NCP1562A, NCP1562B
http://onsemi.com
17
In some instances it may be desired to latch (instead of
auto re--start) the NCP1562 after a cyc le skip event is
detected. This can be easily achieved by adding an external
latch. Figures 35 and 36 show an implementation of an
integrated and a discrete latch, respectively. In general the
circuits work by pulling CSKIP to V
REF
, preventing it from
reaching V
CSKIP(valley)
once the CSKIP voltage reaches the
turn on threshold of the latch. The external latch is cleared
by bringing the UVOV voltage below V
UV
and disabling
V
REF
.
V
REF
C
REF
CSKIP
C
CSKIP
OUTY
V
CC
INA
OE
MC74VHC1GT126
Figure 38. External Latch Implemented using
ON Semiconductor ’s MiniGatet Buffer
The latch in Figure 38 consists of a TTL level tri--state
output buffer from ON Semiconduct or’s MiniGatet
family. The enable (OE) and output (OUTY) terminals are
connected to CSKIP and the V
CC
and INA pins are
connected to V
REF
. The output of the buffer is in a high
impedance mode when OE is low. Once a continuous
current limit condition is detected, the CSKIP timer is
enabl ed and CSKIP begins charging. Once the voltage on
CSKIP reaches the enable threshold of the buffer, the
output of the buffer is pulled to V
REF
, latching the CSKIP
timer. The OE threshold of the buffer is typically 1.5 V.
V
REF
C
REF
C
CSKIP
CSKIP
BSS84L
M2
24.9 kΩ
2N7002L
R
pull--up
M1
Figure 39. External Latch Implemented using
Discrete N and P--Channel MOSFETs
A latch implemented using discrete N and P--channel
MOSFETs is shown in Figure 39. The latch is enabled once
the CSKIP voltage reaches the threshold of M1. Once M1
turns on, it pulls low the gate of M2. CSKIP is then pulled
to V
REF
by M2. It is important to size R
pull--up
correctly. If
R
pull--up
is too big, it will not keep M2 off while V
REF
charges. This will cause the controller to latch duringinitial
power--up. In this particular implementation the turn on
threshold of M1 is 2 V and R
pull--up
is sized to 24.9 k.
Leading Edge Blanking
The current sense signal is prone to leading edge spikes
caused by the powe r switch transitions. The current signal
is usually filtered using an RC low–pass filter to avoid
premature triggering of the current limit circuit. However,
the low pass filter will inevitably change the shape of the
current pulse and also add cost and complexity. The
NCP1562 uses LEB circuitry that blocks out the first 70 ns
(typ) of each current pulse. This removes the leading edge
spikes without altering the c urrent waveform. The blanking
period is disabled during soft--start as the blanki ng period
may be longer than t he startup duty cycle. It is al so disabled
if the output of the Saturation Comparator is low, indicating
that the output is not yet in regulation. This occurs during
power up or during an out put overload condition.
Supply Voltage and Startup Circuit
The NCP1562 internal startup regulator eliminates the
need for e xternal startup components. In addition, this
regula tor increases the efficiency of the supply as it uses no
power when in the normal mode of operation, but instead
uses power supplied by an auxiliary winding. The
NCP1562 incorporates an optimized startup circuit that
reduce s the requirement of the supply capacitor,
particularly important in size constrained applications.
The startup regulator consists of a constant current
source tha t supplies curre nt from the input line voltage
(V
in
) to the supply capacitor on the V
AUX
pin (C
AUX
). The
startup current (I
start
) is typically 10 mA.
Once C
AUX
is charged to 10.3 V (V
AUX(on)
), the startup
regulator is disabled and the outputs are enabled if there are
no UV, OV, cycle skip or therma l shutdown faults. The
startup regulator remains disabled until the lower voltage
threshold (V
AUX(off1)
) of 8.0 V is reached. Once reached,
the startup circuit is enabled. If the bias current requirement
out of C
AUX
is greater than the startup current, V
AUX
will
discharge until reaching the lower voltage threshold
(V
AUX(off2)
) of 7.0 V. Upon reaching V
AUX(off2)
,the
outputs are disabled. Once the outputs are disabled, the bias
current of the IC is reduced, allowing V
AUX
to charge back
up. This mode of operation allows a dramatic reduction in
the size of C
AUX
as not all the power required for startup
needstobestoredbyC
AUX
. This mode of operation is
known as Dynamic Self Supply (DSS). Figure 40 shows the
relationship between V
AUX(on)
,V
AUX(off1)
,V
AUX(off2)
and
UV. As shown in Figure 40, the outputs are not enabled
until the UV fault is removed and V
AUX
reaches V
AUX(on)
.