Bao'An Shenzhen China

nRF51 Crack | Protection Levels & Firmware Recovery

Oct 26, 2026  /  PCB COPY

nrf51 crack: nRF51822 chip on PCB with SWD debug pads for firmware access

What Does an nRF51 Crack Actually Involve?

An nRF51 crack is the process of bypassing or circumventing the readback protection on Nordic Semiconductor’s nRF51 series (nRF51822, nRF51422, nRF51824) to extract the firmware stored in its 128 KB or 256 KB flash. The nRF51 relies on a single configuration register called RBPCONF (Readback Protection Configuration) with two protection levels. When protection is disabled, firmware can be read directly over SWD. When it is enabled, the debug port locks out flash reads, and specialized techniques—glitching, side-channel analysis, or silicon-level probing—are required to recover the binary. Our engineers perform this work only under proof-of-ownership or written authorization for legitimate purposes such as legacy product maintenance, lost-source recovery, or failure analysis.

How RBPCONF Protection Levels Work on the nRF51

nrf51 crack: nRF51822 chip on PCB with SWD debug pads for firmware access

Unlike the three-tier RDP system on STM32 families, the nRF51 keeps things simpler—but that simplicity has its own nuances. The RBPCONF register at address 0x10001004 in the UICR (User Information Configuration Registers) holds two 8-bit fields that control protection behavior.

Field Value Protection State What Changes
PALL (bits 15:8) 0xFF Disabled Full SWD read/write access to flash and RAM. Firmware can be dumped freely.
PALL 0x00 Enabled SWD read of flash and RAM blocked. Debug port remains active for erase-only operations.
PR0 (bits 7:0) 0xFF Disabled Region 0 (code region 0, configurable size) unprotected.
PR0 0x00 Enabled Region 0 protected even from application-level writes, adding a secondary guard against self-modification.

When PALL = 0x00, the only operation the debugger can perform is an ERASEALL command, which wipes the entire flash including UICR—resetting the protection but destroying the firmware in the process. This is the fundamental barrier an nRF51 crack must overcome: extracting the firmware without triggering or relying on the erase path.

Why the nRF51 Protection Is Considered Moderate

The nRF51 was built on an ARM Cortex-M0 core manufactured on a 180 nm process (TSMC). Several factors make its protection moderate rather than strong:

  • No hardware-enforced TrustZone or secure enclave. Protection depends entirely on the RBPCONF register and the debug access port (DAP) gating logic.
  • Single fuse-bit behavior. PALL is a single on/off gate—there is no intermediate “warning” level like STM32’s RDP Level 1 that allows partial debug while blocking flash reads.
  • Known voltage-glitch susceptibility. Published research (including LimitedResults, 2018–2020) demonstrated that a precisely timed voltage glitch on the VDD rail during the DAP initialization sequence can bypass the PALL check, re-enabling full SWD access for the duration of that session.
  • 180 nm process geometry. Larger transistor geometries are inherently more susceptible to fault injection than 40 nm or 28 nm nodes used in newer parts like the nRF52 series with its enhanced APPROTECT.

This does not mean every nRF51 crack succeeds. Success depends on the specific silicon revision, board-level decoupling, and whether the SoftDevice (Nordic’s BLE stack) includes additional runtime integrity checks.

Step-by-Step: How Our Engineers Approach nRF51 Firmware Recovery

1. Initial Assessment (Day 1–2)

We connect to the target board via SWD (SWDIO + SWCLK) using a J-Link or custom probe. The first step is reading the RBPCONF register. If PALL = 0xFF, the firmware is simply dumped—no crack needed. If PALL = 0x00, we proceed to fault injection preparation.

2. Board Preparation

We desolder or isolate bulk decoupling capacitors near the nRF51’s VDD pins. The goal is to reduce the power rail’s ability to absorb a fast voltage transient. On boards with heavy filtering (multiple 100 nF + 10 µF caps in parallel), this step is critical—without it, the glitch pulse is absorbed before reaching the die.

3. Voltage Glitch Injection (Day 2–5)

Using a ChipWhisperer or equivalent glitch platform, we inject a brief (10–50 ns) voltage drop on VDD, timed to coincide with the DAP’s protection-check window during the SWD connection handshake. Parameters we sweep:

  • Glitch width: 8 ns to 60 ns in 2 ns steps
  • Glitch offset: relative to the SWD JTAG-to-SWD switch sequence, typically 0–500 µs window
  • Voltage depth: dropping VDD from 3.0 V to 0.8–1.5 V momentarily
  • Repetitions: 50,000–500,000 attempts per parameter set

When the glitch lands correctly, the PALL check evaluates as “disabled” for that session. We immediately issue a mass flash read over SWD and capture the full binary.

4. Verification & Delivery (Day 5–10)

The extracted binary is verified against the flash size declared in the FICR (Factory Information Configuration Registers). We check CRC consistency, identify the SoftDevice version (S110, S120, S130, or none), and separate the SoftDevice blob from the application firmware. Deliverables include the raw .hex/.bin, a memory map, and—if requested—a disassembly with peripheral register annotations.

What Can Go Wrong

We are transparent about failure modes because honest constraint disclosure matters more than overpromising.

Failure Mode Likelihood Mitigation
Die damage from repeated glitching Low (~2–5% of jobs) We limit total glitch count and monitor current draw to detect latch-up early.
Board-level filtering too aggressive to remove Moderate on multi-layer BGA modules We can depackage and probe the die directly, adding cost and 3–5 extra days.
Newer silicon revisions with tighter timing Occasional (rev. 3 silicon) We adjust glitch parameters; some revisions require electromagnetic fault injection (EMFI) instead of voltage glitching.
Corrupted flash (worn-out cells) Rare Partial recovery possible; we report which sectors are unreadable.

If we determine during the feasibility check that recovery is unlikely, we report that before any charges apply.

nRF51 vs. nRF52: Why the Generation Matters

Voltage glitch setup connected to a BLE board for nRF51 firmware recovery

Engineers sometimes assume the same approach works across Nordic’s lineup. It does not. The nRF52 introduced APPROTECT, a hardware-level protection mechanism that is significantly harder to bypass than RBPCONF. Key differences:

Feature nRF51 (Cortex-M0, 180 nm) nRF52 (Cortex-M4F, 55 nm)
Protection register RBPCONF (UICR) APPROTECT (UICR) + hardware lock
Debug port behavior when locked Erase-only via ERASEALL DAP completely disabled on some revisions
Voltage glitch susceptibility High (180 nm) Lower (55 nm, faster edge recovery)
Typical recovery turnaround 5–10 working days 10–20 working days
Estimated success rate range High Moderate, revision-dependent

If your target is actually an nRF52832 or nRF52840, see our dedicated nRF52 firmware recovery page for specifics on APPROTECT bypass.

Common Use Cases for Authorized nRF51 Firmware Recovery

  • Legacy BLE product maintenance: The nRF51822 shipped in millions of Bluetooth Low Energy devices from 2013–2018. Many original firmware repositories have been lost or the developer is no longer available.
  • Prototype recovery: A single prototype board with firmware that was never backed up—common in startup environments.
  • Failure analysis: Comparing firmware from a failing unit against a known-good binary to identify corruption or unauthorized modification.
  • Second-source qualification: Verifying that a contract manufacturer programmed the correct firmware version before shipping.

All work requires proof of ownership or written authorization. We document the chain of custody and operate under NDA. For details on our legal and ethical framework, visit our STM8L recovery authorization process, which follows the same compliance structure.

Related Chip Families We Support

If you are working across multiple platforms in the same product line, our lab handles a wide range of MCU families beyond Nordic. For Wi-Fi/BLE combo SoCs, we also perform ESP32 firmware extraction and ESP8266 flash readout. For STM32-based designs with different protection tiers, our STM32L1 recovery service covers the ultra-low-power family.

Can the nRF51 firmware be read without any special tools if protection is off?

Yes. If RBPCONF PALL = 0xFF, any standard SWD debugger (J-Link, ST-Link in SWD mode, or a DAPLink probe) can read the full flash contents directly. No glitching or invasive techniques are needed. Simply connect SWDIO, SWCLK, VDD, and GND, then use nrfjprog or OpenOCD to dump the binary.

Will a voltage glitch erase my nRF51’s firmware?

No—a correctly executed voltage glitch temporarily bypasses the protection check without triggering the ERASEALL command. The glitch targets the DAP authentication logic, not the flash controller. However, an improperly timed or excessively deep glitch can cause a latch-up condition that may damage the die, which is why this work should be performed by experienced operators with current-monitoring equipment.

How long does an nRF51 crack typically take?

For a standard nRF51822 with PALL = 0x00 on a board with accessible SWD pads and removable decoupling, typical turnaround is 5–10 working days including verification. Boards with heavy filtering, conformal coating, or BGA-only test points may add 3–5 days for preparation work.

Is nRF51 firmware recovery legal?

Firmware recovery is legal when performed on devices you own or are authorized to service. Common lawful scenarios include recovering your own lost source code, maintaining legacy products, and conducting failure analysis. We require proof of ownership or written authorization before starting any job. Unauthorized extraction of third-party intellectual property is something we do not perform.

What deliverables do I receive after an nRF51 crack?

Standard deliverables include the raw firmware binary (.hex and .bin formats), a flash memory map identifying SoftDevice boundaries and application regions, and a summary report noting the SoftDevice version detected (if any). Optional add-ons include annotated disassembly and peripheral configuration extraction.

Check feasibility for your project

Send the chip marking or board details and tell us what you need to recover. Add clear photos if available.

Request a feasibility review

Working on a board like this?

Send the chip marking or two photos. You get feasibility, lead time and price within 24 hours, and the check costs nothing.

Get a free quote

Related reading

WhatsApp Send board details