RX62T YROTATE-IT-RX62T Motor Control Kit
UM-YROTATE-IT-RX62T Rev.1.00
Page 30 of 51
Jan 15, 2014
15. Start-up procedure
When the motor is in stand-still, the phase of the permanent magnet flux vector cannot be detected with the used
algorithm. So an appropriate start-up procedure has to be applied.
The idea is to move the motor in feed-forward (with higher current than that required to win the load), till a speed at
which the estimation algorithm can work. Then the system can be aligned to the estimated phase, and the current
can be reduced to the strictly necessary quantity.
The following graph illustrates the strategy used (the suffix “
ref
“ stands for
reference
, the suffix “
mea
“ stands for
measured
).
Referring to the graph, the startup procedure (in case of three shunts current reading) is described below.
a)
At the beginning t
0
, the system phase is unknown. No current is imposed to the motor; the system phase is
arbitrarily decided to be
ϑ
a
=0. All the references: id
ref
, iq
ref
and speed
ref
are set to zero.
b)
From the moment t
0
, while the iq
ref
and the speed
ref
are maintained to zero, id
ref
is increased with a ramp till
the value i
start
is reached at the moment t
1
.
The references are referred to an arbitrary (d
a
, q
a
) system based on the arbitrary phase
ϑ
a
. From this moment, the
phase estimation algorithm begins to be performed, and the estimated phase
ϑ
est
is used to calculate the
components of the measured current, referred to the (d, q) system based on the estimated phase, id
mea
and iq
mea
.
The components of the current referred to the arbitrary (d
a
, q
a
) system are controlled to follow the references by the
current PI controllers. On the other hand, since the phase
ϑ
est
is still not correctly estimated, id
mea
and iq
mea
have no
physical meaning. Even if they are not shown in the graph, the applied voltages are subjected to the same treatment
(vd
mea
and vq
mea
are calculated in the algorithm).
c)
At t = t
1
, while iq
ref
is maintained to zero and id
ref
is maintained to its value i
start
, speed
ref
is increased with a
ramp till the value s
start
is reached at the t = t
2
. The system phase
ϑ
a
(t) is obtained simply by integration of
speed
ref
; in the meanwhile, the phase estimation algorithm begins to align with the real system phase.
t
1
t
0
t
2
t
3
t
i
start
s
start
id
0
iq
0
speed
ref
id
ref
iq
ref
id
mea
iq
mea