Occurred when self-tests were enabled, resulting in a longer start-up time for aXiom.
This problem actually only happened when using TH2 4.7.3.1 (or later). TH2 polls u34 for the u01 "hello" report after a reset to indicate that aXiom has come back up again. Asking the bridge to read aXiom when it wasn't ready (nIRQ wasn't asserted) resulted in the SCL line being held low by the AXPB009; likely because the comms timed out when aXiom was clock stretching.
Occurred when self-tests were enabled, resulting in a longer start-up time for aXiom.
This problem actually only happened when using TH2 4.7.3.1 (or later). TH2 polls u34 for the u01 "hello" report after a reset to indicate that aXiom has come back up again. Asking the bridge to read aXiom when it wasn't ready (nIRQ wasn't asserted) resulted in the SCL line being held low by the AXPB009; likely because the comms timed out when aXiom was clock stretching.