Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

[802.3ae_Serial] D4.3 comments



Attached are my final comments to D4.3. Both CSV and HTM formats are included.
 
Note, I have added another controversial one at the end. I created this after cleaning up my power budget review this morning after the con-call. It proposes reducing max TDP for -L to 3.0 dB. Please review.

Tom
Stratos NW
(425) 672-8035 x105

 
BEGIN:VCARD
VERSION:2.1
N:Lindsay;Tom
FN:Tom Lindsay
ORG:Stratos Lightwave;Opto NW
TITLE:Prin. Engr.
NOTE;ENCODING=QUOTED-PRINTABLE:Home: lindsayta@xxxxxxx=0D=0A=0D=0Aom
TEL;WORK;VOICE:(425) 672-8035 x105
TEL;HOME;VOICE:(425) 775-7013
TEL;CELL;VOICE:(206) 790-3240
TEL;WORK;FAX:(425) 672-8014
ADR;WORK;ENCODING=QUOTED-PRINTABLE:;;6100 219th Street SW=0D=0ASuite 520;Mountlake Terrace;WA;98043
LABEL;WORK;ENCODING=QUOTED-PRINTABLE:6100 219th Street SW=0D=0ASuite 520=0D=0AMountlake Terrace, WA 98043
EMAIL;PREF;INTERNET:tlindsay@xxxxxxxxxxxxxxxxxxxx
REV:20020219T195001Z
END:VCARD

P802.3ae-comments_D42.csv

CalcTextForExport
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.2~479~26~T~The present description for eye closure contradicts itself. On one hand, we call out minimal noise and data dependent effects, yet on the other hand, we limit the amounts of sine jitter and sine interference. If one is succesful on the first hand, they will not be able to achieve the required stress values because of the limitations of the second hand.

Relatedly, there is valid concern that allowing more sine interference will unduly stress receivers, as this is not a type of degradation found in real systems.~1. Page 480, change line 26. "...interference, the majority of the vertical..."

2. Move the short paragraph from Page 481, line 3, and insert it at Page 480, line 37 (still separate paragraph).

3. Modify the paragraph just moved to "...at least 5 psec but no more than 15 psec of pulse shrinkage jitter should be achieved. This imposes a limit of less than 1.2 dB of vertical closure from sinusoidal interference, applied after vertical closure created by filtering."

4. Page 480, modify starting at line 45. "Iterate the filter bandwidth and the settings for sinusoidal interference and/or jitter until all constraints are met including jitter (J), vertical closure (VECP), and pulse shrinkage jitter, and that sinusoidal...".
~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~Table 52-18~469~1~T~Table 52-18 specifies 0.5 dB for additional allowable insertion loss. With the new TDP method, this is no longer appropriate.

If it is correct as/where it is, then the min Rx power in Table 52-17 must be decreased by this amount.~Set addtional insertion loss allowed to 0 dB and set allocation for penalties to 4.1 dB.~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.8.XXXX~476~38~E~Clause does not appear to be as removed as it says it is...~Remove.~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.1~479~13~E~Clarification that residual jitter should be minimized.~Modify end of sentence with "...all sources is unavoidable, but should be less than 0.25 UI pk-pk..."~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.1~479~~E~Clean up, strengthen descriptions.~1. Change the first part of the paragraph starting at line 15 (mostly per Piers). "The test pattern generator, filter and E/O converter should together have a frequency response to result in the appropriate level of initial ISI eye closure before the sinusoidal terms are added. The E/O converter should have a linear response - if electrical summing is used, linearity of all elements including the E/O modulator is critical. Summing with an optical..."

2. Add new sentence after the last sentence on line 26. "The reference receiver should have very low noise (be highly sensitive), high linearity, and minimal baseline wander, jitter, or other distortions."~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.2~479~42~E~0.1% of 1% of what? We want to require this to be 1% of the total number of hits, not pk-pk width.~Mostly per Piers - "For this test, VECP is determined by using the 99.9th percentile of the total hits within the histogram of the lower half of the signal and 0.1th percentile of the total hits within the histogram of the upper half of the signal. Jitter is determined by using the 1st and 99th percentiles of the total hits within the jitter histogram.
~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.2~480~20~E~A0 should be italicized.~per comment.~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.2~480~29~E~We really want to minimize such jitter. Clarify that residual jitter should be minimized.~Change "not to exceed" to "less than".~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.3~481~54~E~Not clear.~Change to "The values for sinusoidal jitter must comply with Table 52-19 at all test frequencies."~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~Figure 52-6~473~~E~I would gladly offer a more realistic square wave. The present one is pretty bad...

Further more, it is heavily over-filtered compared to the words in step a).~Separate file. Filter BW used in the simulations is 0.4/UI.~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~52.9.10.1~478~49~E~I am somewhat comfortable with Pier's suggestion to generalize the filter name and order, although I am concerned that it lets in a different type of variation than the ones we already have.

On to my comment - On the same line, we should not be referring to the specific value for VECP at this point.~Delete "penalty (VECP)."~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~Table 52-25~491~~E~Pointing to footnote e in draft 4.2, there was an editors note to remove the sentence "TIA...... is presently in ballot". D4.3 removed the editors note, but didn't remove the sentence.

Per Mike.
~Remove the sentence. ~~X~O
0~Lindsay, Tom~tlindsay@xxxxxxxxxxxxxxxxxxxx~(425) 672-8035 x105~(206) 790-3240~Stratos Lightwave~52~Table 52-12~462~~T~Given all the changes since D4.0, I have been reviewing all the power budget values. In the process, I realized that the budget for -L has become negligibly negative. While this in itself is truly negligible, we then set the max TDP for -L to 3.2 dB while most recently requiring negligible BLW or other distortions in the reference Rx used for the TDP measurement. With a 3.2 dB measured TDP, another 0.2 dB or so could be incurred in a real Rx, resulting in another 0.2 dB or so of negative link margin. This is arguably no longer negligible.

As a further note, the rationale for negligible BLW in the reference Rx was used to limit the TDP for -S to 3.9, and has also been used in calculating link margin with real Rx's in the recent -E study. -L is not exempt.

Supporting intuition, this also brings VECP and max TDP more closely in line.
~Change max TDP to 3.0 dB.~~X~O