InnoSwitch-CH
Over-Temperature Protection
The thermal shutdown circuitry senses the primary die temperature.
This threshold is set to 142 °C with 75 °C hysteresis. When the die
temperature rises above this threshold the power MOSFET is disabled
and remains disabled until the die temperature falls by 75 °C, at which
point it is re-enabled. A large hysteresis of 75 °C is provided to
prevent over-heating of the PC board due to continuous fault condition.
ꢖꢋ ꢖꢐꢓꢧꢍꢐꢟ ꢜꢔꢓꢎ
ꢊꢋ ꢊeꢗꢉꢕꢑꢍꢐꢟ ꢜꢔꢓꢎ
ꢊꢂꢍꢐꢂ
ꢖꢋ ꢖꢉꢏeꢐeꢑ ꢦꢎꢞ ꢊꢏꢓꢂꢗꢔꢓꢕꢙ
ꢊꢋ ꢖꢉꢏeꢐꢓꢕꢙ ꢦꢎ
Current Limit Operation
The current limit circuit senses the current in the power MOSFET.
When this current exceeds the internal threshold (ILIMIT), the power
MOSFET is turned off for the remainder of that switch cycle. The
current limit state-machine reduces the current limit threshold by
discrete amounts under medium and light loads.
ꢖꢋ ꢃꢒꢂꢉꢣReꢁꢂꢍꢐꢂ
ꢊꢋ ꢖꢉꢏeꢐꢓꢕꢙ ꢦꢎ
ꢂꢃRꢄꢅꢆꢆꢇ
ꢊꢋ ꢌꢍꢁ ꢎꢉꢏeꢐeꢑ
ꢒꢎ ꢏꢓꢂꢔꢓꢕ ꢂꢃR
ꢈꢉ
ꢈꢉ
ꢈꢉ
The leading edge blanking circuit inhibits the current limit comparator
for a short time (tLEB) after the power MOSFET is turned-on. This
leading edge blanking time has been set so that current spikes
caused by capacitance and secondary-side rectifier reverse recovery
time will not cause premature termination of the switching pulse.
Each switching cycle is terminated when the Drain current of the
primary power MOSFET reaches the current limit of the device.
ꢖꢋ ꢢꢉeꢁ ꢂꢉ ꢃꢒꢂꢉꢣReꢁꢂꢍꢐꢂ ꢅꢤꢤ
ꢊꢋ ꢥꢟꢎꢍꢁꢁ ꢠꢓꢁꢗꢔꢍꢐꢙꢓꢕꢙ
ꢀeꢁ
ꢂꢃR
ꢖꢋ ꢊꢏꢓꢂꢗꢔꢓꢕꢙ
ꢊꢋ ꢊeꢕꢑꢁ ꢌꢍꢕꢑꢁꢔꢍꢘꢓꢕꢙ ꢖꢒꢚꢁeꢁ
Auto-Restart
In the event of a fault condition such as output overload, output
short-circuit or external component/pin fault, the InnoSwitch-CH
enters into auto-restart (AR) operation. In auto-restart operation the
power MOSFET switching is disabled for tAR(OFF). There are 2 ways to
enter auto-restart after the secondary has taken control:
ꢖꢋ ꢌꢍꢁ Reꢗeꢓveꢑ
ꢌꢍꢕꢑꢁꢔꢍꢘꢓꢕꢙ
ꢖꢒꢚꢁeꢁ
ꢖꢋ ꢜꢉꢕꢂꢓꢕꢒꢉꢒꢁ ꢊꢏꢓꢂꢗꢔꢓꢕꢙ
ꢊꢋ ꢠꢉeꢁꢕꢡꢂ ꢛꢍꢘe ꢜꢉꢕꢂꢐꢉꢚ
1. Continuous switching requests from the secondary for time period
exceeding tAR.
ꢀeꢁ
2. No requests for switching cycles from the secondary for a time
ꢖꢋ ꢊꢂꢉꢎꢁ ꢊꢏꢓꢂꢗꢔꢓꢕꢙꢞ ꢌꢍꢕꢑꢁ
ꢅveꢐ ꢜꢉꢕꢂꢐꢉꢚ ꢂꢉ ꢊeꢗꢉꢕꢑꢍꢐꢟ
period exceeding tAR(SK)
.
The first condition corresponds to a condition wherein the secondary
controller makes continuous cycle requests without a skipped-cycle
for more than tAR time period. The second method was included to
ensure that if communication is lost, the primary tries to restart again.
Although this should never be the case in normal operation, this can
be useful in the case of system ESD events for example where a loss
of communication due to noise disturbing the secondary controller, is
resolved when the primary restarts after an auto-restart off time.
ꢊꢋ ꢌꢍꢁ ꢛꢍꢘeꢕ
ꢜꢉꢕꢂꢐꢉꢚꢝ
ꢖꢋ ꢈꢉꢂ ꢊꢏꢓꢂꢗꢔꢓꢕꢙ
ꢊꢋ ꢠꢉeꢁꢕꢡꢂ ꢛꢍꢘe ꢜꢉꢕꢂꢐꢉꢚ
ꢀeꢁ
The auto-restart alternately enables and disables the switching of the
power MOSFET until the fault is removed. The auto-restart counter is
gated by the switch oscillator in SOA mode the auto-restart off timer
may appear to be longer.
ꢨꢕꢑ ꢉꢤ ꢌꢍꢕꢑꢁꢔꢍꢘꢓꢕꢙꢞ
ꢊeꢗꢉꢕꢑꢍꢐꢟ ꢜꢉꢕꢂꢐꢉꢚ ꢩꢉꢑe
ꢖꢪꢣ7ꢫ1ꢬꢣ110ꢫ1ꢫ
The auto-restart counter is reset once the primary PRIMARY BYPASS
Figure 6. Primary – Secondary Handshake Flowchart.
pin falls below the undervoltage threshold VBPP-VBPP(HYS)
.
Safe-Operating-Area (SOA) Protection
onwards the secondary is in control of demanding switching cycles
when required.
In the event there are two consecutive cycles where the primary
power MOSFET switch current reaches current limit (ILIM) within the
blanking (tLEB) and current limit (tILD) delay time, the controller will
skip approximately 2.5 cycles or ~25 msec. This provides sufficient
time for reset of the transformer without sacrificing start-up time into
large capacitive load. Auto-restart timing is increased when the
device is operating in SOA-mode.
The handshake flowchart is shown in Figure 6.
In the event the primary stops switching or does not respond to cycle
requests from the secondary during normal operation when the
secondary has control, the handshake protocol is imitated to ensure
that the secondary is ready to assume control once the primary
begins switching again. This protocol for an additional handshake is
also invoked in the event the secondary detects that the primary is
providing more cycles than were requested.
Primary-Secondary Handshake Protocol
At start-up, the primary initially switches without any feedback
information (this is very similar to the operation of a standard
TOPSwitch™, TinySwitch™ or LinkSwitch™ controllers). If no
feedback signals are received during the auto-restart on-time, the
primary goes into auto-restart and repeats. However under normal
conditions, the secondary chip will power-up through the FORWARD
pin or directly from VOUT and then take over control. From then
The most likely event that could require an additional handshake is
when the primary stops switching resulting from a momentary line
drop-out or brown-out event. When the primary resumes operation,
it will default into a start-up condition and attempt to detect hand-
shake pulses from the secondary.
4
Rev. J 10/17
www.power.com