Board bring-up checklist for first prototype boards
Twenty-one checks in the order to run them on a new board, from a resistance check before power to a bring-up log the hardware team can act on.
Key characteristics
- Items
- 21 in five groups
- Groups
- Before power, first power, debug and clocks, peripherals, record
- Output
- Ticks on screen, CSV download, print
- Runs
- In your browser with nothing sent
Checklist: Board bring-up checklist for first prototype boards
New boards arrive with two kinds of problems mixed together: hardware faults, and firmware that has never met this hardware. Bring-up is the work of separating them one layer at a time, so that at the end you know which problems belong to whom. This checklist puts the layers in order. Each group builds on the one before, so a failure stops you at the cheapest point to find it.
How to use it
Run it per board, not per batch. The first board of a revision gets every item; later boards get at least the power and debug groups. Tick items as you confirm them, write the measured values in your bring-up log as you go, and download the CSV at the end as a record of what was checked. Ticks are not saved if you leave the page, so download before you close it.
The five groups
Before power
Four checks with the board unpowered: schematic, layout files and bill of materials at the same revision as the board; a visual check for missing, rotated or bridged parts under magnification; the resistance from each supply rail to ground, measured and recorded; and a current-limited bench supply. A rail that measures close to zero ohms is a short you have found before it found something to burn. For first power, set the supply’s current limit a little above what the board should draw with no firmware running, which is usually well below its full operating draw, and raise it once the rails check out.
First power
Each rail measured at its test point and within tolerance, the idle current recorded and compared with the estimate, nothing hot to the touch or under a thermal camera, and the reset line and power-good signals behaving as designed. An idle current far above the estimate is the earliest sign of a part in the wrong state, and it is far easier to chase now than once the firmware is running.
Debug and clocks
The debug probe connects and reads the chip ID, a minimal program runs from flash and toggles a pin, the main and low-frequency clocks are measured rather than assumed, and a serial log or RTT output (a log channel through the debug probe) works. Measuring the clocks matters: a low-frequency crystal that runs slightly off, or not at all, shows up weeks later as drifting timestamps or a device that never wakes from sleep.
Peripherals
Every I2C and SPI device answers at its expected address, ADC readings match a known voltage, the radio sends and receives at short range, flash erase, write and read work across the whole part, and each GPIO on the schematic is exercised once. The full flash test is the one teams skip and later regret: an update path will use every sector eventually.
Record and hand back
A bring-up log with measurements and board serial numbers, hardware issues listed with evidence for the hardware team, rework instructions for any board change, and the pin map and clock tree checked into the firmware repository. This record is what the hardware team acts on before the next revision.
Reading the example
Example (fictional). Loamwick’s Rev C prototype boards were brought up in weeks 1 and 2, right after the firmware review. The first Rev C board went through all 21 items, and the rest went through the power and debug groups. The bring-up log, with idle current per board and the measured low-frequency clock, became the starting point for the sleep current in the battery life calculator, and the hardware issue list went into the Rev D trial boards.
Limits
- It is the firmware side of bring-up. It does not replace the hardware team’s own checks of layout, signal integrity or thermal design.
- It assumes one microcontroller. Boards with several processors, a cellular module or an FPGA need a group per device.
- It stops at “works on the bench”. Temperature, battery and radio range testing belong to the gates in the production readiness checklist.
- Safety comes first. Boards with mains voltage, large batteries or motors need their own safety procedure before any of this.
What to do next
Our board bring-up service runs this list with your hardware team and hands back the log, the issue list and a board support package, and a bring-up sprint is one of the engagements you can start with. How we work shows where bring-up sits between review and build. More in the prototype to production section of Field Update.
Send us your board and your plan.
A firmware review reads your schematics, code, flash layout and update plan, and ends with a written plan: what to build first, what could go wrong in the field and how updates will reach your devices.
Or email sales@fumbar.com, or call (720) 675-4202.