Slow Mo Guys sent the 168 TB Seagate LaCie Thunderbolt 3-based RAID system they had purchased in 2019 to store video footage to Mend it Mark after it failed to power on. Seagate’s support also proved inadequate for the product, whose warranty had expired and whose production had ended. An inspection found that all four fans were nonfunctional and that only a few LEDs on the mainboard were lit. Tests conducted without a schematic determined that not all of the power rails were being activated. The problem appeared to stem from the NXP LPC11U6x-series microcontroller responsible for power management waiting for signals from the fan controllers on the SATA backplane. The RAID began working again after the faulty SI2319 or a similar SOT-23-packaged P-channel MOSFET was replaced. Testing with hard drives is pending; the repair’s material cost was low.
Why it matters
This repair shows that, in high-capacity storage systems whose warranties have expired and whose production has ended, the problem is not always to be found in the hard drives. The fact that the failure was caused by an unexpected interaction between power management and fan control highlights the importance of manufacturer support and technical schematics when diagnosing similar devices. The ability to carry out the intervention at a low material cost brings component-level repair into consideration instead of completely replacing the device; this is particularly relevant to users who rely on a single storage system to access large video archives. However, since testing the RAID with hard drives has not yet been completed, it remains unclear whether the data will be accessible and whether the system will operate stably under load.