On Sun, 2005-09-04 at 19:15 -0400, Tom Becker wrote:
> > ... a simple PWM controller ... Using a brushless DC motor, [] the
> >speed is locked to the basic frequency of the power waveform.
> Huh? DC power waveform? Please explain.
I know others have already responded, but permit mme to (as usual) go
into greater depth and detail (I know, I know, I am a windbag *grin*).
The designation "DC does NOT imply a steady state current flow, not
even over the short term. It refers ONLY to the directionality of that
flow (e.g., one direction only). And even then, this isn't ALWAYS
correct.
AC refers to Alternating Current, a current which has a regular
reversal of current flow. Dc describes Direct Current, a current which
(generally speaking) has but a single direction of current flow.
Look a a Pulse Code MOdulation signal, or a PWM signal, for example.
Bioth examples of Direct Current waveforms, but hardly steady state.
The weaveforms of both are highly complex. Tis is because, while the
direction of the current flow may not be changing,the AMPLITUDE of that
current flow *IS* constantlychanging, in a complex fashion.
Yes, it's true, DC signals have waveforms, sometimes waveforms even
MORE complicated than many AC signals.
In a PWMed brushless DC motor, for example, there are three "phases"
in the coils. Each phase is fed power at some magnitude sequentially,
with specifica timing (e.g., angular realtionships) between the signals
fed to each coil. The compositite signal looks like a squared up, DC
offset three-phase AC power signal, when viewed on an oscilloscope
(primarly because, for all practical purposes, that is exactly what it
is). Indeed, the only really important difference between a PWM for DC
brushless motors, and a PWM for an AC Induction motor is that the
polarity (current direction flow) for the AC motors must be reversed F
times/second (where F is the power line frequency being fed to the
motor).
Indeed, in most PWM speed controls for AC motors, there is no effort
made to make the waveforms appear anything whatever like a sine wave (at
least, not in the Space Vector approach, which is the one I am working
with).
This is one of mnay reaons that PWM speed control is less expensive,
longer lasting, and much, much more efficient than other approaches.
The varying attenuation of the signal magnitude necessary to imimtate a
sine wave is inefficient, electrically speaking. In PWM, you simple let
time average out your voltages and currentflows, so that the power
delievered tothe armature of the motor is approximately a sine wave.
This eliminates a great dela of power lost in wave shaping, and reduces
the need for and cost of heat sinking, etc., as well as permitting you
to use switching devices with lower maximum carrying capacities, thereby
saving additional money. (no need to run 200% safety margins,if your
devices are running cool, because there is no need to derate them as
much)
Tom