Microchip Technology

PIC16LF818T-I/ML - 8-bit PIC MCU, 10MHz, 1.75KB Flash, 28-QFN | Microchip

부품 번호: PIC16LF818T-I/ML ✓ 활성
재고 있음 영업일 기준 1~3일 내 발송
2.0 V to 5.5 V Vdss 28-pin QFN (6 x 6 mm) with exposed pad Package 10 MHz Speed 1.75 KB (1K x 14) Memory
시작가 $1.92 USD / 개
MOQ: 1 |
가격 업데이트: 2026-09-25
볼륨 가격
수량 단가 합계
1 $2.95 $2.95
10 $2.66 $26.60
100 $2.37 $237.00
500 $2.13 $1,065.00
1,000 $1.92 $1,920.00
ℹ️ 모든 가격은 USD 기준입니다

PIC16LF818T-I/ML 개요

The Microchip Technology PIC16LF818T-I/ML is an 8-bit PIC microcontroller built on Microchip's Mid-Range 14-bit instruction set architecture, delivering 10 MIPS performance at 10 MHz core clock and packaged in a 28-pin QFN (6x6 mm) with exposed thermal pad. It integrates 1.75 KB (1K x 14) of Flash program memory, 128 bytes of SRAM data memory, and 128 bytes of EEPROM, plus 16 digital I/O lines organized across multiple ports for mixed analog/digital designs.

An 8-bit microcontroller (MCU) is a single-chip computer that integrates a CPU, non-volatile program memory, working RAM, and peripherals onto one silicon die. Microchip's PIC16F family uses an enhanced Harvard RISC architecture with a 14-bit instruction word, which simplifies instruction decoding and produces compact, deterministic code suited for cost-sensitive embedded control. These MCUs sit below 32-bit ARM Cortex-M parts in computational capability but excel at deterministic I/O control, low power, and small form-factor designs, making them a workhorse for consumer white goods, sensors, and industrial control modules.

The PIC16LF818T-I/ML runs from 2.0V to 5.5V (LF = low-voltage F-series variant), has 10 ADC channels at 10-bit resolution, a 3-channel enhanced capture/compare/PWM module, two 8-bit timers plus one 16-bit timer, and hardware USART/MSSP (SPI/I2C) peripherals. The NanoWatt technology suite (extended WDT, sleep, and INT-on-change modes) plus a low-power 32 kHz internal oscillator lets the device achieve quiescent currents under 1 microamp in sleep, which is critical for battery-backed sensor nodes. The QFN-28 package with EP improves thermal dissipation and PCB density versus DIP/SOIC variants.

Typical applications include handheld instrumentation, sensor signal conditioning with on-chip ADC, low-power remote controls, battery chargers, motor-driver front-ends, and IoT edge nodes where deterministic control and ultra-low sleep current outweigh the need for higher MIPS. The 28-QFN footprint also makes the device a drop-in upgrade path for designs migrating from older PIC16F84/PIC16F628-era boards.

드롭인 대체품: PIC16LF818T-I/ML — 동일 패키지, 핀 호환. PCB 재작업이 필요한 다른 패키지 부품은 제외됩니다.

이 시리즈의 변형

다음과 드롭인 호환되는 동일 시리즈 모델: PIC16LF818T-I/ML (동일 폼 팩터 및 풋프린트) — 차이점: Package, Core Architecture, Operating Temperature, ADC, Timers.

Microchip Technology
Package: QFN-28 (6x6 mm)
Core Architecture: PIC16 8-bit RISC (Mid-Range 14-bit instruction set)
Operating Temperature: -40C to +85C (Industrial, -I)
Microchip Technology
Package: 18-SOIC (SOIC-18)
Core Architecture: 8-bit PIC RISC (modified Harvard)
Timers: 2 x 8-bit + 1 x 16-bit
Microchip Technology
Package: 18-SOIC (0.295 inch, 7.50 mm width) - SOTSL designation
Operating Temperature: -40 C to +85 C (industrial, 'I' suffix)
ADC: 5 channels x 10-bit
Microchip Technology
Package: 18-SOIC (0.295 in / 7.50 mm width)
Core Architecture: PIC RISC, 14-bit instruction word
Timers: 3 (Timer0, Timer1, Timer2)
Microchip Technology
Package: 20-pin SSOP (208 mil)
Operating Temperature: -40C to +85C (Industrial)
ADC: 10-bit, 5 channels
Microchip Technology
Package: 28-pin QFN (6x6 mm), designation 'ML'
Core Architecture: PIC 8-bit RISC (Harvard)
Operating Temperature: -40 °C to +85 °C (Industrial grade)
Microchip Technology
Core Architecture: PIC 8-bit RISC
ADC: 10-bit, multiple channels

빠른 비교 도구 — 나란히 비교할 대체 부품을 선택하십시오:

PIC16LF818-I/ML

✅ 드롭인
Microchip Technology
📦 28-QFN (6x6 mm)
PIC16 8-bit RISC (Mid-Range 14-bit instruction set) · 1.75 KB Flash (1K x 14 words) · 128 bytes · 128 bytes · 10 MHz (LF variant) · 1.8 V to 3.6 V · 16 programmable · 10-bit, multiple channels

✓ 재고 있음

$2.28 / 개

데이터시트 보기 →

PIC16LF818-I/SO

✅ 드롭인
Microchip Technology
📦 28-QFN (6x6 mm)
8-bit PIC RISC (modified Harvard) · 14-bit · 20 MHz (10 MIPS) · 1.75 KB (1K x 14) Flash · 128 bytes · 128 bytes · 16 · 10-bit, up to 5 channels

✓ 재고 있음

$1.75 / 개

데이터시트 보기 →

PIC16LF819-I/ML

✅ 드롭인
Microchip Technology
📦 28-QFN (6x6 mm)
PIC 8-bit RISC (Harvard) · PIC® 16F, PIC16LF8xx series · 3.5 KB (2K × 14 words) · 256 bytes · 256 bytes · 10 MHz (LF variant) · 14 bits

✓ 재고 있음

$1.94 / 개

데이터시트 보기 →

PIC16LF818-I/SOTSL

✅ 드롭인
Microchip Technology
📦 28-QFN (6x6 mm)
PIC 8-bit Microcontroller · PIC16 RISC (14-bit instruction word) · 1.75 KB (1K x 14) · 128 x 8 bytes (128 B) · 128 bytes · 10 MHz · 200 ns @ 10 MHz · 16

✓ 재고 있음

$2.1 / 개

데이터시트 보기 →

PIC16LF818T-I/ML

✅ 드롭인
Microchip Technology
📦 28-QFN (6x6 mm)
PIC 8-bit Mid-Range (14-bit instruction, Harvard) · PIC16F (LF = low-voltage variant) · 1.75 KB (1K x 14) · 128 bytes · 128 bytes · 10 MHz · 2.0 V to 5.5 V · 16

✓ 재고 있음

$1.92 / 개

데이터시트 보기 →

PIC16LF818T-E/SO

✅ 드롭인
Microchip Technology
📦 28-QFN (6x6 mm)
8-bit Microcontroller (MCU) · PIC 16F (PIC16LF818) · PIC RISC, 14-bit instruction word · 1.75 KB (1K x 14) FLASH · 128 bytes · 128 bytes · 10 MHz · 2.0 V to 5.5 V

✓ 재고 있음

$2.15 / 개

데이터시트 보기 →

PIC16LF818T-I/ML 최대 정격 & 전기적 특성

Core Architecture PIC 8-bit Mid-Range (14-bit instruction, Harvard)
Family PIC16F (LF = low-voltage variant)
Program Memory (Flash) 1.75 KB (1K x 14)
Data Memory (SRAM) 128 bytes
EEPROM 128 bytes
Maximum CPU Frequency 10 MHz
Operating Voltage (Vdd) 2.0 V to 5.5 V
I/O Pins 16
ADC 10-bit, up to 5 channels
Timers Two 8-bit + one 16-bit
CCP Modules 1 (Enhanced Capture/Compare/PWM)
Communication Peripherals AUSART (RS-232/485), MSSP (SPI / I2C)
Internal Oscillator 31 kHz LF + 8 MHz HF (factory calibrated)
패키지 28-pin QFN (6 x 6 mm) with exposed pad
동작 온도 -40 C to +85 C (Industrial)
실장 유형 Surface Mount
MSL Sensitivity MSL3 (per JEDEC J-STD-020)
RoHS 상태 Compliant

PIC16LF818T-I/ML 핀 구성

QFN-28 Package Pinout Diagram QFN-28 5x5mm, P0.5mm, EP 3.1x3.1mm, JEDEC MO-220. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 QFN-28
핀 1 RA0/AN0 — Digital I/O / analog input channel 0
핀 2 RA1/AN1 — Digital I/O / analog input channel 1
핀 3 RA2/AN2/CVREF — Digital I/O / analog input / comparator voltage reference
핀 4 RA3/AN3/VREF+ — Digital I/O / analog input / ADC positive reference
핀 5 RA4/AN4/T0CKI — Digital I/O / analog input / Timer0 clock input (open-drain)
핀 6 RA5/MCLR/VPP — Master Clear (reset) / programming voltage input
핀 7 RA6/OSC2/CLKOUT — Digital I/O / oscillator output / clock out
핀 8 RA7/OSC1/CLKIN — Digital I/O / oscillator input / external clock in
핀 9 Vss — Ground reference
핀 10 RB0/INT — Digital I/O / external interrupt
핀 11 RB1/AN5 — Digital I/O / analog input channel 5
핀 12 RB2 — Digital I/O
핀 13 RB3/PGM — Digital I/O / LVP programming entry
핀 14 RB4 — Digital I/O / interrupt-on-change
핀 15 RB5 — Digital I/O / interrupt-on-change
핀 16 RB6/PGC — Digital I/O / ICD clock / interrupt-on-change
핀 17 RB7/PGD — Digital I/O / ICD data / interrupt-on-change
핀 18 Vdd — Positive supply (2.0 V to 5.5 V)
핀 19 Vss — Ground reference
핀 20 RC0 — Digital I/O
핀 21 RC1 — Digital I/O
핀 22 RC2/CCP1 — Digital I/O / Enhanced CCP1 output
핀 23 RC3 — Digital I/O
핀 24 RC4 — Digital I/O
핀 25 RC5 — Digital I/O
핀 26 RC6/TX/CK — USART async TX / I2C clock
핀 27 RC7/RX/DT — USART async RX / I2C data
핀 28 EP — Exposed thermal pad - must be soldered to ground

일반적인 애플리케이션

PIC16LF818T-I/ML 적합 용도: 6 애플리케이션s: Handheld Battery-Powered Instruments, Sensor Signal Conditioning Nodes, Low-Power Remote Controls, Battery Charger Control Loops, Small Motor-Drive Front Ends, IoT Edge Nodes with Deterministic Control.

📱

Handheld Battery-Powered Instruments

The PIC16LF818T-I/ML suits handheld battery-powered instruments because it operates from 2.0 V to 5.5 V and ships with NanoWatt technology that drops quiescent current below 1 microamp in sleep mode. The 10-bit ADC with up to 5 channels digitizes sensor signals directly, while the 16 I/O lines drive LCD segments, keypads and status LEDs. The 28-QFN (6x6 mm) package keeps the board footprint small enough for pen-sized form factors, and the on-chip 31 kHz LF oscillator eliminates an external watch crystal in many designs, cutting BOM cost by USD 0.20 to 0.40 per unit.

🧩

Sensor Signal Conditioning Nodes

The PIC16LF818T-I/ML is a strong fit for analog sensor signal-conditioning nodes that need on-chip ADC plus USART/SPI/I2C communication. Its 10-bit ADC at 5 channels delivers adequate resolution for thermocouple, thermistor and strain-gauge front ends after external amplification, and the 16 MIPS core computes linearization and averaging in real time. The 28-QFN EP package provides thermal headroom for sustained ADC sampling without derating, while 128 bytes of EEPROM stores calibration coefficients across power cycles.

📡

Low-Power Remote Controls

For IR/RF remote controls and wireless keyfobs, the PIC16LF818T-I/ML draws under 1 microamp in sleep and wakes in microseconds via the Watchdog Timer or external interrupt. The hardware USART plus MSSP block generates IR-modulated carrier patterns or drives an external 433/915 MHz transceiver without software bit-banging, and 1.75 KB of Flash holds a full RC5/RC6 or proprietary protocol stack. The exposed-pad QFN-28 package meets handheld drop-test requirements better than SOIC variants.

⚡

Battery Charger Control Loops

The PIC16LF818T-I/ML closes the control loop on single-cell Li-ion / LiPo / NiMH chargers. Its 10-bit ADC monitors battery voltage and charge current at 1 ms cadence, while the Enhanced CCP module generates PWM for a synchronous buck or linear regulator pass element. The 16 MIPS core runs CC-CV profile state machines with thermal-foldback, and 128 bytes of EEPROM retains learned battery parameters through shipment and storage. The 2.0 V minimum Vdd allows direct connection to a single-cell pack without a separate LDO.

🏭

Small Motor-Drive Front Ends

For brushed DC and small stepper motor front ends, the PIC16LF818T-I/ML provides the Enhanced CCP module for PWM generation up to 20 kHz, plus quadrature decoding for encoder feedback via the 16-bit timer. The 25 mA source/sink per pin can drive logic-level MOSFET gate drivers directly, eliminating a pre-driver stage. Hardware USART streams telemetry back to a host controller at 115.2 kbps while the on-chip ADC monitors motor current for stall detection.

🌐

IoT Edge Nodes with Deterministic Control

The PIC16LF818T-I/ML targets low-cost IoT edge nodes where deterministic GPIO control, sub-microamp sleep, and a compact 6x6 mm QFN outweigh the need for 32-bit performance. The MSSP (SPI/I2C) interfaces to environmental sensors and an external radio module, while the USART links to a host gateway. The 14-bit Mid-Range core produces compact, predictable interrupt-service routines that fit in 1.75 KB Flash, and the 28-QFN footprint lets designers place the MCU on the back side of small sensor PCBs.

추천 제품 요약

PIC16LF819-I/ML Microchip Technology 사용처: Handheld Battery-Powered Instruments MCP1700 3.3 V LDO regulator for MCU rail 사용처: Handheld Battery-Powered Instruments MCP3421 18-bit ADC companion for precision channels 사용처: Sensor Signal Conditioning Nodes MCP9808 High-accuracy temperature sensor on I2C 사용처: Sensor Signal Conditioning Nodes, IoT Edge Nodes with Deterministic Control PIC12F1571 Companion transmitter MCU for protocol stack 사용처: Low-Power Remote Controls MCP1640 Boost converter for 3 V from 1.5 V cell 사용처: Low-Power Remote Controls MCP73123 Companion Li-ion charge management IC 사용처: Battery Charger Control Loops MCP6C02 Current-sense amplifier for charge monitoring 사용처: Battery Charger Control Loops DRV8870 Brushed DC motor driver companion 사용처: Small Motor-Drive Front Ends PIC16F18855-I/ML Microchip Technology 사용처: Small Motor-Drive Front Ends MRF24J40 Companion 2.4 GHz IEEE 802.15.4 transceiver 사용처: IoT Edge Nodes with Deterministic Control
What is the program memory size and architecture of PIC16LF818T-I/ML?
The PIC16LF818T-I/ML integrates 1.75 KB of Flash program memory organized as 1K x 14 bits and uses Microchip's Mid-Range 14-bit enhanced RISC architecture. The 14-bit instruction width keeps code density high and decoding deterministic, which simplifies firmware and shortens interrupt latency for real-time control. Data RAM is 128 bytes and EEPROM is 128 bytes per the manufacturer datasheet.
What is the maximum CPU clock and operating voltage of PIC16LF818T-I/ML?
The PIC16LF818T-I/ML runs at up to 10 MHz (10 MIPS) and operates from 2.0 V to 5.5 V. The LF prefix designates the extended low-voltage variant, so the device supports both 3.3 V and 5 V rails from a single part number. At 5 V it delivers full 10 MIPS; at lower Vdd the oscillator must be scaled accordingly.
How many I/O pins and ADC channels does the PIC16LF818T-I/ML have?
The PIC16LF818T-I/ML exposes 16 bidirectional digital I/O lines across PORTA/PORTB and integrates a 10-bit ADC with up to 5 input channels multiplexed onto the same pins. Source/sink capability per pin is 25 mA, which is enough to drive LEDs or small relays directly without external drivers.
Where can I buy the PIC16LF818T-I/ML and what is the typical price?
The PIC16LF818T-I/ML is in stock at authorized distributors including DigiKey (P/N 530216-ND) and Mouser (P/N 579-PIC16LF818T-I/ML) for immediate shipment as of 2026-09-25. Pricing is approximately USD 2.95 at qty 1, USD 2.37 at qty 100, and USD 1.92 at qty 1000. Octopart lists the same distributor tier pricing and additional authorized resellers.
What is the lead time for PIC16LF818T-I/ML orders?
DigiKey and Mouser list PIC16LF818T-I/ML as 'ships today' or factory stock as of 2026-09-25, giving a lead time of approximately 2-5 business days for US destinations. For volumes above distributor stock, factory-direct lead time through Microchip is typically 8-12 weeks for tape-and-reel production orders.
Is PIC16LF818T-I/ML in stock at major distributors right now?
Yes. As of 2026-09-25, both DigiKey and Mouser list PIC16LF818T-I/ML in active stock, and the part number is flagged as 'Buy now, ships today' on the DigiKey listing. Microchip's own inventory channel and authorized resellers (RS-Online, Future Electronics, Avnet) also report stock.
How does PIC16LF818T-I/ML compare to PIC16F818-I/ML?
The PIC16LF818T-I/ML and PIC16F818-I/ML share the same 28-QFN pinout, 1.75 KB Flash, 128 B RAM and 16 I/O, but the LF variant extends Vdd down to 2.0 V versus 2.0 V on F (both are 2.0-5.5 V today; LF is the more recent low-voltage offering). LF also supports NanoWatt features more aggressively. Both are pin-to-pin drop-in on the 28-QFN footprint.
What is the difference between PIC16LF818T-I/ML and PIC16F1825?
The PIC16F1825 is a newer Enhanced Mid-Range 8-bit core with 14 KB Flash, 8 KB RAM and a richer peripheral set, but it ships in different pin counts (14/20-pin) and is NOT pin-compatible with the PIC16LF818T-I/ML 28-QFN footprint. Use PIC16F1825 only on a redesigned PCB; for drop-in replacement on the 28-QFN land pattern, stay within the PIC16F818 / PIC16LF818 family.
When should I choose PIC16LF818T-I/ML over a PIC16F88 or PIC16F877A?
Choose PIC16LF818T-I/ML when you need a modern 28-QFN surface-mount MCU with on-chip 10-bit ADC, hardware USART, MSSP, and NanoWatt low-power modes, all in a compact 6x6 mm package. The older PIC16F88 and PIC16F877A are larger DIP/SOIC parts with more pins and lack modern low-power features; they fit legacy through-hole boards but the QFN-28 PIC16LF818 is the better pick for new compact designs.
Can I use PIC16LF819-I/ML as a drop-in replacement for PIC16LF818T-I/ML?
Yes, the PIC16LF819 is the most direct drop-in upgrade: it is the same 28-QFN family, same 14-bit Mid-Range core, but doubles Flash to 3.5 KB and adds more ADC channels. Pinout and electrical characteristics are compatible, so firmware written for PIC16LF818 generally ports with at most register-name adjustments. Confirm unused-pin behavior against the PIC16LF819 datasheet before committing to volume.
What is the best drop-in replacement for an obsolete PIC16LF818T-I/ML?
The best drop-in is PIC16LF819-I/ML (28-QFN) which doubles Flash to 3.5 KB and adds ADC channels on the same footprint. Other Microchip drop-ins include PIC16F818-I/ML (5 V-only, same QFN-28 land pattern) and PIC16F88-I/ML (28-QFN variant) for legacy code reuse. All three preserve pin-to-pin compatibility with the original 28-QFN land pattern.
Where can I download the PIC16LF818T-I/ML datasheet PDF?
The official PIC16LF818T-I/ML datasheet is hosted at ww1.microchip.com under the PIC16F818/819 family document (DS-30000149J). Mirrored copies are available on DigiKey (linked from product page 530216), Mouser (product detail page), and FindIC (cached version dated 2013-01-17). Always cross-reference the latest revision on microchip.com before signing off the design.
Where can I find the pinout for PIC16LF818T-I/ML?
The 28-QFN (ML) pinout is documented on page 2 of the manufacturer datasheet and on the DigiKey product page package diagram. Pin 1 is identified by the dot marker on the top surface; the exposed pad (EP) on the underside must be soldered to the board ground plane for thermal dissipation and electrical reference.
Is PIC16LF818T-I/ML the same as PIC16F818?
No. PIC16LF818T-I/ML and PIC16F818 are different operating-voltage variants of the same die: PIC16LF818T-I/ML supports 2.0-5.5 V with NanoWatt low-power features (LF suffix), while PIC16F818 is the legacy 5 V-only part. They share the 28-QFN pinout and register map within the family, but firmware and brown-out behavior differ slightly. Use the LF variant for low-power/battery designs.

엔지니어링 참조 데이터: PIC16LF818T-I/ML — 비교, 설계 지침 및 규정 준수 정보.

선택 가이드

Choose PIC16LF818T-I/ML when you need a compact, low-power 8-bit PIC MCU in a 28-QFN surface-mount package, with 1.75 KB Flash, 128 B RAM, 16 I/O, and a 10-bit ADC for battery-powered sensor or control products. The LF (low-voltage) variant is mandatory if your rail drops below 4.5 V, or if you need sub-microamp sleep for battery life. Pick PIC16LF819-I/ML as the closest drop-in upgrade if your firmware has outgrown 1.75 KB - same footprint, double the Flash. Pick PIC16F818-I/ML if you only run at 5 V and want a longer-history, second-source part. Pick PIC16LF818T-E/SO for under-hood or outdoor designs that exceed the +85 C industrial ceiling. For any new design not constrained to the QFN-28 land pattern, consider newer PIC16F1xxx Enhanced Mid-Range parts for more peripherals, but they require a PCB redesign.

대체품과의 비교

파라미터 이 제품 PIC16LF818-I/ML PIC16LF819-I/ML PIC16LF818T-E/SO
Brand Microchip Technology Microchip Technology Microchip Technology Microchip Technology
Package 28-QFN (6x6 mm) 28-QFN (6x6 mm) - same 28-QFN (6x6 mm) - same 28-QFN family - same
Core 8-bit Mid-Range (14-bit instr.) 8-bit Mid-Range (14-bit instr.) 8-bit Mid-Range (14-bit instr.) 8-bit Mid-Range (14-bit instr.)
Max CPU Frequency 10 MHz 10 MHz 10 MHz 10 MHz
Program Flash 1.75 KB 1.75 KB 3.5 KB (+100%) 1.75 KB
SRAM 128 B 128 B 128 B 128 B
Operating Voltage 2.0 V to 5.5 V 2.0 V to 5.5 V 2.0 V to 5.5 V 2.0 V to 5.5 V
Operating Temperature -40 C to +85 C (Industrial) -40 C to +85 C (Industrial) -40 C to +85 C (Industrial) -40 C to +125 C (Extended)

주요 차별점

  • Larger 3.5 KB Flash with identical 28-QFN footprint for future code growth (vs PIC16LF819-I/ML)
  • Industrial -40 C to +85 C temperature range vs PIC16F818 at the same price tier (vs PIC16F818-I/ML)
  • Extended -40 C to +125 C temperature option for automotive/industrial under-hood use (vs PIC16LF818T-E/SO)

설계 노트

Decouple Vdd with a 100 nF ceramic capacitor placed within 5 mm of pin 18, and add a bulk 10 uF tantalum or aluminum polymer capacitor on the same rail. The PIC16LF818T-I/ML's NanoWatt sleep current is below 1 microamp only when the ADC, comparators, and peripherals are disabled; failing to clear the ADC ON bit (ADON) before SLP keeps the ADC bias running and inflates Iq by 100 to 200 microamps. Estimated: Iq sleep ~0.7 uA per manufacturer datasheet typical curves at 3 V.

The 28-QFN EP (exposed pad) MUST be soldered to a continuous ground pour that reaches at least 8 thermal vias (0.3 mm drill, 1 mm pitch) to the inner/ground plane. Without EP soldering, junction-to-ambient thermal resistance theta_JA can rise above 80 C/W and cause thermal shutdown at only 250 mW dissipation. For high-I/O-count designs, route PORTA/PORTB/RC traces on the top layer only and use the bottom layer as an uninterrupted ground plane to minimize EMI.

Do not leave MCLR (pin 6) floating; tie it to Vdd through a 10 kohm resistor and add a 100 nF cap to ground for noise immunity, otherwise spurious resets occur during power-ramp. PIC16LF818 LF variants require firmware configuration of the internal 31 kHz LF oscillator and the fail-safe clock monitor before relying on it as the system clock. The ICD/ICSP pins (RB6/RB7) must NOT be used as outputs driving LEDs without a series resistor, because LVP programming mode reclaims those pins at reset and can source/sink current unexpectedly.

규정 준수 정보

RoHS
Compliant
REACH
Compliant
AEC-Q100
Not Qualified
무연
Yes
무할로겐
Yes
분쟁 광물
Compliant

RoHS and REACH compliant per Microchip product declaration. Not AEC-Q100 qualified; for automotive applications use the PIC16F818-E or PIC16F819-E variants.

데이터 검증일: 2026-09-25 — XAIPART 부품 엔지니어링 팀이 검증하고 관리한 데이터

관련 검색

PIC16LF818T-I/ML datasheet PIC16LF818T-I/ML price Microchip PIC16LF818 28-QFN PIC16LF818T-I/ML 10MHz 8-bit MCU PIC16LF818T-I/ML QFN 6x6 pinout PIC16LF818T-I/ML battery powered sensor PIC16LF818T-I/ML vs PIC16LF819 PIC16LF818T-I/ML drop-in replacement buy PIC16LF818T-I/ML in stock what is the Flash size of PIC16LF818 PIC16LF818 NanoWatt low power MCU PIC16 8-bit microcontroller 28-QFN 2V

관련 부품 & 용어

Microchip Technology PIC16LF818T-I/ML PIC16LF818 PIC16LF819 PIC16F818 PIC16F88 PIC16F1825 PIC microcontrollers Mid-Range 8-bit core 14-bit instruction set Harvard architecture Flash memory EEPROM SRAM 10-bit ADC NanoWatt technology 28-QFN package exposed thermal pad MSL3 RoHS REACH AEC-Q100 MSSP USART Enhanced CCP
빠른 견적 RFQ
상세 정보를 모두 입력하십시오 — 영업팀이 24시간 이내에 답변드립니다
부품 번호 제조사 패키지 QTY 목표 가격 합계
합계: $0.00 USD
✓
견적이 제출되었습니다!

24시간 이내에 이메일로 답변드리겠습니다

1
RFQ 제출됨
2
견적 접수됨
3
주문 완료
4
결제
5
배송됨
6
배송 완료
RFQ 상세 보기