This issue is following the discussion start in #53
In summary, it appears that atmega subordinate devices are not synchronizing to the clock generator device and instead add a offset in the order of 200 microseconds.
This bug can be demonstrated by measuring the clock serial output from a clock generator (example 1 a atmega generator, TimestampGeneratorGen3 and example 2 a rp2040 generator, WhiteRabbit) and the corresponding heartbeat LED toggle of a subordinate device (in this case a atmega behavior board)
Example 1)
Example 2)
Under a perfect scenario, if the two board were to be perfectly synchronized (and the led heartbeat routine is instantaneous), one would expect the led toggle to take place exactly 672 after the start of the last byte, as per harp standard (https://github.com/harp-tech/protocol/blob/main/SynchronizationClock.md#serial-configuration)
This issue is following the discussion start in #53
In summary, it appears that atmega subordinate devices are not synchronizing to the clock generator device and instead add a offset in the order of 200 microseconds.
This bug can be demonstrated by measuring the clock serial output from a clock generator (example 1 a atmega generator,
TimestampGeneratorGen3and example 2 a rp2040 generator,WhiteRabbit) and the corresponding heartbeat LED toggle of a subordinate device (in this case a atmegabehavior board)Example 1)
Example 2)
Under a perfect scenario, if the two board were to be perfectly synchronized (and the led heartbeat routine is instantaneous), one would expect the led toggle to take place exactly 672 after the start of the last byte, as per harp standard (https://github.com/harp-tech/protocol/blob/main/SynchronizationClock.md#serial-configuration)