[EFM-OAM] comment 586
I have a question on Comment #586 made against D2.0. The comment was
against 57.3.1.5, and the values of the two timers defined there. Both
timers are defined as having nominal values of 5s and 1s respectively, but
no ranges are provided for the timers. The comment requested that a range
be provided for each of them.
The response was to reject, stating that sometimes timers are described as
having just nominal values. Examples in 23.2.4.3, 32.3.4.2, 36.2.5.1.7, and
40.3.3.3 were provided.
Subclause 23.2.4.3 (100BASE-T4 PHY) defines a timer that governs the
transitions between states in a state machine. It seems that this 40ns timer
is based off of the defined TX_CLK of 25MHz, which does have a tolerance
associated with it. Subclause 32.3.4.2, 40.3.3.3, and 36.2.5.1.7 have
similarly defined timers for 100BASE-T2, 1000BASE-T, and 1000BASE-X
respectively.
In each of these examples, it appears that there is a clear range over which
the timer should exist, and therefore a clear measurement range.
For the timers in Clause 57, I can see so such range. Also, we're talking
about timers on the order of seconds and not nanoseconds. I would like to
have a range such that I can test that the device properly implements these
timers. For example, look at A_timer defined in 40.4.5.2. This is a 1.3
second timer with a range of +/- 25%. I would expect something similar for
the timers in Clause 57.
If the timers remain the same and do not get a range, then from a compliance
testing stand point, what does a nominal duration of 1 or 5 seconds really
mean?
Eric Lynskey
InterOperability Laboratory
University of New Hampshire