Silicon Labs

EFM32GG230F1024G-E-QFN64 - 48MHz ARM Cortex-M3 MCU, 1MB Flash | Silicon Labs

Référence : EFM32GG230F1024G-E-QFN64 ✓ Actif
En stock Expédié sous 1 à 3 jours ouvrés
1.85 V to 3.8 V Vdss 0.6 uA Id 64-QFN (9x9 mm) with Exposed Pad Package 48 MHz Speed 1 MB (1M x 8) Memory
À partir de $9.6 USD / unité
MOQ : 1 |
Prix mis à jour : 2026-10-21
Tarifs dégressifs
Qté Prix unitaire Total
1 $15.25 $15,25
10 $13.78 $137,80
100 $12.2 $1 220,00
500 $10.85 $5 425,00
1 000 $9.6 $9 600,00
ℹ️ Tous les prix sont en USD

EFM32GG230F1024G-E-QFN64 Présentation

The Silicon Labs EFM32GG230F1024G-E-QFN64 is an ARM Cortex-M3 Giant Gecko microcontroller IC running at 48 MHz with 1 MB (1M x 8) of Flash and 128 KB of RAM, housed in a 64-pin QFN (9x9 mm) package with exposed thermal pad. It is a 32-bit single-core MCU engineered for energy-sensitive applications, offering autonomous peripherals, AES hardware encryption, and rich analog integration while consuming as little as 0.6 uA in Stop (EM2 Deep Sleep) mode and 180 uA/MHz in Run mode.

What is a 32-bit low-power microcontroller? A microcontroller (MCU) is a system-on-chip that integrates a CPU core, program memory (Flash), data memory (RAM), and peripherals (timers, ADC, communication interfaces) into a single device. Low-power MCUs extend this concept by including multiple energy modes (Run, Sleep, Stop, EM4 shutoff), fast wake-up times, and peripheral reflex logic that lets timers and sensor interfaces operate without waking the CPU. The EFM32GG Giant Gecko family sits in the hierarchy as: Cortex-M3 MCU -> ARM 32-bit MCU -> low-energy MCU -> embedded microcontroller -> semiconductor IC.

Key features include 1024 KB Flash, 128 KB RAM, 56 digital I/O pins, 4 x 16-bit timers, AES-128/256 hardware crypto accelerator, and a peripheral reflex system. The wide supply voltage range of 1.85 V to 3.8 V supports both single-cell Li-ion battery applications and 3.3 V industrial rails, while integrated LDO and DC-DC support enables flexible power tree design with sub-microamp sleep currents critical for battery-powered IoT endpoints.

The Giant Gecko architecture pairs the Cortex-M3 core with Silicon Labs' LESENSE, LETIMER, RTCC, and AUTOSELECT peripherals, allowing sensor interfaces and GPIO events to operate autonomously while the CPU remains in deep sleep. Hardware AES acceleration and a True Random Number Generator (TRNG) target secure connected applications. The 9x9 mm QFN-64 footprint with exposed pad provides 56 usable GPIO - generous for a small-outline MCU package.

Typical applications include battery-powered IoT sensors, smart energy metering, wireless sensor nodes, industrial HMI controllers, and home automation gateways where years of battery life and rich peripheral integration are required.

When designing with this part, place a 1 uF X7R bypass capacitor adjacent to each VDD pin and a 4.7 uF bulk capacitor near the package. The QFN-64 thermal pad must be soldered to a continuous ground pour for both electrical ground and thermal dissipation (theta_JA typically 28 C/W on a 4-layer JEDEC JESD51-7 board).

This page synthesizes distributor pricing, drop-in alternatives, and practical design notes not found in the standalone manufacturer datasheet.

Alternatives de remplacement direct pour EFM32GG230F1024G-E-QFN64 — même boîtier, compatible broche à broche. Les références à boîtier différent nécessitant une retouche du PCB sont exclues.

Variantes de cette série

Modèles de la même série directement compatibles avec EFM32GG230F1024G-E-QFN64 (même facteur de forme et même empreinte) — différant par Package, Series, Flash Memory, RAM, Low-Power Features.

Silicon Labs
Package: 64-QFN, 9 x 9 mm
Series: EFM32 Gecko
Flash Memory: 32 kB
Comparer avec EFM32GG230F1024G-E-QFN64 →
Silicon Labs
Package: 64-QFN, 9 mm x 9 mm
Series: EFM32 Gecko
Flash Memory: 64 kB
Comparer avec EFM32GG230F1024G-E-QFN64 →
Silicon Labs
Package: 64-QFN, 9 x 9 mm
Series: Tiny Gecko
Flash Memory: 8 KB
Comparer avec EFM32GG230F1024G-E-QFN64 →
Silicon Labs
Package: 64-QFN (9 x 9 mm)
Series: EFM32 Tiny Gecko
Flash Memory: 8 kB
Comparer avec EFM32GG230F1024G-E-QFN64 →

Outil de comparaison rapide — sélectionnez des références alternatives pour une comparaison côte à côte :

EFM32GG230F1024-QFN64T

✅ Remplacement direct
📦 64-QFN (9x9)
Same die and 64-QFN package, tape-and-reel packaging variant ('T' suffix); pinout identical

📋 Alternative de référence (hors catalogue)

EFM32GG230F1024G-E-QFN64R

✅ Remplacement direct
📦 64-QFN (9x9)
Same die and 64-QFN package, reel ('R') packaging suffix; pinout identical

📋 Alternative de référence (hors catalogue)

EFM32GG230F512G-E-QFN64

✅ Remplacement direct
📦 64-QFN (9x9)
Same die family in 64-QFN package, Flash reduced to 512 KB (vs 1024 KB); otherwise pin-to-pin

📋 Alternative de référence (hors catalogue)

EFM32GG230F1024-QFN64

✅ Remplacement direct
📦 64-QFN (9x9)
Same die and 64-QFN package, tray packaging (no -T/-R suffix); pinout identical

📋 Alternative de référence (hors catalogue)

EFM32GG230F1024G-E-QFN64 Valeurs maximales et caractéristiques électriques

Manufacturer Silicon Labs
Series EFM32 Giant Gecko
Core Processor ARM Cortex-M3
Core Size 32-Bit Single-Core
Maximum Clock Speed 48 MHz
Program Memory Size (Flash) 1 MB (1M x 8)
RAM Size 128 KB
Number of I/O Pins 56
Supply Voltage (Vcc/Vdd) 1.85 V to 3.8 V
Boîtier 64-QFN (9x9 mm) with Exposed Pad
Type de montage Surface Mount
AES Hardware Encryption Yes (AES-128/256)
Timers 4 x 16-bit
Stop Mode Current (EM2) 0.6 uA
Run Mode Current 180 uA/MHz
Connectivity / Peripherals I2C, SPI, UART, USB, ADC, DAC, GPIO
Statut RoHS Compliant

Configuration des broches de EFM32GG230F1024G-E-QFN64

QFN-64 (8x8mm, EP) Package Pinout Diagram QFN-64 8x8mm, P0.5mm, EP 5.1x5.1mm, JEDEC MO-220. Pin 1 by dot. 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 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 QFN-64 (8x8mm, EP)
Broche 1 VDD — Main supply voltage input (1.85 V to 3.8 V)
Broche 2 VSS — Ground reference
Broche 3 RESETn — Active-low reset input
Broche 4 SWDIO — Serial Wire Debug data I/O
Broche 5 SWCLK — Serial Wire Debug clock input
Broche 6 SWO — Serial Wire trace output (optional)
Broche 7 GPIO — General-purpose I/O with peripheral routing
Broche 8 GPIO — General-purpose I/O
Broche 9 GPIO — General-purpose I/O
Broche 10 GPIO — General-purpose I/O
Broche 11 GPIO — General-purpose I/O
Broche 12 GPIO — General-purpose I/O
Broche 13 GPIO — General-purpose I/O
Broche 14 GPIO — General-purpose I/O
Broche 15 GPIO — General-purpose I/O with ADC capability
Broche 16 GPIO — General-purpose I/O with ADC capability
Broche 17 GPIO — General-purpose I/O
Broche 18 GPIO — General-purpose I/O
Broche 19 GPIO — General-purpose I/O
Broche 20 GPIO — General-purpose I/O
Broche 21 GPIO — General-purpose I/O
Broche 22 GPIO — General-purpose I/O
Broche 23 GPIO — General-purpose I/O
Broche 24 GPIO — General-purpose I/O
Broche 25 GPIO — General-purpose I/O
Broche 26 GPIO — General-purpose I/O
Broche 27 GPIO — General-purpose I/O
Broche 28 GPIO — General-purpose I/O
Broche 29 GPIO — General-purpose I/O
Broche 30 GPIO — General-purpose I/O
Broche 31 GPIO — General-purpose I/O
Broche 32 GPIO — General-purpose I/O
Broche 33 GPIO — General-purpose I/O
Broche 34 GPIO — General-purpose I/O
Broche 35 GPIO — General-purpose I/O with peripheral routing
Broche 36 GPIO — General-purpose I/O with peripheral routing
Broche 37 GPIO — General-purpose I/O
Broche 38 GPIO — General-purpose I/O
Broche 39 GPIO — General-purpose I/O
Broche 40 GPIO — General-purpose I/O
Broche 41 GPIO — General-purpose I/O
Broche 42 GPIO — General-purpose I/O
Broche 43 GPIO — General-purpose I/O
Broche 44 GPIO — General-purpose I/O
Broche 45 GPIO — General-purpose I/O
Broche 46 GPIO — General-purpose I/O
Broche 47 GPIO — General-purpose I/O
Broche 48 GPIO — General-purpose I/O
Broche 49 GPIO — General-purpose I/O
Broche 50 GPIO — General-purpose I/O
Broche 51 GPIO — General-purpose I/O
Broche 52 GPIO — General-purpose I/O
Broche 53 GPIO — General-purpose I/O
Broche 54 GPIO — General-purpose I/O
Broche 55 GPIO — General-purpose I/O
Broche 56 GPIO — General-purpose I/O
Broche 57 GPIO — General-purpose I/O
Broche 58 GPIO — General-purpose I/O
Broche 59 GPIO — General-purpose I/O
Broche 60 GPIO — General-purpose I/O
Broche 61 GPIO — General-purpose I/O
Broche 62 GPIO — General-purpose I/O
Broche 63 GPIO — General-purpose I/O
Broche 64 GPIO — General-purpose I/O

Applications typiques

EFM32GG230F1024G-E-QFN64 convient pour 6 applications: Battery-Powered IoT Sensor Nodes, Smart Energy Metering, Wireless Sensor Network Gateway, Industrial Human-Machine Interface (HMI), Home Automation and Security Hub, Portable Medical and Health Monitoring Devices.

🧩

Battery-Powered IoT Sensor Nodes

The EFM32GG230F1024G-E-QFN64 fits battery-powered IoT sensor nodes because of its 0.6 uA EM2 Deep Sleep current and 180 uA/MHz Run mode draw - combining these with a typical 0.1% active duty cycle yields 5-10 year battery life on a CR2032 coin cell. The 1 MB Flash accommodates over-the-air (OTA) update images plus application code, while 128 KB RAM supports protocol stacks such as Bluetooth Low Energy, Thread, Zigbee, or Matter. The Cortex-M3 at 48 MHz delivers enough compute to run sensor fusion, encryption (AES hardware accelerator offloads secure TLS handshakes), and local decision logic without external radios. Place the MCU adjacent to the sensor ICs, with a 1 uF X7R bypass on each VDD pin and a 4.7 uF bulk decoupling capacitor near the exposed pad - this layout minimizes sleep-current spikes from rail collapse.

⚡

Smart Energy Metering

In smart electricity, water, and gas meters, the EFM32GG230F1024G-E-QFN64 serves as the metering compute and connectivity host. The integrated 12-bit ADC (typical of Giant Gecko) combined with the AES-128/256 hardware accelerator enables tamper-resistant metrology with secure data logging, while the wide 1.85-3.8 V supply range supports direct connection to battery-backed or harvested power rails. The Cortex-M3 at 48 MHz handles metrology algorithms, time-of-use tariff logic, and DLMS/COSEM or wireless M-Bus protocol stacks within the 1 MB Flash budget. The 128 KB RAM is sufficient for buffering daily load profiles and supporting display refresh. The 64-QFN exposed pad provides 28 C/W theta_JA on a 4-layer JEDEC board - adequate for indoor meter enclosures without active cooling.

🌐

Wireless Sensor Network Gateway

As a wireless sensor network gateway host, the EFM32GG230F1024G-E-QFN64 routes data from sub-GHz or 2.4 GHz sensor nodes to backhaul IP networks. The 56 GPIO pins and rich peripheral set (multiple UART, SPI, I2C) let designers connect a companion radio SoC (such as EFR32) plus Ethernet PHY, SD card storage, and local user interfaces. The 128 KB RAM holds packet buffers for time-slotted protocols, while 1 MB Flash stores gateway firmware plus stack libraries. The Cortex-M3 at 48 MHz is sufficient for protocol bridging, encryption using the on-chip AES, and local alarm processing even when forwarding thousands of packets per hour. The 64-QFN (9x9 mm) package fits inside small gateway enclosures, and the MCU's low sleep current keeps standby gateway power below 100 mW.

🏭

Industrial Human-Machine Interface (HMI)

The EFM32GG230F1024G-E-QFN64 drives industrial HMI panels that combine a small TFT or OLED display, capacitive touch, and fieldbus communication. The 1 MB Flash stores graphical assets, fonts, and the application state machine, while the 128 KB RAM holds frame buffers for small displays and protocol state. The Cortex-M3 at 48 MHz refreshes typical QVGA panels over SPI at 30+ FPS using only basic DMA channels, and EBI (external bus interface) on compatible Giant Gecko variants supports larger parallel LCDs. AES hardware encrypts remote firmware updates over Modbus TCP or EtherCAT, addressing industrial cybersecurity requirements. The 64-QFN (9x9 mm) package footprint keeps the HMI controller compact, and the wide -40 C to +85 C industrial temperature range supports factory floor deployment.

🏠

Home Automation and Security Hub

In home automation gateways and security panels, the EFM32GG230F1024G-E-QFN64 provides the local compute and connectivity backbone. The 1 MB Flash stores Z-Wave, Zigbee, or Matter stack libraries plus application firmware with room for future feature additions. The 128 KB RAM supports simultaneous radio coexistence and secure handshake buffers. On-chip AES-128/256 hardware acceleration secures wireless commissioning and prevents device spoofing. The 56 GPIO drive LED indicators, keypads, sirens, and sensors without external expanders. The Cortex-M3 at 48 MHz runs event-driven automation logic with sub-millisecond response. The 64-QFN (9x9 mm) package and 1.85-3.8 V supply simplify power tree design from a single 3.3 V rail or USB-powered supply.

💊

Portable Medical and Health Monitoring Devices

For portable medical devices such as blood glucose meters, pulse oximeters, and ambulatory ECG monitors, the EFM32GG230F1024G-E-QFN64 offers Cortex-M3 compute, low sleep current, and AES hardware encryption for HIPAA/GDPR-relevant patient data. The 0.6 uA EM2 current lets wearable devices sleep most of the day, waking only for measurements - extending coin-cell battery life from weeks to months. The 1 MB Flash stores firmware plus on-device lookup tables and measurement algorithms, while 128 KB RAM buffers raw sensor samples between BLE transmissions. The integrated 12-bit ADC and on-chip comparators interface directly to biomedical front-ends. The 64-QFN (9x9 mm) package fits inside small handheld enclosures. Note: for FDA-classified devices, validation against IEC 60601-1 and IEC 62304 remains the designer's responsibility.

Récapitulatif des produits recommandés

EFR32MG21A010F1024IM32 Wireless SoC for sub-GHz/2.4 GHz IoT connectivity Utilisé dans : Battery-Powered IoT Sensor Nodes SI7021-A20-GMR I2C temperature/humidity sensor Utilisé dans : Battery-Powered IoT Sensor Nodes AFE031 Analog front-end for PLC metering Utilisé dans : Smart Energy Metering MAX31865 RTD-to-digital converter for thermal metering Utilisé dans : Smart Energy Metering EFR32FG14P231A256GM48 Sub-GHz/2.4 GHz Flex Gecko radio transceiver Utilisé dans : Wireless Sensor Network Gateway DP83848IVV Texas Instruments Utilisé dans : Wireless Sensor Network Gateway FT813 Embedded video engine for TFT displays Utilisé dans : Industrial Human-Machine Interface (HMI) STMPE610 I2C touch screen controller Utilisé dans : Industrial Human-Machine Interface (HMI) EFR32MG24A020F1024IM48 Matter/Zigbee/Thread radio co-processor Utilisé dans : Home Automation and Security Hub ATECC608B-MAHCZ-T Secure element for authentication and key storage Utilisé dans : Home Automation and Security Hub MAX30102 Pulse oximeter and heart-rate sensor module Utilisé dans : Portable Medical and Health Monitoring Devices AFE4400 Analog front-end for heart-rate oximetry Utilisé dans : Portable Medical and Health Monitoring Devices
What is the operating voltage of the EFM32GG230F1024G-E-QFN64?
The EFM32GG230F1024G-E-QFN64 operates across a supply voltage range of 1.85 V to 3.8 V, making it compatible with single-cell Li-ion, 3 V coin cells, and 3.3 V industrial rails. According to the Silicon Labs EFM32 Giant Gecko datasheet, the device includes integrated LDO and DC-DC regulators so designers can select the most efficient internal rail topology for their application.
How much Flash and RAM does the EFM32GG230F1024G-E-QFN64 have?
The EFM32GG230F1024G-E-QFN64 integrates 1 MB (1024 KB) of Flash program memory and 128 KB of RAM. The 'F1024' suffix in the part number directly indicates the 1024 KB Flash capacity. Per the Silicon Labs Giant Gecko datasheet, this memory configuration supports rich application code with substantial RAM for protocol stacks like Bluetooth, Thread, or Matter.
What is the maximum clock speed of the EFM32GG230F1024G-E-QFN64?
The EFM32GG230F1024G-E-QFN64 runs the ARM Cortex-M3 core at up to 48 MHz. This is the standard performance tier of the Giant Gecko family. According to the Silicon Labs datasheet, the core uses a 1.25 DMIPS/MHz Cortex-M3 implementation, yielding approximately 60 DMIPS peak performance for embedded control and IoT workloads.
How much current does the EFM32GG230F1024G-E-QFN64 draw in sleep mode?
The EFM32GG230F1024G-E-QFN64 consumes as little as 0.6 uA in Stop mode (EM2 Deep Sleep with RTC running) and 180 uA/MHz in Run mode. According to the Silicon Labs EFM32GG datasheet, this ultra-low sleep current enables battery lifetimes of 5-10 years on a single CR2032 coin cell in typical IoT sensor duty-cycled applications.
Where can I buy the EFM32GG230F1024G-E-QFN64?
The EFM32GG230F1024G-E-QFN64 is available from authorized distributors including DigiKey (8 units in stock as of 2026-10-21), Mouser, Octopart, and Symmetry Electronics. Pricing starts at $15.25 USD for qty 1 and drops to $9.60 USD per unit at 1000-piece quantities. Stock is constrained at distribution; lead time may apply for volume orders.
What is the price of EFM32GG230F1024G-E-QFN64?
Pricing for the EFM32GG230F1024G-E-QFN64 (as of 2026-10-21) starts at $15.25 USD per unit at qty 1, scaling down to $13.78 at qty 10, $12.20 at qty 100, $10.85 at qty 500, and $9.60 at qty 1000 pieces. These prices reflect DigiKey listing data; volume quotes should be requested directly from Silicon Labs for OEM production quantities.
What is the lead time for EFM32GG230F1024G-E-QFN64?
Distribution stock for the EFM32GG230F1024G-E-QFN64 is currently limited to 8 pieces at DigiKey as of 2026-10-21. Lead time for replenishment typically runs 8-14 weeks from Silicon Labs for volume orders. Designers evaluating alternatives in case of stock shortage should consider the EFM32GG230F1024-QFN64T (tape and reel variant) or the EFM32GG230F1024G-E-QFN64R (reel packaging) for direct drop-in options.
Is the EFM32GG230F1024G-E-QFN64 in stock?
Stock status for the EFM32GG230F1024G-E-QFN64 (as of 2026-10-21) is limited - DigiKey lists 8 units in inventory, and Mouser and Octopart show similar low stock levels. For prototype quantities the device is available; for production volumes, planners should contact Silicon Labs directly to confirm allocation or consider ordering the tape-and-reel EFM32GG230F1024-QFN64T variant which shares the same die and package.
What is the difference between EFM32GG230F1024G-E-QFN64 and EFM32GG230F1024-QFN64T?
The EFM32GG230F1024G-E-QFN64 includes the 'G-E' suffix denoting lead-free (Pb-free) matte tin lead finish plus an exposed pad (EP), while the EFM32GG230F1024-QFN64T variant adds a 'T' suffix indicating tape-and-reel packaging. Both parts share the same 64-QFN (9x9 mm) package, same die, same 1 MB Flash, 128 KB RAM, and 48 MHz Cortex-M3 core - they are drop-in replacements of each other.
When should I choose EFM32GG230F1024G-E-QFN64 over EFM32LG230F256G-E-QFN64?
Choose the EFM32GG230F1024G-E-QFN64 (Giant Gecko) when you need the full 1 MB Flash and 128 KB RAM ceiling, AES hardware encryption, and 48 MHz Cortex-M3 performance. Choose the EFM32LG230F256G-E-QFN64 (Leopard Gecko) when your application fits within 256 KB Flash and 32 KB RAM at a lower unit cost. Both share the 64-QFN footprint but the GG outperforms by ~4x in memory headroom.
What is the best drop-in replacement for the EFM32GG230F1024G-E-QFN64?
The best drop-in replacements for the EFM32GG230F1024G-E-QFN64 are same-family EFM32GG230 variants in the same 64-QFN (9x9) package, namely the EFM32GG230F1024-QFN64T (tape-and-reel variant of the same die) and the EFM32GG230F1024G-E-QFN64R (reel-packaged variant). Both share identical silicon, pinout, and electrical specifications - they are mechanically and electrically interchangeable.
Can EFM32GG230F1024G-E-QFN64R replace EFM32GG230F1024G-E-QFN64?
Yes, the EFM32GG230F1024G-E-QFN64R is a direct drop-in replacement for the EFM32GG230F1024G-E-QFN64. The only difference is the 'R' packaging suffix, which denotes a reel packing format rather than tray. Both parts use the same 64-QFN (9x9 mm) Giant Gecko die with identical 1 MB Flash, 128 KB RAM, and 48 MHz Cortex-M3 specifications - the pinout is identical.
Where can I download the EFM32GG230F1024G-E-QFN64 datasheet PDF?
The manufacturer datasheet for the EFM32GG230F1024G-E-QFN64 is available as the EFM32 Giant Gecko Family datasheet at https://www.silabs.com/documents/public/data-sheets/efm32gg-datasheet.pdf on the Silicon Labs website. The companion EFM32GG Reference Manual at https://www.silabs.com/documents/public/reference-manuals/EFM32GG-RM.pdf provides full peripheral register descriptions. Both are required for board bring-up and firmware development.
What is the pinout configuration of the EFM32GG230F1024G-E-QFN64?
The EFM32GG230F1024G-E-QFN64 uses a 64-pin QFN package (9x9 mm) with an exposed thermal pad that must be soldered to the ground plane. Per the Silicon Labs datasheet, the package provides 56 usable GPIO pins plus dedicated supply, ground, reset, and debug (SWD) pins. Pin 1 is identified by the package marker dot. Refer to the official QFN-64 pinout diagram in the datasheet for exact signal mapping.
What are the key specifications of the EFM32GG230F1024G-E-QFN64 that engineers should know?
Key specifications: ARM Cortex-M3 core at 48 MHz, 1 MB Flash, 128 KB RAM, 56 GPIO, 1.85-3.8 V supply, 0.6 uA EM2 sleep current, 180 uA/MHz run current, AES-128/256 hardware encryption, integrated LDO and DC-DC, 64-pin QFN (9x9 mm) with exposed pad, RoHS compliant. The part is part of the Giant Gecko low-energy MCU family designed for years-of-battery-life IoT and industrial sensor applications.

Données de référence d'ingénierie pour EFM32GG230F1024G-E-QFN64 — comparaison, recommandations de conception et informations de conformité.

Guide de sélection

Choose the EFM32GG230F1024G-E-QFN64 when designing battery-powered IoT endpoints, smart meters, or wireless sensor gateways that require full 1 MB Flash headroom for protocol stacks (BLE, Zigbee, Matter), hardware AES acceleration for secure connectivity, and years of battery life enabled by the 0.6 uA EM2 sleep current. Choose the EFM32GG230F1024-QFN64T or EFM32GG230F1024G-E-QFN64R drop-in variants when only the packaging format (tape-and-reel vs tray) differs - all three share identical silicon. Choose the EFM32GG230F512G-E-QFN64 when your firmware fits within 512 KB Flash and you want roughly 10-20% unit cost savings, accepting reduced code/data headroom. All Giant Gecko family variants share the 64-QFN footprint, so PCB layout and pinout remain compatible across the family. Avoid this part only if you need Cortex-M4/M33 cores - migrate to the EFM32TG11 or EFR32xG22 families for newer architectures.

Comparaison avec les alternatives

Paramètre Ce produit EFM32GG230F1024-QFN64T EFM32GG230F1024G-E-QFN64R EFM32GG230F512G-E-QFN64
Brand Silicon Labs Silicon Labs Silicon Labs Silicon Labs
Package 64-QFN (9x9) 64-QFN (9x9) - same 64-QFN (9x9) - same 64-QFN (9x9) - same
Core ARM Cortex-M3 ARM Cortex-M3 ARM Cortex-M3 ARM Cortex-M3
Max Clock Speed 48 MHz 48 MHz 48 MHz 48 MHz
Flash Size 1024 KB 1024 KB 1024 KB 512 KB
RAM Size 128 KB 128 KB 128 KB 128 KB
Supply Voltage 1.85 V to 3.8 V 1.85 V to 3.8 V 1.85 V to 3.8 V 1.85 V to 3.8 V
AES Hardware Crypto Yes Yes Yes Yes
EM2 Sleep Current 0.6 uA 0.6 uA 0.6 uA 0.6 uA
Packaging Tray (G-E suffix) Tape and Reel (-T suffix) Reel (-R suffix) Tray

Principaux différenciateurs

  • 1 MB Flash and 128 KB RAM in 64-QFN (vs EFM32LG230F256G-E-QFN64)
  • On-chip AES-128/256 hardware encryption (vs EFM32G230F64 (Gecko, no AES variant))
  • 0.6 uA EM2 Deep Sleep current (vs STM32F103 Cortex-M3 MCU)

Notes de conception

Place one 1 uF X7R ceramic bypass capacitor as close as possible to every VDD pin and a single 4.7 uF X5R bulk capacitor near the package. Estimated: with the device running at 48 MHz and typical workload, average supply current is around 8-10 mA active. Select the internal LDO mode for 1.85-3.0 V supply rails and the DC-DC mode for supply voltages above 2.4 V to maximize efficiency. For battery applications drawing microamp sleep currents, disable all unused peripherals and route GPIO to a defined state to eliminate parasitic current paths.

The QFN-64 package includes an exposed thermal pad on the bottom that serves as the primary ground connection and thermal dissipation path. This pad must be soldered to a continuous ground pour on the top layer with at least 9 thermal vias (0.3 mm drill, 0.6 mm pad) arrayed in a 3x3 grid to inner ground planes. Without proper soldered thermal pad attachment, the device may exceed junction temperature ratings (typical theta_JA is 28 C/W on a 4-layer JEDEC JESD51-7 board). Keep high-speed traces (SWD, SPI, external bus) length-matched within 5 mm to avoid signal skew.

Common pitfalls: (1) Forgetting to solder the exposed thermal pad causes intermittent resets under load due to poor ground return; (2) leaving SWDIO/SWCLK pins floating during boot may put the MCU in a debug state - add 100 kohm pull-downs if unused; (3) selecting the wrong memory model in the linker script causes the bootloader region to be overwritten - reserve the top 16 KB of Flash for the bootloader; (4) enabling peripheral interrupts before configuring nested vectored interrupt controller (NVIC) priorities leads to hard faults - configure priorities first. Refer to Silicon Labs application note AN0003 for the recommended bring-up sequence.

Informations de conformité

RoHS
Compliant
REACH
Compliant
AEC-Q100
Not Qualified
Sans plomb
Yes
Sans halogène
Unknown
Minerais de conflit
Compliant

RoHS and lead-free status inferred from 'G-E' suffix in MPN indicating lead-free matte tin finish; consult manufacturer datasheet for full compliance certificates.

Données vérifiées le : 2026-10-21 — données vérifiées et sélectionnées par l'équipe d'ingénierie composants de XAIPART

Recherches associées

EFM32GG230F1024G-E-QFN64 EFM32GG230F1024G-E-QFN64 datasheet Silicon Labs EFM32GG230F1024G-E-QFN64 ARM Cortex-M3 48MHz 1MB Flash MCU 64-QFN 9x9 microcontroller low power EFM32GG230F1024G-E-QFN64 IoT battery powered EFM32GG230F1024G-E-QFN64 vs EFM32LG230F256G-E-QFN64 EFM32GG230F1024G-E-QFN64 replacement EFM32GG230F1024G-E-QFN64 buy price Giant Gecko microcontroller sleep current 0.6uA EFM32GG230F1024G-E-QFN64 pinout QFN64 low energy 32-bit MCU AES hardware encryption EFM32GG230F1024G-E-QFN64 lead time stock

Composants & termes associés

Silicon Labs EFM32GG230F1024G-E-QFN64 EFM32GG230F1024-QFN64T EFM32GG230F1024G-E-QFN64R EFM32GG230F512G-E-QFN64 ARM Cortex-M3 32-bit microcontroller low-power MCU Giant Gecko family EFM32 Giant Gecko VQFN-64 QFN-64 package family embedded microcontroller semiconductor IC AES-128 AES-256 hardware encryption RoHS compliant deep sleep mode EM2 energy mode IoT sensor node sub-GHz radio BLE protocol stack Thread/Matter protocol Zigbee protocol
RFQ de devis rapide
Renseignez tous les détails — notre équipe commerciale vous répondra sous 24 heures
Référence Fabricant Boîtier QTY Prix cible Total
Total : 0,00 USD
✓
Devis envoyé !

Nous répondrons à votre e-mail sous 24 heures

1
RFQ envoyée
2
Devis reçu
3
Commande passée
4
Paiement
5
Expédié
6
Livré
Voir les détails de la RFQ