Re: [CAD_CAM_EDM_DRO] Re: 20 Amp 80 V $99 DC Servodrive

Mariss Freimanis wrote:

>
> The servodrive is much more tolerant of flakey accel/decel ramps and
> ragged pulse trains than a step motor. A step motor has a "servo-
> lock" range of +/- 1.8 deg versus about +/- 11 deg for a 1000 line
> encoder. The step motor resonates during accel/decel which subtracts
> from the already meager 1.8 deg margin, the servo does not, so its
> margin is preserved. Also the PID algorithm provides low-pass
> filtering or averaging of the pulse train so a ragged step input
> looks awfully smooth on the encoder, which of course shows what the
> motor is doing.

It sounds like this system only keeps an 8-bit signed difference, and so
it
will lose position in blocks of 128 encoder counts, is what you are
saying,
here. I know the small processors have a lot of housekeeping to do
without
adequate hardware assist, but I think a 128 count difference could lead
to
a loss of position during bind-ups, crashes and emergency stops. This
could
lose the great advantages of closed-loop operation.

If the NC computer was ALSO reading the encoder, then this would cause
some
very odd transients, when the positioning loop tried to correct for a
loss of
128 counts in the velocity loop, but at least the displayed position
would not
be premanently offset.

>
> You still need to accel/decel or you will over-shoot the target
> position. It will settle into position, but that does no good if you
> are milling something, since the damage would already be done. If
> your ramp rates are reasonable (still much steeper than for a
> stepper), the servo will follow within 1 count. If you have a 'scope
> you can look at the position error TP and see exactly where the motor
> is relative to the moving command position. Every 40 mV it is off the
> 5 volt reference is a 1 count error.

My system occasionally shows following errors of about .060" (on severe
accelerations to full rapid speed or e_stops due to impending crashes)
which comes out to 1200 counts at 20,000 counts/inch.

> Right now we think it will stay around $99

This is quite remarkable, as the boards for my PWM servo amp run about
$30
each, and with the component cost adds up to almost $200, not counting
assembly. I don't know how you do it, but that is quite a bargain!

Of course, I have a lot of analog input circuitry (instrumentation amp
inputs for
the velocity command in, and the Tach input), another super
instrumentation
amp for measuring the current on the PWM outputs as they are swinging +-

80 V at 100 KHz). I also have LED bar graphs for velocity command and
current
output, and there are two servo loops in the amp, the inner one controls

armature current, the outer one controls velocity. This amp also
features
a balanced differential 2-pole output filter to remove the 100 KHz PWM
carrier, which would provide a strong radio interference source if not
filtered. It has a transformer-coupled 'charge-pump' to power the
high-side
FET drivers, as well as the high-side fault current sensing circuit, so
that
faults to ground on the motor output wires will cause a shutdown rather
than blowing a high-side FET.

Good luck,

Jon