mc1451a-e ETC-unknow, mc1451a-e Datasheet - Page 23

• Manufacturer Part Number: mc1451a-e
• Description: Advanced Brushless Motor Control Chipset
• Manufacturer: ETC-unknow

where:
All multiplied error quantities are saturated to fit within a 16 bit number
so that no discontinuities in the output signal occur at values beyond +/-
2 15 . The integral term is actually maintained to an accuracy of 24 bits,
but only the top 16 bits are used. This results in a more useful range for
Ki, the integral gain.
The result of this calculation is a 16 bit number. The top 10 bits of this
result become the output value if the motor output mode is PWM (10
bits magnitude), and the entire word is used if the mode is DAC16.
Care should be taken when setting a Ki value for the first time. If
the system has already been running and the integration value is
unknown, an abrupt 'jump' may occur when the Ki value is set to a
non-zero value. To avoid this set the I_LM (integration limit) to 0,
set the Ki to the desired value, and then set I_LM to the desired
integration limit value. This will 'clear' all prior integration values,
smoothly enabling the integration function from that point
forward.
Motor Limit
In addition to setting various PID gain values the MC1231A also allows
the maximum value output by the filter to be set. This motor limit value
is set using the command SET_MTR_LMT. It can be read back using
the command GET_MTR_LMT.
The specified motor limit affects the filter output such that if the
magnitude of the filter output value (positive as well as negative)
exceeds the motor limit than the output value is maintained at the motor
limit value. Once the filter output value returns below the specified limit
than normal servo filter values are output
The motor limit is only applied during closed loop servo operations,
when the servo filter controls the motor output value. It does not affect
the output motor value applied during open loop operations
Motor Bias
When using an axis which has a net force in one direction or the other
(such as a vertical axis which experiences the force of gravity) the
motor bias function of the PID compensation filter may be useful. By
adding a constant bias value to the filter output, the overall position
error of the filter can be reduced by directly compensating for the
constant force.
The motor bias value is set using the command SET_MTR_BIAS. It can
be read back using the command GET_MTR_BIAS.
E n is the position error at sample time n
TP n is the target position at sample time n
AP n is the actual position at sample time n
Int(E n ) is the integration sum at time sample n
TrgtVel is the current desired velocity in counts/sample
MtrBias is the motor bias value
23
The motor bias value is applied to the filter value at all times when the
chipset is in closed loop mode. If the chipset transitions to open loop
mode (MTR_OFF command is given or a motion error occurs with
automatic motor stop enabled) than the motor bias value will be output
to the motor by itself, until a manual motor command value is given
(SET_MTR_CMD command), at which point this host-provided motor
command value, without modification by the motor bias value, becomes
the active motor command.
The following example illustrates this: If the chipset is in closed loop
mode with a motor bias value of 100, then if a motor off command is
given (MTR_OFF), then the output motor command will be exactly 100.
Thereafter if the host sends a manual motor command of 200 (using the
SET_MTR_CMD command), then the output motor command will be
200. At this instant the chipset is returned to closed loop mode however
(MTR_ON command), the motor bias value will again be added to the
filter output.
If the specified motor bias value does not properly compensate for
the offsetting DC load, then after a motion error with automatic
motor stop enabled or after a MTR_OFF command the axis may
move suddenly in one direction or another. It is the responsibility
of the host to select a motor bias value such that safe motion is
maintained.
Parameter Loading & Updating
Various profile & servo parameters must be specified by the host for an
axis to be controlled in the desired manner. To facilitate precisely
synchronized motion, these parameters and related control commands
are loaded into the chip using a double-buffered scheme. In this
scheme, the parameters and action commands being loaded are not
acted upon (copied from the double-buffered to the active registers)
until an update signal is given.
This update signal can consist of either a "manual" update command,
or one of several conditional breakpoints. Whichever update method is
used, at the time the update occurs, all of the double buffered registers
and commands will be copied to the active registers. Conversely,
before the update occurs, loading the double-buffered registers or
executing the double buffered commands will have no effect on the
system behavior.
The double buffered registers are listed below.