Handy Circuit Gives Systems Flexible Fault Protection

Aug. 4, 2003
Many applications must include a capability to automatically disconnect power from an operating circuit. Such applications include thermal shutdown of high-voltage power supplies in radar and X-ray systems, shutdown to limit inrush current during...

Many applications must include a capability to automatically disconnect power from an operating circuit. Such applications include thermal shutdown of high-voltage power supplies in radar and X-ray systems, shutdown to limit inrush current during power-up or hot-swapping of printed-circuit cards, and shutdown to ensure that a card is properly seated before power is applied. In general, power should also remain off following any system fault.

Figure 1 shows a flexible circuit that limits current or removes power in response to a command from the user or other fault-indicating signal. It accommodates manual-reset (MR), overtemperature, and interlock-switch inputs. U1, for instance, is a circuit-breaker IC designed to offer protection in hot-swap applications. Residing on either the backplane/host side or the removable-card/remote-device side of the backplane connectors, it guards against startup faults when a card or board is inserted into a rack or host with the main power supply turned on.

Two main fault conditions are possible. Discharged filter capacitors on the card or remote device can provide a low impedance to ground that momentarily collapses the host power supply. Or, a card only partially seated in its connector can generate erroneous data. U1 prevents the first condition by regulating inrush current during a programmable startup period, allowing the system to stabilize safely. During normal operation, two internal comparators provide short-circuit and overcurrent protection (DualSpeed/BiLevel capability). The second condition is handled by routing U1's current output (ON) through two pins at the outer edges of the card (Fig. 2).

An internal charge pump generates controlled gate drive for an external n-channel MOSFET power switch, Q1 (Fig. 1, again). Following a fault condition, U1 latches Q1 off until cleared by an external-reset signal. To indicate fault conditions, connect an LED (or other fault indicator like an audible alarm) to U1's open-drain status output (STAT). After a fault is detected, you can reset U1 by pulsing its ON pin low for 20 µs minimum.

The switch debouncer (U2) has inputs protected to ±15 kV for ESD and to ±25 V for circuit faults. Its input (shown connected to a normally closed pushbutton switch) can clear a fault on U1 or supply a manual reset. An internal pull-up to VCC eliminates the need for an external pull-up resistor. Replacing the pushbutton switch with an interlock or single-pole, single-throw (SPST) switch supplies a power-disconnect signal, which can implement a chassis-intrusion interlock or allow a technician to power-down the unit before swapping circuit cards. Because U2's push-pull output can't be directly OR-connected with other signals joined to U1's ON pin, Q2 is added to create an OR-able open-drain connection.

A temperature switch (U3) adds thermal protection, and its open-drain output can be linked either directly (wire-OR'd) to other signals on U1's ON pin, or separately, back to the microprocessor. With its TO220 package option, U3 can be bolted to a heatsink or surface-mounted (SOT23 option) close to known heat-generating sources such as R1, Q1, or the main load.

In this circuit, U3 provides circuit protection by issuing a power-disconnect signal when the circuit temperature becomes critical. A low-power linear regulator (U4) supplies 5 V to the low-voltage components U2 and U3, but U4 isn't meant to provide system power to the load.

Sponsored Recommendations

Comments

To join the conversation, and become an exclusive member of Electronic Design, create an account today!