Hi group
I have the following issue which I don't have an idea about this.
We have 3 cmts (bsr2000, bsr1000 and bsr2000). Each cmts have different DS freq, 555, 561 and 549 Mhz,respectively. The problem is that some cm from 555 are trying to make ranging on 561. The config file is created to redirect modems to 555 but the cable modem gets on send blinkin and trying 561 until I don't know why the cable modem goes online but after long time. When gets online, navigation gets slow.
What could be the reason?
Thanks in advance.
Here I send some screen shots of one cable modem logs.
Generally it's bad to have cable modems able to see multiple downstreams from different CMTS's. If the CM attempts to range on the DS you dont want and each cmts uses the same upstream, then you have a CM thats actually transmitting upstream when it shouldn't. It becomes an interferring signal to the modems on the correct CMTS.
You should isolate the forward combining so that each CMTS is isolated and feeds a different area of the plant.
This can also cause unwanted upstream laser clipping.
-Tom
to isolate downstreams but what if it is impossible? I have some quite a few plants with shared/multiple downstreams & CMTS'a and it works ok. In this case I'd rather check what is MER/SN in 555 downstream and check if there is any interference with downstream signal. Also check if levels are correct.
I would have to agree with Tom on this one.. the modem config for that modem should tell it to change downstream channels and it should just move, but you've got to keep track of what modem gets what file in regards to downstream channel, speeds, cpe's etc.... and there are how many modems total for all three systems?
also you're using up channel space that could be used for other services. narrow cast the downstream is the better way to go. and upstream is going to be an issue too. three cmts's also means three upstreams. you're running docsis 1.1 and 2.0, chances are good your not doing 6.4Mhz wide on the 2.0 system because you haven't got enough clean spectrum (unless your running Euro).
Hi group
Sorry for the delay in my answer. I must say that the issue has "disappeared" without any reason at all. What I did was to change US freqs on US ports in one cmts. But I think that is not a solution. I'm very concerned about this become a mayor issue due to we don't know what the cause of the problem. About narrow cast each DS, is a very complex thing to do in our head-end for our design but I will give the advice to cable guys. The strange thing is that our design has been working for around a year without any issues.
I will give some snapshots to get feedback.
Thanks in advance.
What were the upstream frequencies used on all ports on each CMTS before you made the change?
Hi group
Thanks mtntrailseeker. The next are the freqs before last change:
BSR 2000 (1)
US0: 36 Mhz Channel width 3.2 Mhz qpsk
US1: 33.6 Mhz Channel width 3.2 Mhz qpsk
US2: 33.6 Mhz Channel width 3.2 Mhz qpsk
US3: 33.6 Mhz Channel width 3.2 Mhz qpsk
BSR 1000 (2)
US0: 36 Mhz Channel Width 3.2 Mhz QPSK
US1: 36 Mhz Channel width 3.2 Mhz Qpsk
Us2: 36 Mhz Channel width 3.2 Mhz Qpsk
US3: 33 Mhz Channel width 3.2 Mhz Qpsk
BSR2000 (3)
US0: 36 Mhz Channel width 3.2 Mhz qpsk
US1: 36 Mhz Channel width 3.2 Mhz qpsk
US2: 36 Mhz Channel width 3.2 Mhz qpsk
US3: 36 Mhz Channel width 3.2 Mhz qpsk
Any advice?
Thanks in advance
Lets say you have a modem properly registered on CMTS #3 (bsr2000). If the modem goes offline and into a channel scan process and finds a downstream channel from CMTS #2 it will begin to range based on the UCD (upstream channel descriptor) it detects for that downstream channel. Problem is the UCD's are the same for both cmts (using the same upstream freq). This modem is attempting to range with the wrong CMTS, it will power range through all the transmit levels it can, from min to max level because it never hears a reply from the CMTS. Why, because it's listening to the wrong downstream channel.
Now lets say the modems properly registered on CMTS #2 are all at a transmit level of +40 upstream. The modem that is incorrectly ranging will transmit at levels below this and ABOVE. The ranging attempts at levels above this will be basically an interferring signal to the properly registered modems. Keep in mind that this ranging activity is BEFORE the modem gets a configuration file pushing it to a different downstream.
Broadcasting DS's from multiple cmts's is just plain bad. It causes longer modem registration times especially following a plant outage.
You could try using unique upstream frequencies for as many ports as you can, but you probably dont have a lot of good SNR upstream space for this. The other problem doing this is modems can still transmit upsrream where they shouldn't. If you have other upstream services besides these 3.2 data channels, "rougue" modems could disrupt services there as well.
What is the problem with narrowcasting the downstream signals out of the headend??? Is there only one fiber transmitter or one coax leg that leaves the building???
Admit it mtntrailseeker, you miss being out in the field ;-)
Anyway, in the start of the post, if memory serves me right.. there is an issue where there isn't enough optical transmitters or head end equipment to narrow cast the downstream channels. not your fault, your just working with what you have. but it is what it is... let your manager know that even though it is working, it's not done right and with an outage in side of town, it would cause issues with service with the whole system and when those customers call in to bitch about it or change service provider when they finally get the option... he'll know why.
the equipment pays for itself in little to no time. need to invest in the head end or field to keep it going. was fine with one cmts and I'm sure it's paid for itself. added more cmts's and customers. time to do it right now.
what will you do when you decide to go to docsis 3.0? you know you can't have 24 downstream channels to cover the same area of cable plant..
~Carl
--quote--
Broadcasting DS's from multiple cmts's is just plain bad. It causes longer modem registration times especially following a plant outage.
--/quote--
This really isn't true if your DS frequency is in the modems config file. The most it should range is twice, once to get the config, then second to range into the correct cmts. You just have to make sure you don't have overlapping upstreams, or you get the "interference effect" as you said. Though, even a modem that is ranging won't cause interference because it will read the upstream channel descriptors that the cmts it locks onto is putting out before it ever transmits.
The other problem I see is that at 3.2mhz, a channel centered on 36mhz is too close to a channel centered on 33.6 or even 33mhz. So, you need to subract half of 3.2 from your center frequency to get the bottom edge of your channel, which would be 34.4, which even assuming you had no buffer between the channels, the next place you could center a channel would be 32.8mhz without it overlapping. Generally, its a good idea to leave .5mhz between channel edges, so your next safe distance is 32.3, but to make it easy, lets just say 32mhz. (subtract 4 to get your next center) when they overlap slightly in this manner, you will see a low SNR reading, as it will see the overlap as noise.
So, assuming he's feeding all 3 downstreams to all segments, and each upstream port on the cmts's represent different physical upstreams (but parallel, meaning port 0, on all 3 cmts's is coming from the same node or physical segment) Then you would want your upstream frequencies to look like so:
BSR 2000 (1)
US0: 36 Mhz Channel width 3.2 Mhz qpsk
US1: 36 Mhz Channel width 3.2 Mhz qpsk
US2: 36 Mhz Channel width 3.2 Mhz qpsk
US3: 36 Mhz Channel width 3.2 Mhz qpsk
BSR 1000 (2)
US0: 32 Mhz Channel Width 3.2 Mhz QPSK
US1: 32 Mhz Channel width 3.2 Mhz Qpsk
US2: 32 Mhz Channel width 3.2 Mhz Qpsk
US3: 32 Mhz Channel width 3.2 Mhz Qpsk
BSR2000 (3)
US0: 28 Mhz Channel width 3.2 Mhz qpsk
US1: 28 Mhz Channel width 3.2 Mhz qpsk
US2: 28 Mhz Channel width 3.2 Mhz qpsk
US3: 28 Mhz Channel width 3.2 Mhz qpsk