Re: [BP] 802.3ap D3.0 comments
howard,
I have no strong feelings about normative vs informative channel
spec, but link budget closure is consider vital. I intend to submit
a technical (not TR) like:
" I do not feel comfortable with our ICR specification. While it
is could work as stated i do not like the fact that the basic equation
assumes the thru channel, victim and aggressor transmitters are better
than minimum spec, and only applies in general if corrections are added.
Possible modifications could be:
1. Remove equations 69B-24 and 69B-25, the paragraphs explaining
them, beginning at page 194, line 36 and ending page 195 line 18, and
table 69B-2. Replace equation 69B-26 with:
ICRfit = 23.3 - 18.7log(f/5 GHz)
(Assuming a maximum value of 3dB for PILD. The 23.3 value may change
if this assumption is wrong.)
2. Remove equations 69B-24 and 69B-25, the paragraphs explaining
them, beginning at page 194, line 36 and ending page 195 line 18, and
table 69B-2. Replace equation 69B-26 with:
ICRfit = 23.3 - 18.7log(f/5 GHz) + Bsys
add:
"If the system designer has no assurance that transmitter
variability any better than specified under the appropriate port type
transmitter specification and no assurance that the receiver
interference tolerance will be any better than specified for the
appropriate port receiver specification, he should a system bonus (Bsys)
of 0. If better than specified parts will always be used compute
Bsys as:
Bsys = 20*log10 ((minimum transmitter amplitude to be used/
maximum transmitter amplitude to be used)/(
minimum transmitter amplitude allowed by spec/
maximum transmitter amplitude allowed by spec)) +
20*log10 (minimum expected interference tolerance/
specified interference tolerance)
3*log10((minimum transmitter rise time to be used/
maximum transmitter rise time to be used)/
(minimum transmitter rise time allowed by spec/
maximum transmitter rise time allowed by spec))
"
3. Rename 60B4.6 "Interfernece"
Change the first paragraph to:
"In order to limit interference at TP4, the differential
crosstalk due to near-end and far-end aggressors and self interference
are specified to meet the BER objective defined in 69.1.2."
add a new paragraph "Self interfernece"
"The self interference due to through channel irregularities at TP4 is
calculated with the equation:
SI(f) = 14.3-10*log10 ( 1.6* ILD(f) ^2)
Change Equation 69B-17 to
PSXT = -10log(10 ^(-PSNEXT/10) + 10 ^(-PSFEXT/10) + 10 ^(-SI/10))
Remove equations 69B-24 and 69B-25, the paragraphs explaining
them, beginning at page 194, line 36 and ending page 195 line 18, and
table 69B-2. Replace equation 69B-26 with:
ICRfit = 20.3 - 18.7log(f/5 GHz) + Bsys
add:
"If the system designer has no assurance that transmitter
variability is any better than specified for the appropriate port type
transmitter and no assurance that the receiver interference tolerance
will be any better than specified for the appropriate port receiver, he
should a system bonus (Bsys) of 0. If better than specified parts will
always be used compute Bsys as:
Bsys = 20*log10 ((minimum trnasmitter amplitude to be used/
maximum trnasmitter amplitude to be used)/(
minimum transmitter amplitude allowed by spec/
maximum transmitter amplitude allowed by spec)) +
20*log10 (minimum expected interference tolerance/
specified interference tolerance)
3*log10((minimum transmitter rise time to be used/
maximum transmitter rise time to be used)/
(minimum transmitter rise time allowed by spec/
maximum transmitter rise time allowed by spec))
"
"
charles
|--------------------------------------------------------------------|
| Charles Moore
| Avago Technologies
| ISD
| charles.moore@avagotech.com
| (970) 288-4561
|--------------------------------------------------------------------|
Howard A. Baumer wrote:
> I anticipate that the initial sponsor ballot on 802.3ap
> will produce comments related to the need for a normative
> channel model, and link budget closure. If you are interested
> in these topics, and would like to participate in the
> development of a proposal to address them, please contact
> me. If we work on this proactively, our upcoming comment
> resolution meeting in Knoxville will be more productive.
>
> Howard Baumer