| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Mike, If that is the main value proposition then
I agree that 19 (or 18.75) dB should do the trick. However, the differential
cost of better transceivers at 10G is likely considerably lower than at 40G. So
increasing the percentage of candidate channels might lower the overall
deployment cost for the full network. Paul From: Mike Dudek [mailto:mike.dudek@xxxxxxxxxx] I think we need to be careful what we
mean by “It just reduces the
expected percentage of channels that would be upgrade
candidates”. The question has to be asked what you are upgrading
from. Pete’s analysis seems to me to be saying that if we choose a
19dB budget then we could upgrade 100% of the 40km 10GBASE-ER channels which is
obviously very important from an Ethernet standpoint. Mike Dudek QLogic Corporation Senior Manager Signal Integrity Aliso 949 389 6269 - office. Mike.Dudek@xxxxxxxxxx From: Kolesar, Paul [mailto:PKOLESAR@xxxxxxxxxxxxx] Right you are Pete. It just reduces
the expected percentage of channels that would be upgrade candidates.
That aspect is really what is in trade against various power budgets. As
with the multimode reach discussions, this question really should be addressed
with respect to the cost of the alternatives. Lower channel coverage
capability means either: 1) the channel must be upgraded or replaced, or 2) the user must deploy transceivers that are more capable than
40GBASE-ER4. We should see some cost analysis of these
scenarios to guide our decision. Paul From: Anslow, Peter [mailto:panslow@xxxxxxxxx]
Paul, Yes, I took the multimode number by
mistake from: The maximum link distances for multimode
fiber are calculated based on an allocation of 1.5 dB total connection and
splice loss. Thanks for checking that. It does
not, however affect the conclusion. The corrected response is therefore: The maximum channel insertion loss for
10GBASE-ER from Table 52-24 is 11 dB with 2 dB for connectors at 1550 nm. This is
(11-2)/40 = 0.225 dB/km This covers about 35% of installed links on
Figure 10-8 of G.Sup39. 0.225 dB/km at 1550 nm is equivalent to 16.75 dB for a 40km link at
1264.5 nm. If we add 2 dB for connectors we get 18.75 dB which fits well with
the proposed 19 dB maximum channel insertion loss for 40GBASE-ER4. Regards, Pete
Anslow | Senior Standards Advisor From: Kolesar, Paul [mailto:PKOLESAR@xxxxxxxxxxxxx]
Pete, Thanks for the loss analysis. Clause
52 says the connection loss allocation for 10GBASE-ER is 2 dB. The maximum link distances for single-mode fiber are
calculated based on an allocation of 2 dB total connection and splice loss at 1310 nm for 10GBASE-L, and 2 dB for 30 km
total connection and splice loss at 1550 nm for
10GBASE-E. Using 1.5 dB would not necessarily ensure
an upgradable channel. Paul From: Anslow, Peter [mailto:panslow@xxxxxxxxx]
Hi, As noted in the draft minutes of
yesterday’s SMF Ad Hoc meeting, there were a couple of questions related
to the maximum channel loss: ·
What coverage of
links is the 18 dB loss for 100GBASE-ER4 equivalent to? ·
What loss does
the channel loss for 10GBASE-ER translate to for 40GBASE-ER4 wavelengths? The answer to the first question is that
(18-2)/40 = 0.4 dB/km at 1294.53 nm (lowest wavelength for 100GBASE-ER4) is
equivalent to about 50% of installed links on Figure 10-8 of G.Sup39. The maximum channel insertion loss for
10GBASE-ER from Table 52-24 is 11 dB with 1.5 dB for connectors at 1550
nm. This is (11-1.5)/40 = 0.2375 dB/km This covers about 50% of installed links
on Figure 10-8 of G.Sup39. 0.2375 dB/km at 1550 nm is equivalent to
17.25 dB for a 40km link at 1264.5 nm. If we add 1.5 dB for connectors
(assuming that we are going to upgrade an existing 10GBASE-ER link) we get
18.75 dB which fits well with the proposed 19 dB maximum channel insertion loss
for 40GBASE-ER4. Regards, Pete
Anslow | Senior Standards Advisor |