Phase Control
Delaying the current transfer from the preceding t0 the succeeding
phase reduces the average value of the direct voltage. The delay
{s accomplished by controlled diodes, or by reactors in serles
‘connection with regular diodes. The voltage reduction will be
Computed from the time integral of the withheld voltage, and from
the angle of delay.
‘With regular diodes the current transfer between two phases stats as s00n
as the potential ofthe succeeding phase rises above that of the preceding
phase. Thus the d-c circuit is given the highest possible potential at any
Fastant, and the direct voltage is a maximum. If, however, the transition
is delayed, the direct voltage is determined by a lower potential during ¢
fetta petiod of time, and the average value is reduced. This is illustrated
by fig. 7.1 where at crs the potential of phase 2 exceeds that of phase I»
and the voltage across the diode of phase 2, shown in section (c), changes
ig. 7a. Voltages ina 3-puike midpoint connetion operating wih phase contol,
0Phase Control 81
from the reverse to the forward direction. A regular diode cannot absorb
forward voltage and would immediately start to conduct. However,
devices such as controlled diodes, thyratrons, or mercury are rectifiers
have to be “fired” before they conduct, so that the current transfer can be
delayed. Reactors with a strongly marked transition from the unsaturated
to the saturated state are also able to cause such a delay.
Computing the Voltage Reduction from the Time Integral
of the Withheld Voltage
During the delay of the current transfer a certain time integral of voltage
is withheld from the d-c circuit. It corresponds to the crosshatched area
in fig. 7.10, and is identical with the time integral of the voltage that is
absorbed by the rectifying element or the control reactor while delaying
the transfer (represented by the crosshatched area in fig. 7.Lc). This
voltage-time integral can also be regarded as the source of the voltage
reduction during a certain interval and is therefore equal to the average
value of the reduction, Z,, multiplied by the duration of this interval. As
an oxample, ina 3-pulse midpoint connection (fig. 7.14) such a delay occurs
three times during each cycle. Hence for a 3-pulse midpoint connection,
4 duration of _ time integral of voltage withheld from (7)
Fe 4 cycle ~ dec circuit during one delay
This relation permits the computation of E, from the value of the integral.
‘The same relation can be applied to each commutating group of a 3-phase
bridge or a double-wye connection with interphase transformer. In the
bridge connection two 3-pulse commutating groups operate in series, and
the direct voltages as well as the reductions E, have to be added, whereas
in an interphase transformer connection the commutating groups are
paralleled, and the direct voltages as well as the voltage reductions are
‘the same for the individual groups or for the whole system.
‘This method of computing the voltage reduction can be used ifthe delay
is caused by devices such as control reactors that absorb a certain volage-
‘time integral (measured in voltseconds) before they permit the current flow.
AA different approach is necessary if the rectifying elements are “fired”
with a certain angle of delay, but this will be explained later.
Example (Voltage reduction in a magnetically controlled interphase
transformer connection)
-In a magnetically controlled double-wye connection with interphase trans-
former a voltage-time integral of 200 mVsec is absorbed by the reactors82 D-C Characteristics
uring each delay. Compute the voltage reduction F,, assuming that
(f= 60 ops, and that
(@) The interphase transformer is in operation.
() The interphase transformer is not in operation because the load
current is too small to excite the core of the interphase transformer,
Solution, (@) We compute for each commutating group, according to
relation (7.1),
200 {msec}
x [mVsec]
Eg = 200 [mVsee] x 180 [seo™*] = 36 [V]
‘This result represents the reduction for each group as well as the reduction
for the whole system, because the two groups are paralleled.
(0) With the interphase transformer not in operation the voltage-time
integral of 200 mVsec is withheld from the d-c cireuit six times during each
cycle. Hence
£,= 2)
From this we conclude that the control characteristic changes at a current
level where the interphase transformer loses
Aut) = Bn 608 st tg influence. ‘This should be considered
‘when designing the control system.
Computing the Voltage Reduction from
the Angle of Delay
In many cases the voltage reduction is con
trolled by the instant of “firing,” and the
dogree of phase control is a function of the
angle of delay, «. For this condition we
derive a relation between the remaining
direct voltage, Euy — andthe angle «, which i simpler than computing
he reduction .. For one interval of operation, 2a/g, the remaining
ottege is represented by the erosshatched area in ig. 7.2, assuming that
Teo seal! area below the ero lines regarded as a negative contribution and
{s therefore subtracted from the large area, Guided by this figure we
compute:
inlets
cos ot dotPhase Control 8
‘With a few conversions this results in
Egg ~ By = Ey Lin
Ey, (telation 6.3)
aei (72)
‘This relation is surprisingly simple, It is illustrated by fig. 7.3; with an
‘equal eaual
Hy DX Bycora
« o
Fig. 73. Direct voltage with phase contol
accurate scale the cosine curve of fig. 7.36 is presented in the chart of fig.
7.4, Equation (7.2) is valid for any pulse number except q = 1 and does
Jb not need any statement concerning the connection as long as the operation
is regular,
Influence of the D-C Circuit
lation (7.2) was derived under the assumption that the d-c circuit stays
pon the potential of the conducting phase until the transfer to the succeeding
pphase occurs, This condition is satisfied under regular operation where the
'inductance of the load circuit is assumed infinite so that the current flow
is maintained while the direct voltage assumes instantaneous values of
‘opposite polarity. If, however, the inductance is not infinite, the current
‘decays during those intervals, and the conditions on which equation (7.2)
Lis based will be satisfied only ifthe current does not reach zero before the
succeeding phase is being fied,
In a resistive circuit the current is proportional to the voltage, and both
uantities remain zero from the instant of voltage reversal to the instant84 D-C Characteristics
of firing the succeeding phase. With a free-wheeling diode the situation
is similar, since the current is bypassed during the intervals of negative
‘emf and the voltage across the d-c terminals stays close to zero during
these intervals. In both eases the control characteristic changes at a certain
‘angle «, and another expression substitutes for (7.2). It will be derived in
Chapter 13.
104
o a or
Fig. 74. Control characteristic for “regula operation.”
With capacitive load or with a d-c circuit with back-emf and a low value
of impedance, the diodes will nt start to conduct before the alternating
‘oltage exceeds the voltage across the d- circuit, and the relation between,
the direct voltage and the angle of phase control is completely changed.
‘This will be investigated in Chapter 14.
Situation in a Bridge Connection
‘Although a bridge connection does not present anything novel, there may
tee some difficulties in understanding how the results of this chapter can
be applied either to the total system or to the individual commutatingPhase Control as
groups. Therefore the 3-phase bridge connection has been selected as a
sample for this category of circuits, and fig. 7.5 presents the voltage curves
for this connection. In fig. 7.Sa the circuit is conceived as two 3-pulse
‘commutating groups in series connection, and all potentials are referred
to the center point of the transformer, assuming a secondary wye connec-
tion. The total direct voltage is the difference between the potentials of
20
r (fee
Bu
«
Bip cos
Ey
) Bq ze
Bios Ba|
Direct voltage of a 3-phase bridge connection operating with phase contro
indieate that values correspond,
the positive and the negative output terminals. In fig, 7.56 the circuit is
treated as a G-pulse unit, and the total direct voltage is plotted over a
straight line, Comparing both illustrations, we find
1: By of the total system, as wel as E,, ~ E,, i the same in both cases.
2. The double-crosshatched area has the same size in both cases. It
represents the time integral of the voltage that is absorbed by the
rectifying element or by the control reactor while delaying the current
transfer to the succeeding phase, In fig. 7.Sa this voltage-time
integral is withheld from the d-c cireut six times during each cycle;
three times at the positive side, and three times atthe negative side.
In fig. 7b itis also six times retained,
3. The basic law, (Ea, — Eup = C08 a, can be applied either to the
total system or to each individual group, as shown atthe right of these
illustrations,86 D-C Characteristics
Example (3-Phase bridge connection)
‘The direct voltage of a 3-phase bridge connection with By, = 100 V is
reduced by phase control to 60 V. Determine
(a) the angle of phase control,
(}) the time integrat of the voltage absorbed during each delay by the
device that causes the delay, f= 60 eps.
Solution, (a) We compute
Fy — Bs INI 9.6 = cosa
‘eo 100 [V]
a= 53°
‘This value is shown in fig. 74 at (Eye — Ba)/ Eas = 0.6.
(8) The voltage reduction is 40 V, and the voltage-time integral under
consideration is six times withheld from the d-e cirouit during each cycle
Hence the value of this integral is
= 111 [msec]Reactive Voltage Drop
The current transfer between the commutating phases causes the
reactive voltage drop. For computing this voltage drop, we interpret
the transfer as the result of a short circuit in the commutating loop
: and define the “line-to-neutral commutating inductance”.
In Chapter 1 it was shown that a rectifier connection must consist of at
least two alternately conducting branches if a continuous current flow is
demanded. The periodic transfer of the current from one branch to the
other causes a reduction of the direct
voltage andis therefore of special interest.
This transfer is called commutation and
will now be investigated.
‘The Principle of Commutation
For the following investigation we select
‘the 3-pulse midpoint connection as the
“commutating group” of most of the
‘more complex rectifier systems. This con-
nections shown in fig. 8.1, and the curves Fig. 8:1. Commutation ia a 3+
of potential and current are presented by Pulse midpoint connection,
ig. 8.2. Starting at oy, of fig. 8.2 the
“Potential of phase 2exoeeds that of phase 1. This causes the diode of phase
{2to conduct since otherwise it would have to absorb voltage in the forward
| direction. The diode of phase 1 is also conducting, because it still carries
the load current. Thus the potential difference between phases 1 and 2 is
| suddenly forced to be absorbed in a circuit that contains only the resistance
land the inductance of the transformer windings and of the bus bars of
hases 1 and 2 (disregarding the diodes, which have nearly the same
forward voltage drop). ‘This condition is comparable to a short circuit.
fit existed permanently, the current curve would be displaced in relation
7