[802.3_100GNGOPTX] Derivation of improved TxVEC parameter M

[Piers Dawe <piersd@xxxxxxxxxxxx>]

On yesterday's MMF call I was asked to give the reference for the noise term M that appears in the improved TxVEC document we were reviewing (the same term appears in e.g. "802.3bm TxVEC improved revision 5 (post meeting)", http://ieee802.org/3/bm/public/mmfadhoc/meetings/jun26_14/802.3-95-TxVECimproved5.pdf ).

The equation for M comes from the terms for mode partition noise and modal noise in slide 18 of "Improved 100GBASE-SR4 transmitter testing", http://ieee802.org/3/bm/public/mmfadhoc/meetings/may15_14/Improved100GBASE-SR4txTestingV3.pdf .

The term for mode partition noise can be quickly found from the 10 Gigabit Ethernet link model, http://ieee802.org/3/ae/public/adhoc/serial_pmd/documents/10GEPBud3_1_16a.xls , or a variant such as "Example MMF Link Model", http://ieee802.org/3/bm/public/may13/ExampleMMF%20LinkModel%20%20130503.xlsx .

The term for modal noise, 0.01 * average level of whole signal, is an approximation of the sigma/worst_1 = 0.0075 derived in "Modal noise in 100GBASE-SR4", http://ieee802.org/3/bm/public/jan14/dawe_04_0114_optx.pdf , from "ExampleMMF LinkModel  130503.xlsx" referenced above.
(The same derivation from 10GEPBud3_1_16a.xls gave sigma/worst_1 = 0.0073.)

Piers