effectively look
like higher power AC motors -- ?).

The 3-phase VFD (Variable Frequency Drive) seems simple enough in
principle, but I'm wondering what's in the best cost effective versions.

The one I want will take single phase 230V AC and convert it to 3-phase
motor drive output that is variable in freq.

First question: what happens to the input? If it is single phase, I
assume it goes through a full wave rectifier to get it as smooth as
possible. Then what? Caps wouldn't help much at this kind of power I'd
guess, and commercial units are small. Just ignore the bumps?

bridge rectifier + big cap.

for a single-phase input, the crest factor will be fairly nasty. Better
drives have a small (3-5%) line choke to increase the conduction angle.

A fairly hefty DC bus cap is used, typically sized for lifetime - it has
to deal with the extremely high RMS input current, as well as the
high-frequency output current.


I think the drive to the three phases is a form of PWM, probably
microprocessor generated, and done by IGBTs.

almost exclusively digital PWM nowadays. it used to be done with 3
reference sinewaves, a triangular carrier and 3 comparators. And a whole
bunch of other hardware, to deal with the rat**** waveform quality
caused by very low switching frequencies.


So, It's obvious I don't know a lot about this except my first-level
assumptions. Can anyone provide a basic description of what is happening
in these VFDs and what might make better or worser implementations?

2 types of drive: scalar (V/F) & vector control.

Scalar control maintains constant machine flux by keeping the ratio of
output voltage to frequency constant. At half speed, the drive outputs
half voltage (done by halving the PWM duty cycle).

Vout = Vrated*Fout/Frated

Vrated, Frated are nameplate voltage, frequency of machine.

at very low speeds, the stator IR drop becomes significant, so some form
of "boost" is added - the output voltage is Vboost + Vrated*F/Frated

dynamic response isnt too great, but can be improved with speed
feedback (shaft encoder).

Scalar control is based on a steady-state model of the AC machine, and
as such ignores dynamic effects. For what you want, scalar control is
just fine.


Vector control is based on a full-order dynamic model of the machine,
and requires some form of speed feedback - shaft encoder or speed
observer (so-called sensorless operation). Sensorless drives may or may
not work well at zero speed (observers get tricky here, and all drives
are not created equal), but with a shaft encoder, a vector drive can
easily do shaft position control, and provide full torque at zero speed
- its actually kinda neat to set the shaft speed to zero, and try and
turn it by hand (for a small machine). A well-implemented vector
controller will *not* allow the shaft to move.

Vector-controlled induction machines routinely replace DC drives
nowadays, even in demanding applications like steel rolling mills.


Thanks. Hope it generates some interesting observations. I know this can
get deep, but at a first level I'd like to hear the the basic theory
about how the input power might be adapted and controlled.


HTH

Cheers
Terry