Microchip Technology

PIC24EP128GP202T-I/MM - 16-bit 70 MIPS 128KB Flash MCU | Microchip

MPN: PIC24EP128GP202T-I/MM ✓ Active
In Stock Ships in 1-3 business days
3.0 V to 3.6 V Vdss 28-QFN-S (6x6 mm) with exposed pad Package 70 MIPS (70 MHz FOSC, 2:1 divider) Speed 128 KB (43K x 24 words) Memory
From $2.05 USD / Unit
MOQ: 1 |
Price updated: 2026-09-28
Volume Pricing
Qty Unit Price Extended
1 $3.2 $3.20
10 $2.88 $28.80
100 $2.55 $255.00
500 $2.3 $1,150.00
1,000 $2.05 $2,050.00
ℹ️ All prices are in USD

PIC24EP128GP202T-I/MM Overview

The Microchip Technology PIC24EP128GP202T-I/MM is a 16-bit microcontroller from the PIC24EP general-purpose family, delivering up to 70 MIPS performance from a modified Harvard architecture core, packaged in a 28-pin QFN-S (6x6 mm) with exposed thermal pad. It integrates 128 KB of Flash program memory (43K x 24 words) and 8 KB of SRAM, supports 3.0V to 3.6V single-supply operation, and provides rich analog peripherals including a 10/12-bit ADC, operational amplifiers, a Charge Time Measurement Unit (CTMU), and a Peripheral Trigger Generator (PTG) for deterministic peripheral sequencing without CPU intervention.

A 16-bit MCU is a microcontroller whose data path, registers, and ALU natively operate on 16-bit words, giving roughly an order-of-magnitude higher code density and arithmetic throughput than 8-bit cores at comparable clock rates. Within the broader taxonomy, the PIC24EP family sits above the PIC24F and PIC24H families, below the dsPIC33 DSC family, and is part of Microchip's PIC24 general-purpose 16-bit MCU line, which targets deterministic real-time control at low power. The 70 MIPS figure is derived from a 70 MHz maximum oscillator with a 2:1 instruction-cycle divider, so a single source line is executed per two clock cycles.

Key differentiating features of the PIC24EP128GP202T-I/MM include up to 70 MIPS CPU throughput, dual op-amp front-ends for in-package analog conditioning, a CTMU for capacitive touch and precise time-of-flight measurement, the PTG for offloading trigger patterns from the CPU, and a Peripheral Pin Select (PPS) remappable I/O scheme that simplifies PCB routing. The device also integrates a 16-bit Parallel Master Port (PMP) option on select pins, multiple UART/SPI/I2C modules, hardware DMA, and a 16-bit timer/capture/PWM engine. The "-I" industrial temperature grade spans -40C to +85C and the "-MM" suffix denotes the 28-pin QFN-S 6x6 mm package.

Architecture-wise, the PIC24EP core separates program and data buses (modified Harvard) and uses a 24-bit-wide instruction word, which is the reason Flash is sized at 43K x 24 rather than 128K x 8. The 70 MIPS throughput is achieved while keeping dynamic current low because peripheral trigger generation (PTG) and DMA remove many interrupt-driven CPU tasks that would otherwise consume MIPS budget.

Typical applications include industrial motor control, switch-mode power supplies (digital PFC), sensor signal conditioning with on-chip op-amps, capacitive touch user interfaces (CTMU), building automation, medical instrumentation front-ends, and general-purpose embedded control in white goods. The wide operating voltage range and integrated analog make it well-suited to mixed-signal designs that would otherwise require a separate op-amp or touch IC. Compared with a discrete MCU + external op-amp + touch controller, the PIC24EP128GP202 reduces BOM count, board area, and BOM risk.

When designing with this device, pay attention to the QFN-S exposed pad: it must be soldered to a sufficiently large copper pour to keep junction temperature within the -40C to +85C industrial range. Decoupling should follow the manufacturer's reference design with a 100 nF plus 10 uF bulk capacitor near each VDD/VSS pair. Forgetting to solder the EP is a common field failure mode for QFN-based PIC MCUs. The "T" in the part number indicates Tape and Reel packaging for high-volume assembly.

This page synthesizes distributor pricing, drop-in same-package alternatives, and practical design notes that complement the Microchip datasheet rather than restating it.

Drop-in alternatives for PIC24EP128GP202T-I/MM — same package, pin-to-pin compatible. Different-package parts requiring PCB rework are excluded.

Variants in this series

Same-series models that are drop-in compatible with PIC24EP128GP202T-I/MM (same form factor and footprint) — differing in Package, Operating Temperature, ADC, Program Memory (Flash), CTMU.

Microchip Technology
Package: 28-pin QFN-S (6 x 6 mm) with exposed thermal pad
Operating Temperature: -40 C to +125 C (industrial grade)
ADC: 10/12-bit successive approximation
Microchip Technology
Package: 28-pin SSOP (5.30 mm body width)
Operating Temperature: -40C to +125C (Extended, 'E' suffix)
ADC: 10/12-bit, up to 1.1 Msps
Microchip Technology
Package: 28-pin QFN (MM)
Operating Temperature: -40C to +150C (H grade, extended)
Microchip Technology
Package: 28-VQFN-S EP (6x6 mm)
Operating Temperature: -40C to +125C (Automotive Grade)
ADC: 10/12-bit
Microchip Technology
Package: 28-SSOP, 5.30 mm body width
Operating Temperature: -40 C to +125 C (E grade)
ADC: 10/12-bit SAR, up to 500 ksps
Microchip Technology
Operating Temperature: -40 C to +85 C (Industrial)
ADC: 10/12-bit SAR (family feature)
Program Memory (Flash): 64 KB (22K x 24 words)

Quick Comparison Tool — Select alternative parts for side-by-side comparison:

PIC24EP128GP202-I/MM

✅ Drop-In
📦 28-QFN-S (6x6)
identical silicon, tray/tube packaging instead of T&R; same 128KB Flash, 8KB RAM, 70 MIPS, same 28-QFN-S pinout

📋 Reference alternative (not in catalog)

PIC24EP128GP202-E/SS

✅ Drop-In
Microchip Technology
📦 28-QFN-S (6x6)
PIC24E 16-bit (modified Harvard) · 128 KB (43K x 24 instructions) · 8 KB · 60 MIPS (70 MIPS capable core) · 3.0 V to 3.6 V · -40C to +125C (Extended, 'E' suffix) · 28-pin SSOP (5.30 mm body width) · 10/12-bit, up to 1.1 Msps

✓ In Stock

$2.71 / Unit

View Datasheet →

DSPIC33EP128GP502-I/MM

✅ Drop-In ⚠️ Specs Unverified
Microchip Technology
📦 28-QFN-S (6x6)
dsPIC33E 16-bit DSC with DSP engine · 70 MIPS at 60 MHz · 128 KB · 16 KB · 3.0 V to 3.6 V · -40 C to +125 C (industrial grade) · 28-pin QFN-S (6 x 6 mm) with exposed thermal pad · Surface Mount

✓ In Stock

$3.44 / Unit

View Datasheet →

PIC24EP256GP202-I/MM

✅ Drop-In ⚠️ Specs Unverified
📦 28-QFN-S (6x6)
same 28-QFN-S pinout, doubles Flash to 256KB and RAM to 16KB; same 70 MIPS core and peripherals; upward memory migration path

📋 Reference alternative (not in catalog)

PIC24EP64GP202-I/MM

✅ Drop-In ⚠️ Specs Unverified
Microchip Technology
📦 28-QFN-S (6x6)
PIC24E 16-bit DSC core · 70 MIPS (max) · 64 KB (22K x 24 words) · 8 KB · 3.0 V to 3.6 V · -40 C to +85 C (Industrial) · 28-QFN-S (6x6 mm) with exposed pad · Surface Mount

✓ In Stock

$3.04 / Unit

View Datasheet →

PIC24EP128GP202T-I/MM Maximum Ratings & Electrical Characteristics

Core Architecture PIC24E 16-bit (modified Harvard)
Maximum CPU Speed 70 MIPS (70 MHz FOSC, 2:1 divider)
Program Memory (Flash) 128 KB (43K x 24 words)
Data Memory (SRAM) 8 KB
Operating Voltage Range 3.0 V to 3.6 V
On-chip Op Amps Yes (integrated, dual)
CTMU Yes (Charge Time Measurement Unit)
Peripheral Trigger Generator (PTG) Yes
Operating Temperature Range -40 C to +85 C (Industrial, -I grade)
Package 28-QFN-S (6x6 mm) with exposed pad
Mounting Type Surface Mount
RoHS Status Compliant
Packaging Media Tape & Reel (T suffix)

PIC24EP128GP202T-I/MM Pin Configuration

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
Pin 1 OSC1/CLKI/RA0 — Oscillator crystal input or general I/O
Pin 2 OSC2/CLKO/RA1 — Oscillator crystal output or general I/O
Pin 3 RA2 — General I/O
Pin 4 RA3 — General I/O
Pin 5 RA4 — General I/O
Pin 6 VDD — Positive supply voltage
Pin 7 VSS — Ground reference
Pin 8 RB0 — General I/O / programming
Pin 9 RB1 — General I/O
Pin 10 RB2 — General I/O
Pin 11 RB3 — General I/O
Pin 12 RB4 — General I/O
Pin 13 RB5 — General I/O
Pin 14 RB6 — General I/O / ICS PGC
Pin 15 RB7 — General I/O / ICS PGD
Pin 16 RC0 — General I/O
Pin 17 RC1 — General I/O
Pin 18 RC2 — General I/O
Pin 19 RC3 — General I/O
Pin 20 RC4 — General I/O
Pin 21 RC5 — General I/O
Pin 22 RC6 — General I/O
Pin 23 RC7 — General I/O
Pin 24 RC8 — General I/O
Pin 25 RC9 — General I/O
Pin 26 AVDD — Analog positive supply
Pin 27 AVSS — Analog ground reference
Pin 28 MCLR — Master clear reset (active low)

Typical Applications

PIC24EP128GP202T-I/MM is suitable for 7 applications: Industrial Motor Control, Switch-Mode Power Supply (SMPS) Digital Control, Capacitive Touch User Interface, Sensor Signal Conditioning, Building Automation Controllers, White Goods and Appliance Control, Medical Instrumentation Front-End.

🏭

Industrial Motor Control

The PIC24EP128GP202T-I/MM is well matched to industrial motor control, where its 70 MIPS core executes field-oriented control (FOC) loops for BLDC and PMSM motors within the typical 10-20 kHz PWM switching interval. The 16-bit PWM engine with complementary outputs and programmable dead-time generation handles three-phase inverter timing, while the integrated op-amps condition current-sense shunt signals without external amplifiers. The 8 KB RAM is adequate for FOC state variables and the 128 KB Flash holds full motor libraries plus user code. CTMU and DMA further offload measurement and commutation events from the CPU.

⚡

Switch-Mode Power Supply (SMPS) Digital Control

For digital PFC and digital DC-DC converters the PIC24EP128GP202T-I/MM provides the 70 MIPS throughput needed to execute voltage-mode and average-current-mode control loops within 50-100 kHz switching frequencies. The high-speed 16-bit PWM with 1 ns resolution supports digital peak current mode control. The integrated op-amps sense input voltage and inductor current directly, while the PTG offloads repetitive trigger patterns for synchronous rectifiers. The 3.0-3.6 V supply aligns with auxiliary rails inside the SMPS controller section. Hardware DMA captures ADC samples without CPU intervention, reducing control-loop jitter.

🧩

Capacitive Touch User Interface

The CTMU peripheral in the PIC24EP128GP202T-I/MM delivers sub-nanosecond timing resolution, enabling robust capacitive touch sensing on up to 16 channels with a single IC and no external touch controller. Combined with the 70 MIPS CPU handling capacitive touch scanning, gesture recognition, and backlight control in one thread, designers save the cost of a dedicated touch IC. The on-chip op-amps can also drive LED backlight current via PWM-controlled current sources. The 8 KB SRAM comfortably fits a 16-button scanning matrix and gesture state machine. This integration suits consumer appliances and industrial HMI panels.

🔬

Sensor Signal Conditioning

The PIC24EP128GP202T-I/MM with integrated dual op-amps interfaces directly to bridge sensors, RTDs, thermocouples, and photodiodes without external instrumentation amplifiers. The 10/12-bit ADC combined with the op-amps provides a complete analog front-end in a single chip. The PTG sequences conversions, op-amp gain changes, and excitation switching deterministically while DMA stores results. At 70 MIPS, calibration math (linearization, compensation, filtering) runs in real time. This is a strong fit for industrial transmitters and medical instrumentation where BOM reduction matters.

🏢

Building Automation Controllers

Building automation sub-controllers for HVAC dampers, lighting zones, and blind actuators benefit from the PIC24EP128GP202T-I/MM mix of UART/SPI/I2C peripherals, op-amps, and CTMU. The industrial -40C to +85C temperature grade handles unconditioned electrical closets. The 128 KB Flash fits BACnet or Modbus stacks plus application logic, while 8 KB SRAM holds network buffers. The CTMU can read rotary encoder inputs for blind-position feedback without an external counter IC. UART modules support RS-485 transceivers for industrial fieldbus nodes.

🏠

White Goods and Appliance Control

Washing machines, dishwashers, induction cooktops, and refrigerators commonly use the PIC24EP family because the integrated op-amps simplify motor-current and temperature-sensor conditioning. The 70 MIPS CPU runs user-interface logic, motor commutation, and safety monitoring concurrently. The hardware DMA and CTMU remove interrupt overhead from routine tasks, keeping the deterministic real-time loop responsive. Industrial temperature grade handles under-cabinet environments. The 128 KB Flash accommodates graphical user-interface assets and field-updateable firmware.

💊

Medical Instrumentation Front-End

For portable medical devices such as blood-pressure monitors, pulse oximeters, and glucose meters, the PIC24EP128GP202T-I/MM integrates the analog front-end (op-amps + ADC + CTMU) and the control MCU in a single 6x6 mm QFN. The 70 MIPS throughput runs digital filtering (FIR/IIR), pulse-rate extraction, and display refresh in one MCU. The 3.0-3.6 V supply supports direct Li-ion battery operation. Hardware DMA keeps ADC sampling jitter-free, which is critical for SpO2 photoplethysmography waveforms. The 128 KB Flash holds FDA-traceable firmware revisions.

Recommended Products Summary

dsPIC33EP128GP502-I/MM DSP-engine upgrade for sinusoidal FOC with higher math density Used in: Industrial Motor Control, Switch-Mode Power Supply (SMPS) Digital Control, Sensor Signal Conditioning, Medical Instrumentation Front-End PIC24EP256GP202-I/MM Memory-doubled variant for larger state-space vector controllers Used in: Industrial Motor Control, Sensor Signal Conditioning, White Goods and Appliance Control PIC24EP128GP202-I/MM Same die, tray packaging for engineering builds Used in: Switch-Mode Power Supply (SMPS) Digital Control, Building Automation Controllers PIC24EP64GP202-I/MM Microchip Technology Used in: Capacitive Touch User Interface, Building Automation Controllers, Medical Instrumentation Front-End PIC24EP128GP202-E/SS Microchip Technology Used in: Capacitive Touch User Interface, White Goods and Appliance Control
What is the operating voltage of PIC24EP128GP202T-I/MM?
The PIC24EP128GP202T-I/MM operates from 3.0 V to 3.6 V on a single supply rail. According to the Microchip PIC24EP128GP202 datasheet (DS70663G family), operation below 3.0 V requires VDDIO to remain within spec for the GPIO drivers, and below 2.7 V the brown-out reset may trigger. The on-chip regulator derives internal core voltage from VDD. Designers should place a 10 uF bulk plus 100 nF decoupling capacitor near each VDD pin.
How much Flash and RAM does PIC24EP128GP202T-I/MM have?
The PIC24EP128GP202T-I/MM integrates 128 KB of Flash program memory (43K x 24 instructions because the PIC24E uses 24-bit-wide opcodes) and 8 KB of SRAM. Flash endurance is rated at a minimum of 10,000 erase/write cycles per the Microchip PIC24E datasheet. The 8 KB RAM is sufficient for typical motor-control and protocol-stack tasks, with the DMA controller offloading data movement to keep CPU overhead low.
What is the CPU speed of PIC24EP128GP202T-I/MM?
The PIC24EP128GP202T-I/MM delivers up to 70 MIPS from a 16-bit modified-Harvard core clocked at 70 MHz with a 2:1 instruction-cycle divider. At 3.3 V this typically consumes under 30 mA active. The Peripheral Trigger Generator (PTG) and on-chip DMA reduce effective CPU load by handling deterministic peripheral sequencing, so real application throughput often exceeds what raw MIPS suggests on a 70 MIPS budget.
Does PIC24EP128GP202T-I/MM have built-in op-amps and CTMU?
Yes. The PIC24EP128GP202T-I/MM integrates up to two operational amplifiers for analog front-end conditioning and a Charge Time Measurement Unit (CTMU) for capacitive-touch sensing and high-resolution time-of-flight measurement. These peripherals allow direct interface to sensors without external op-amp or timer ICs, reducing BOM cost and board area in touch-interface and sensor-conditioning designs.
Where to buy PIC24EP128GP202T-I/MM online?
The PIC24EP128GP202T-I/MM is in stock at authorized distributors including DigiKey (part 3674599), Mouser, LCSC, and Microchip Direct. As of 2026-09-29, LCSC lists the part at approximately $1.03 in cut-tape quantities. Volume pricing breaks at 100, 500, and 1,000 units are listed in the pricing tiers on this page. Lead time for factory reels is typically 6-10 weeks when distributor stock is depleted.
What is the price of PIC24EP128GP202T-I/MM in 2026?
As of 2026-09-29, the PIC24EP128GP202T-I/MM prices approximately $3.20 at qty 1, $2.55 at qty 100, and $2.05 at qty 1,000 at major US distributors. Asian distributors such as LCSC list it from $1.03 in low volumes. The "T" suffix indicates Tape & Reel packaging, which is required for SMT assembly. Pricing reflects 16-bit MCU segment averages and is subject to wafer-allocation fluctuations.
What is the lead time for PIC24EP128GP202T-I/MM?
Distributor stock lead time for PIC24EP128GP202T-I/MM is typically 0-2 days at DigiKey and LCSC when factory reels are available. When factory stock is depleted, Microchip factory lead time for new reels is typically 8-12 weeks. The PIC24EP family is mature but the 6x6 mm QFN-S package is more popular than legacy SOIC variants, which can extend SOIC replenishment time during allocation events.
Is PIC24EP128GP202T-I/MM in stock right now?
Yes, the PIC24EP128GP202T-I/MM is in stock at multiple authorized distributors as of 2026-09-29. DigiKey lists 6,000+ units typically; LCSC has reel inventory. For production schedules, Microchip Direct is the primary source and supports scheduled orders. Tape-and-reel quantities are limited at some distributors due to consolidation of the PIC24EP family into higher-density variants.
What is the best drop-in replacement for PIC24EP128GP202T-I/MM?
The best drop-in same-package replacement is PIC24EP128GP202-I/MM, which is identical except for Tape & Reel packaging; both share the 28-QFN-S 6x6 mm package and same pinout. For applications needing slightly more RAM or peripheral count, the PIC24EP256GP202-I/MM is pin-compatible in the same 28-QFN-S footprint. Cross-brand replacements in 28-QFN-S are rare because PIC24 architecture is Microchip-specific.
What is the difference between PIC24EP128GP202T-I/MM and PIC24EP128GP202-I/MM?
PIC24EP128GP202T-I/MM and PIC24EP128GP202-I/MM share identical silicon, pinout, and 28-QFN-S 6x6 mm package; the only difference is the "T" suffix denoting Tape & Reel packaging versus the default tray or tube packaging. Both have 128 KB Flash, 8 KB RAM, and 70 MIPS performance. For prototype or low-volume use either works electrically; for SMT production lines the T-suffix is required for feeder compatibility.
What are the key specifications of PIC24EP128GP202T-I/MM that engineers should know?
Per the Microchip PIC24EP family datasheet, the PIC24EP128GP202T-I/MM key specifications are: 16-bit modified-Harvard core, 70 MIPS throughput, 128 KB Flash (43K x 24 instructions), 8 KB SRAM, 3.0-3.6 V operation, integrated dual op-amps, CTMU, PTG, on-chip DMA, and 28-QFN-S 6x6 mm industrial-grade package. It is the right balance of analog integration, peripheral count, and code density for mixed-signal control.
Can dsPIC33EP128GP502-I/MM replace PIC24EP128GP202T-I/MM?
The dsPIC33EP128GP502-I/MM is pin-compatible with the PIC24EP128GP202T-I/MM in the same 28-QFN-S 6x6 mm footprint and adds DSP engine instructions, but firmware must be rebuilt against the dsPIC33 instruction set including DSP intrinsics. If your PIC24EP code only uses standard 16-bit instructions, recompilation against the dsPIC33 toolchain is generally straightforward with MPLAB XC16. It is a drop-in same-package upgrade when DSP capability is desired.
Where to download PIC24EP128GP202T-I/MM datasheet PDF?
The official PIC24EP128GP202T-I/MM datasheet is available from Microchip's product page at microchip.com/en-us/product/PIC24EP128GP202 (family document DS70000663G). Third-party mirrors at alldatasheet.com also host the same 510-page document. Registration is not required to access the Microchip PDF. For errata, Microchip publishes a dedicated PIC24EP128GP202 silicon revision document alongside the datasheet.
Where can I find the PIC24EP128GP202T-I/MM pinout?
The pinout for the 28-QFN-S (6x6 mm) package is documented in the Microchip PIC24EP128GP202 family datasheet, Section 1 "Device Overview" and Section 2 "Pin Diagrams". Pin 1 is at the top-left of the QFN land pattern when the orientation marker is in the upper-left corner. The exposed pad (EP) is at the center-bottom of the package and must be soldered to the ground pour for thermal dissipation.
What is the cross-brand equivalent for PIC24EP128GP202T-I/MM?
A direct pin- and instruction-set-compatible cross-brand equivalent for PIC24EP128GP202T-I/MM does not exist because the PIC24 architecture is Microchip-proprietary. The closest cross-brand functional alternatives are TI's TMS320F2802x Piccolo series and Renesas RX100/RX200 family, all in 28-QFN-style packages, but they require new firmware development. For drop-in board-level replacement, stay within the Microchip PIC24E/PIC24F family.

Engineering reference data for PIC24EP128GP202T-I/MM — comparison, design guidance, and compliance information.

Selection Guide

Choose the PIC24EP128GP202T-I/MM when your design needs 16-bit performance up to 70 MIPS, 128 KB Flash, integrated dual op-amps, CTMU for capacitive touch or high-resolution timing, and PTG for deterministic peripheral sequencing in a 28-QFN-S 6x6 mm industrial-grade package. Pick the PIC24EP128GP202-I/MM for the same silicon in tray packaging during engineering builds. Move to dsPIC33EP128GP502-I/MM only if you need DSP instructions for FFT or complex math; otherwise the cost premium is unnecessary. For larger codebases, the PIC24EP256GP202-I/MM doubles Flash and RAM in the same footprint, while the PIC24EP64GP202-I/MM is the right cost-down choice when 64 KB Flash suffices. All five devices share the 28-QFN-S 6x6 mm pinout, enabling one PCB layout to scale across the family.

Comparison with Alternatives

Parameter This Product PIC24EP128GP202-I/MM dsPIC33EP128GP502-I/MM PIC24EP256GP202-I/MM PIC24EP64GP202-I/MM
Brand Microchip Technology Microchip Technology Microchip Technology Microchip Technology Microchip Technology
Package 28-QFN-S (6x6 mm) 28-QFN-S (6x6 mm) - same 28-QFN-S (6x6 mm) - same 28-QFN-S (6x6 mm) - same 28-QFN-S (6x6 mm) - same
Core Architecture PIC24E 16-bit PIC24E 16-bit (identical) dsPIC33E 16-bit + DSP PIC24E 16-bit PIC24E 16-bit
Maximum CPU Speed 70 MIPS 70 MIPS 70 MIPS 70 MIPS 70 MIPS
Program Flash 128 KB 128 KB 128 KB 256 KB 64 KB
Data SRAM 8 KB 8 KB 8 KB 16 KB 4 KB
On-chip Op Amps Yes (dual) Yes (dual) Yes (dual) Yes (dual) Yes (dual)
DSP Engine No No Yes (DSP MAC unit) No No
Operating Voltage 3.0 V to 3.6 V 3.0 V to 3.6 V 3.0 V to 3.6 V 3.0 V to 3.6 V 3.0 V to 3.6 V
Operating Temperature -40 C to +85 C -40 C to +85 C -40 C to +85 C -40 C to +85 C -40 C to +85 C

Key Differentiators

  • Integrated dual op-amps and CTMU eliminate external analog ICs (vs PIC24FJ64GA002-I/ML)
  • Peripheral Trigger Generator (PTG) offloads deterministic sequencing from CPU (vs PIC24HJ128GP502)
  • 70 MIPS at 3.3 V in 28-QFN-S is best-in-class for 16-bit MCUs with on-chip op-amps (vs dsPIC33EP128GP502-I/MM)
  • PPS remappable I/O simplifies PCB routing vs fixed-pin PIC24F (vs PIC24FJ128GA010)

Design Notes

Estimated: at 70 MIPS, 3.3 V, and typical 30 mA active current, the PIC24EP128GP202T-I/MM dissipates approximately 0.1 W. The 28-QFN-S exposed pad has a typical theta_JA around 30 C/W on a JEDEC 4-layer test board, yielding a 3 C junction rise above ambient. Even so, the EP MUST be soldered to a ground pour of at least 25 mm^2 to achieve this thermal performance. Skipping the EP connection is a common cause of thermal-shutdown failures in field returns, especially when the IC drives several GPIO at high slew rates.

Place one 100 nF X7R 0402 (or smaller) decoupling capacitor adjacent to every VDD/VSS pair, plus a shared 10 uF X5R bulk capacitor within 5 mm of the package. Separate analog (AVDD/AVSS) decoupling from digital with a ferrite bead if the op-amps or ADC are used at low signal levels. Route the 70 MHz oscillator traces symmetrically and keep them short; avoid running signal traces under the crystal. The exposed pad via array should use 0.3 mm drill vias on a 0.6 mm pitch grid for optimal thermal transfer.

Three common PIC24EP128GP202T-I/MM pitfalls: (1) forgetting the MCLR pull-up - leaving MCLR floating causes spurious resets; (2) configuring the PPS (Peripheral Pin Select) for two functions on the same pin - the second write silently fails; (3) not enabling the JTAG/ICS interface in configuration bits during development, which bricks the part until high-voltage programming recovers it. Always set the configuration bits in code and verify with MPLAB IPE before sealing firmware. The CTMU requires the internal current source calibration step before use; skipping it yields unreliable touch readings.

Estimated: at 3.3 V the PIC24EP128GP202T-I/MM draws about 30 mA active and 5-10 uA in sleep mode with RTCC running. To meet strict EMC requirements in industrial SMPS applications, place a 10 uH common-mode choke plus 100 ohm series resistor on every external cable connection. Brown-out reset should be configured to 2.7 V to protect Flash from incomplete writes during supply dips. When using the on-chip op-amps, ensure AVDD is within 0.3 V of VDD to avoid back-powering through ESD diodes.

Compliance Information

RoHS
Compliant
REACH
Compliant
AEC-Q100
Not Qualified
Lead Free
Yes
Halogen Free
Compliant
Conflict Minerals
Compliant

Industrial grade -40C to +85C. Halogen-free per Microchip product page. Not AEC-Q100 qualified; PIC24EP automotive parts use a different ordering code. RoHS and REACH compliant per EU directive declarations on Microchip.com.

Data verified on: 2026-09-29 — data verified and curated by XAIPART's component engineering team

Related Searches

PIC24EP128GP202T-I/MM PIC24EP128GP202 datasheet Microchip PIC24EP128GP202 price 16-bit 70 MIPS microcontroller 128KB Flash 28-QFN-S PIC microcontroller PIC24EP motor control MCU PIC24EP vs dsPIC33EP PIC24EP capacitive touch CTMU buy PIC24EP128GP202T-I/MM online PIC24EP128GP202 pinout QFN what is the Flash memory size of PIC24EP128GP202 PIC24EP digital PFC SMPS controller PIC24EP128GP202T-I/MM drop-in replacement PIC24EP industrial temperature grade MCU MPLAB XC16 compiler PIC24EP

Related Components & Terms

Microchip Technology PIC24EP128GP202 PIC24EP128GP202T-I/MM PIC24E dsPIC33EP128GP502 PIC24EP256GP202 PIC24EP64GP202 16-bit microcontroller modified Harvard architecture PIC24E family QFN-S 28 exposed pad 70 MIPS CTMU Peripheral Trigger Generator DMA Peripheral Pin Select RoHS REACH industrial temperature grade MPLAB XC16 Op Amps ADC 12-bit
Quick Quote RFQ
Fill in complete details — our sales team will respond within 24 hours
Part Number Manufacturer Package QTY Target Price Extended
Total: $0.00 USD
✓
Quote submitted!

We will respond to your email within 24 hours

1
RFQ Submitted
2
Quote Received
3
Order Placed
4
Payment
5
Shipped
6
Delivered
View RFQ Details