how-to block ads
NetFixerFrom my cold dead handsPremiumReviews:
|reply to telcodad |
Re: [Connectivity] Net Drops Constantly
said by telcodad:I have stress tested my current SB6121 by moving it from its normal -3.5 db splitter leg to a -7 db splitter leg. When on the -3.5 db leg, my upstream levels read ~-45-48 dBmV. When on the -7 db leg, my upstream levels read ~-49-52 dBmV. I have left it on the -7 db leg (with at least one upstream channel at the -52 dBmV level) for multiple days, and my connection remained solid.
While I'll agree that "very high" might have been a little too much, if some people are saying that they have had issues when their upstream levels hit 50 dBmV, then it is more than just "marginal." You also have to allow for a few dB of "headroom," given the normal, daily fluctuations most HFC systems experience.
I have also tested my old (backup) D-Link DCM202 modem in a similar fashion, and its upstream level readings consistently read ~ 3 dBmV lower on either splitter leg than my SB6121. So, if my old D-Link had started failing at the -50 dBmV reading, would it really have failed at -50 dBmV, or would it have failed at -53 dBmV, or perhaps both of my cable modems are so far off in their readings that it would have actually failed at -55 dBmV ?
Cable modems are not expensive cable line testers, and their line level readings are not to be taken seriously. At best, the day to day readings on the same cable modem on the same connection, are good as a relative measure of conditions on a day to day basis for that specific modem on that specific connection. Those readings bear no real or reliable relationship to actual values measured with a high quality tester.
We can never have enough of nature.
We need to witness our own limits transgressed, and some life pasturing freely where we never wander.
Given the variations in the architectures of and equipment used in the various HFC systems Comcast has cobbled together over the years, I would think it is also possible that in some systems (like yours), upstream power levels of 51 dBmV or slightly more may not be a problem, while in other systems (like in ExoticFish 's and owlyn 's), levels that high do cause issues.
The differences between different brands/models of modems can also enable some of them to perform OK at >50 dBmV levels.
Like they say - YMMV.
EDIT: Looks like the upstream power limit is determined by the modem and, as some posters have stated before, the number of upstream channels. From page 3 of the data sheet for the Moto SB6121 (»www.motorola.com/staticfiles/Sup···heet.pdf ):
Upstream Operating Level Range Level range per channel:
(Multiple Transmit Channel mode disabled, or only Multiple Transmit Channel mode enabled with one channel in the TCS)
Pmin to +57 dBmV (32 QAM, 64 QAM)
Pmin to +58 dBmV (8 QAM, 16 QAM)
Pmin to +61 dBmV (QPSK)
Level range per channel (two channels in the TCS)
Pmin to +54 dBmV (32 QAM, 64 QAM)
Pmin to +55 dBmV (8 QAM, 16 QAM)
Pmin to +58 dBmV (QPSK)
Level range per channel (three or four channels in the TCS)
Pmin to +51 dBmV (32 QAM, 64 QAM)
Pmin to +52 dBmV (8 QAM, 16 QAM)
Pmin to +55 dBmV (QPSK)
EGThe wings of lovePremium
|reply to NetFixer |
said by NetFixer:Yes !!Exactly !! Perhaps that is why some posters should not constantly and automatically say that "50 dB is fine" ?
Cable modems are not expensive cable line testers, and their line level readings are not to be taken seriously.