dslreports logo
site
spacer

spacer
 
    All FAQs Site FAQ DSL FAQ Cable Tech About DSL Distance DSL Hurdles »»
spc

spacer




how-to block ads



When a cable system is set up for two-way, it is swept. Sweep is the process of verifying that basically all of your amps in the field have the same signature as the node. You are basically trying to re-create the node out to the last amp. The node is the fiber to cable converter that every neighborhood has. Once the amps in a node have all been swept, they are known to have unity gain. All of the amp chips in the return are set and the amount of RF needed to transmit through them to the headend are the same at the input to the amp chip. Now, where it differs is at the taps and other system passives. Ultimately, you are setting your system so that a preset amount of RF will equal 0 dBmV at the CMTS. The CMTS wants to see the RF signals from the modems all at the same dB level. The CMTS cannot discriminate the signals from modems if the values are all different. It is here that the CMTS can dictate the amount of RF needed from each modem. So, when a DOCSIS modem is synchronizing, it first gets the forward info, it then ranges upstream. Ranging basically consists of calls to the CMTS where the modem says, "can you hear me?"... this goes on until the CMTS acknowledges the modem. Once the CMTS can hear the modem, it fine tunes the modems transmit power and the modem can then use DHCP to get an IP, grab a config file, time of day, and boom... your modem is synchronized. The modem receives maintenance packets on a regular interval to adjust the transmit power. This happens OFTEN. (the packets).. The return transmit power should NEVER move more than a dB. If you are looking at the 192.168.100.1 page and your transmit power moves more than a dB or two, you have a problem (usually a loose fitting). Now, how do you compute what the return should be? Easy. First you need to know the unity gain value. Then you need to know the tap value you feed from. (you'll need to know all of the passives in the plant to your tap too. Good techs can figure this out with a print). Then, you take the unity gain figure, add it to the tap value, and that should give you the transmit value at the tap. Most drops to the house under 200 feet should lose no more than 2 dB. Then add your splitter loss and you will have your transmit power value.

Let me make this example. Lets say you live at a house by a tap that is right by an amp. Lets say the tap is a 23 value. Lets use 16 as the unity gain figure (its not 16 in a lot of systems). 16 + 23 = 39 dBmV. So, if the modem was hooked up at the tap, the 192.168.100.1 page should say the transmit is 39 dBmV. Now, add 1 or 2 for the drop to the house (cable) and then add splitter loss. Lets say the modem is hooked up to a two way in the house box outside and you have a two way inside with a tv next to the modem. Do the math... 39 + 1 + 3.5 + 3.5 = 47 dBmV. Your modem should be transmitting 47 dBmV in this scenario.

So, for those of you who have return problems (like really high) The best test you can do is go to the ground block where the cable reaches the house and plug the modem in there. (We have test equipment to do all of this but for you the modem at the house block will work)... You'll need a LONG cat5 cable for this but. Get the modem working right out of the house box with no house cable or splitters. If your modem transmit power is over 45dBmV at the ground block, you probably have a unity gain problem that the maintenance people need to fix. If it is lower than that, start doing some math and look at the splitters in the way. Remember when you add or remove splitters, the transmit value in the 192.168.100.1 page will move roughly the same amount. All of this assumes that you have good wire and solid fittings.

Thanks to TechnoScott See Profile

*This FAQ is based on user knowledge from a volunteer core of BroadbandReports' members. This FAQ in no way constitutes official information from Comcast or any of its affiliates.

Expand got feedback?

by Big_D See Profile edited by Johkal See Profile
last modified: 2008-11-01 14:56:44