SPA1 commissioning sequence
17
2.3.2 Applying acceleration feedback
Acceleration feedback is another level of control that selectively increases the apparent inertia of the
machine, allowing us to increase the gain still further before the machine becomes unstable. Machines
react in different ways to acceleration feedback so the amount of gain increase can vary between zero and
ten times.
NOTE:
Experience has shown that the use of acceleration feedback is usually not required and therefore
not recommended for machines that have an uncompensated gain of
≥
0.2.
1.
The first stage is to increase the proportional uncompensated gain value so that the machine is just
unstable using the uncompensated gain step, refer to 2.3.1, record which axes are unstable.
2.
It is now necessary to activate the acceleration feedback term (
K
A
) in the axes that are unstable.
•
Enter the machine configuration screen
•
Select the "ServoConfiguration" tab and scroll down to locate the edit boxes for the
acceleration feedback terms
K
A(X)
,
K
A(Y)
and
K
A(Z)
.
•
Set the initial acceleration feedback value (
K
A
) for the axis that has shown signs of instability
to 0.00005. Repeat for each unstable axis (Note: For any stable axes the acceleration
feedback value remains set to zero).
•
Check for stability by running the “Uncompensated Gain Test refer to 2.3.1
3.
If the machine is stable, increasing the uncompensated gain by a further 10 – 20%. If the machine
is displaying instability as for the Y axis in figure 21, increase the value of
K
A
(refer to step 2) in the
unstable axes by a further 0.00005 (to do this follow step 2) and repeat the uncompensated gain
test.
•
If the acceleration feedback value is too high the machine will produce an audible singing and
must be reduced by 10 to 20 % in the unstable axes. Run the “Uncompensated Gain Test
refer to 2.3.1 to ensure the machine is stable. If it is not stable repeat this step.
4.
Repeat these steps, raising the values of
K
A
and
K
P1
until final values of
K
A
and
K
P
are found (see
figure 22). The final value of
K
P
obtained during this process is referred to as
K
P3
.
This will be the new “Uncompensated Gain”.
NOTE:
It is necessary to increase the amount of acceleration feedback slowly in iterative steps as shown
in figure 12, so that the maximum value of
K
P
can be found. This is in the peak between gain induced
instability line and acceleration induced instability line. The consequence of increasing the acceleration
feedback in one large step without increasing the proportional gain on intermediate steps is illustrated in
figure 13, note the final value of
K
P
is far lower than in that obtained in figure 12.
Once the values for acceleration feedback are determined they must be left unchanged. Their values may
be different, but the value of
K
P3
must be the same for each axis.
