ICMASS.COM - Global Electronic Component Solutions Provider

Q: Where can I buy the CAT823?

Contact ICMASS for current pricing and availability. We stock the CAT823 family and can match the threshold variant to your rail - 3.0V, 3.3V, or 5.0V systems. Same-day quotes available.

All Categories

Language

Home > Blog > Blog

CAT823 Pinout, Specifications & Common Mistakes | ICMASS

2026/8/24 15:04:28

CAT823 Pinout, Specifications & Common Mistakes | ICMASS

The CAT823 is a 5-lead voltage supervisor in SOT-23-5 with reset, watchdog, and manual-reset functions. Its pins: 1 RESET, 2 GND, 3 MR, 4 WDI, 5 VCC — the same map as the MAX823.

It holds your MCU in reset below the 2.93V trip point, releases after 140ms, and reboots the system if the watchdog isn't fed. Three functions, five pins, one tiny package.

CAT823 Pinout and Pin Functions

Pin	Name	Type	Function
1	RESET	Output, push-pull, active low	Held low while V_CC is below the trip point and during the 140ms reset timeout. Drives the MCU's reset pin directly — no pull-up needed.
2	GND	Ground	Common ground. Keep the return short to the MCU's ground pin.
3	MR	Input, active low, debounced	Manual reset. Internal 52kΩ pull-up; a pushbutton to GND generates a full 140ms reset pulse.
4	WDI	Input, watchdog	Watchdog input. Must toggle before the timeout or the supervisor resets. Left open or tri-stated, the watchdog is disabled.
5	VCC	Power	Supply and the monitored rail, 1.2V–5.5V. Decouple with 0.1µF right at the pin.

SOT-23-5 layout: pins 1–3 on the left, 4–5 on the right, pin 1 marked with the package dot. The map matches the MAX823 — a socket-level second source.

Key Specifications

Parameter	Value
Reset Threshold (V_trip)	2.93V (2.85V min / 3.00V max)
Reset Timeout (t_RP)	140ms minimum
Supply Voltage (V_CC)	1.2V – 5.5V
Reset Output	Push-pull, active low
Watchdog	WDI input, disabled when open/tri-stated
Manual Reset	MR input, internal 52kΩ pull-up, debounced
Quiescent Current	4µA typical
Operating Temperature	-40°C to +85°C
Max Power Dissipation	571mW

Numbers that matter: the 2.93V trip with 140ms timeout is the whole behavior of the part — everything else is packaging. Per the onsemi CAT823 datasheet, the threshold spreads 2.85–3.00V, so design for the worst case: a rail that dips to 2.90V will still trigger the reset.

Reset threshold window — where reset actually fires on a 3.0V rail:

V_trip min (worst case, triggers early)2.85V

V_trip typical2.93V

V_trip max (triggers late)3.00V

The 2.85V worst-case trip is only 0.15V below a nominal 3.0V rail — a regulator with 5% droop at load transients can sit inside the window. Size the rail for margin, not for the typical number.

Typical Application Circuit

VCC and GND to the rail, RESET to the MCU's reset pin, MR to a pushbutton, WDI to a free GPIO that toggles in the main loop. That's the entire circuit — the 52kΩ pull-up and the debounce are internal.

The 0.1µF capacitor at VCC is not optional: the trip window is only 0.15V wide at the worst case, and rail noise in that band is a false reset.

Full five-pin hookup: pin 1 RESET straight into the MCU reset pin (push-pull, no pull-up, no series cap); pin 2 GND; pin 3 MR to a debounced pushbutton; pin 4 WDI from a GPIO that must toggle in the firmware's main loop; pin 5 VCC decoupled with 0.1µF.

Reset timing: on power-up, RESET follows VCC low, then releases exactly 140ms after VCC crosses the trip point on the way up. A brownout dip below the trip re-asserts RESET immediately — and the 140ms counter restarts from the recovery crossing.

Common Mistakes When Using CAT823

1. Adding a capacitor to the RESET line "for noise." The push-pull output doesn't need one, and a cap turns the supervisor into an RC reset circuit — the exact failure mode it exists to replace.

Real forum case: a dev board's "noise suppression" cap generated endless reset complaints until it was removed.

2. Using the 2.93V part on a 5V rail. The trip never fires on a healthy 5V supply, so the whole supervisor is silent — brownouts included. Match the threshold variant to the rail.

3. Not re-feeding the watchdog after a reset. If the firmware's init doesn't start toggling WDI, the first watchdog timeout after reboot fires again — a reset loop. Feed WDI in the init routine, not just in the main loop you think survived.

4. Blaming the supervisor for a supply problem. A reset loop caused by VCC sag under load is a power problem: the MCU wakes, draws current, the rail drops, reset fires, load drops, rail recovers — repeat. The supervisor is the detector, not the cure.

5. Treating the push-pull output like an open-drain one. RESET needs no pull-up — and it can't be wire-ORed with other reset sources. If your design shares a reset bus, you need the open-drain variant of the family, not this one.

6. Skimping on decoupling. The worst-case trip window is only 0.15V below nominal. 0.1µF right at VCC, 10–100nF at the MCU, and a ground return that doesn't wander under load currents — the forum's standard first-line fix for spurious resets.

Frequently Asked Questions

Q1: What are the CAT823 pin functions?

A: Pin 1 RESET (push-pull active-low output), pin 2 GND, pin 3 MR (manual reset with internal 52kΩ pull-up), pin 4 WDI (watchdog input), pin 5 VCC. Same map as the MAX823 in SOT-23-5.

Q2: What happens if MR is left floating?

A: Nothing — the internal 52kΩ pull-up holds it high and the debounce ignores stray pulses. A pushbutton to GND is the whole interface; no external pull-up or RC needed.

Q3: Does the RESET output need a pull-up resistor?

A: No — the CAT823's RESET is push-pull, so it drives high and low by itself. Open-drain supervisors need pull-ups; this one doesn't. What it can't do is share a wire-OR reset bus with other outputs.

Q4: What happens if I never feed WDI?

A: After the watchdog timeout, the supervisor resets the system — once. Then it keeps resetting every timeout until WDI starts toggling. Leaving WDI open or tri-stated disables the watchdog entirely. Choose deliberately: a board that doesn't feed WDI has no watchdog at all.

Q5: Why does my board reset by itself?

A: The three classic causes are rail noise in the trip window, a sagging supply under load, and missing decoupling. Check the rail with a scope at the moment of reset, then check decoupling before suspecting the part. Forum consensus: the supervisor reports supply problems; it doesn't cause them.

Q6: Is the SOT-23-5 pinout the same as the SC-70 version?

A: No — the SC-70 5-lead package uses a different pin arrangement for RESET and MR. The SOT-23-5 map here (1 RESET, 2 GND, 3 MR, 4 WDI, 5 VCC) does not transfer to the SC-70 body. Always check the package drawing before layout.

Q7: Where can I buy the CAT823?

A: Contact ICMASS for current pricing and availability. We stock the CAT823 family and can match the threshold variant to your rail — 3.0V, 3.3V, or 5.0V systems. Same-day quotes available.