Welcome to the Spectrum Community

Get answers, share solutions, help your community.

Start here

Gig service issues climbing uncorrectables 45mins after modem is reset

Asjb123Asjb123 Posts: 14
in Connectivity May 09, 2019

So ive had a bunch of truck rolls and techs to the house they all say signals are good just wondering if this dip in the 500mhz area is normal? there has been no issues with the upstream just Cm status event type 16 And event type 24.

Screenshot (27).png

also the correctables have stopped after noticing my coaxial drop was not properly grounded and left with a wobbly bonded ground wire... but as i said earlier just wondering if that dip before the 500mhz is normalScreenshot (32).png

 

Screenshot (33).png

 

Comments

  • Asjb123Asjb123 Posts: 14
    May 10, 2019

    CD62A207-2C91-479C-B8A4-F790B82B632D.png

    So I am assuming that dip is not supposed to be there considering this is a factory reset log with 3 hours of up time 

  • reds91185reds91185 Posts: 2,452
    May 10, 2019

    Signals look good to me. Are you having any connections issues? A few corrected /uncorrected are normal and could be a natural part of the boot process. Several million is cause for concern.

  • James_MJames_M Posts: 2,110 ADMIN
    May 10, 2019

    @Asjb123 ,  

     

    I'm seeing the logs are from a Netgear modem.  What is the modem make and model you are using?  

  • Asjb123Asjb123 Posts: 14
    May 10, 2019
    Netgear cm1000 and after fixing the ground issue there is only now 20 uncorrectables ofdm channel has been stable for 3 days. . But whatever tech installed the bond for my coax drop did it improperly which was causing static electricity in the foil shields to discharge on the modem causing the interference... it also explained why the techs could not figure out why I was having issues because their meter can only test center of the cable. . But so far gig service is working 100% now on cm1000
  • James_MJames_M Posts: 2,110 ADMIN
    May 10, 2019

    Thanks.  Currently, there are no customer owned modems that are approved on our network for Gig speeds.  The Netgear CM1000 is considered minimally qualified and speeds using this modem will not be guaranteed.   If the issue returns, we suggest using the Spectrum issued modem and you can manage the network using your own router.  

  • Asjb123Asjb123 Posts: 14
    May 10, 2019

    F3D70B6F-2B49-438A-A749-3BC4284B0653.png

    As these are taken with QoS enabled on router a+ with bufferbloat. . And speedtests were taken via WiFi ..

     

    Only reason I would never use your guys modem is you lock the user out of the modem UI so nothing can be checked and your staff is less than minimally trained to handle any situations... unless it’s taking money 

     

  • karlbeckmankarlbeckman Posts: 2,208 ✭✭✭✭
    May 12, 2019

    Usually a dip like that is caused by a premium channel filter that was installed way back in the old analog days and forgotten ever since.  It should be on the line tap where the pole drop ine heads toward your residence. 

  • Asjb123Asjb123 Posts: 14
    May 12, 2019
    Yeah I was just curious if it was normal but ofdm channel runs at 314mhz then the 24 channels start at 507mhz to 645mhz which I am assuming the ofdm uses for the channel bonding
  • RAIST5150RAIST5150 Posts: 667 ✭✭✭✭
    May 13, 2019

    @karlbeckman wrote:

    Usually a dip like that is caused by a premium channel filter that was installed way back in the old analog days and forgotten ever since.  It should be on the line tap where the pole drop ine heads toward your residence. 


    Still find  lot of thkse notch filters in our market. Usually see the deep dips around 500mhz or lower (roughly the old broadcast channels 14-36).  Some will be decades old cylinders for HBO and such (no color bands), others will have a green band to choke out the upper channels for the internet only customers so they couldn't scan the QAM channels above 23 or so.

     

    Really wish they could be more proactive about finding these things and removing them now that they have gone full digital.  Some of them are causing up to a 12db drop in the first octave of channels, ultimately rendering channels unusable on some lines, but the modems sometimes try to lock them anyway if there is trouble in the upper octaves.  End result is people have to routinely reboot their modem's because throughput tanks as the modem starts having fewer and fewer channels to work with.

  • Asjb123Asjb123 Posts: 14
    May 14, 2019

    So I understand Cm1000 isn’t recommended for the gig service connection was running flawlessly for 5-6 days decided to check on it.... and before my multiple power cycles I had 5000+ uncorrectables 2000 correctables but here is the log from last power cycle.

    9E480B9C-305C-424C-8176-936DE428565A.png

     

    There were x4 of the critical  Mac sync timeout, x8 event type 24, x3 of type 16 also before reset there was event type 5 on Chan 11-17There were x4 of the critical Mac sync timeout, x8 event type 24, x3 of type 16 also before reset there was event type 5 on Chan 11-17

     

    30599796-4CCE-4CCC-BF9C-659F4263A44E.png

    Also a new coaxial cable was ran from the drop to the cable modem, also the coaxial cable I noticed It feels bunched up aren’t they suppose to be smooth? Also there is a -12db forwardpath attenuator to reduce a hot signal coming into the house. 

  • Asjb123Asjb123 Posts: 14
    May 14, 2019

    Here’s link if you’d like to see it

     

    For any  spectrum techs reading this I found this browsing around the internet this is taken from another post.   

     

     

     

    FAILED_TO_RECEIVE_SYNC: T04.0 Failed to Receive MAC SYNC frame within time-out period

    Explanation The cable modem initially was able to acquire MAC framing but failed to receive the MAC SYNC frame within the timeout period. This error message is DOCSIS event message is T04.0, SYNC Timing Synchronization failure.

    Recommended Action Check the RF plant for cabling and connector problems that could be generating noise on the downstream. If using a Cisco CMTS, you can use the show cable flap-list command to determine if other cable modems on the upstream are having problems. You can also use the show interfaces cable upstream command and examine the noise, microreflection, and uncorrectable error counters to determine the level of noise on the upstream.

    Some modems (Motorola Surfboards especially) record every little thing in the log and users panic. I see the first one and T3 and T4 time out frequently and my connection is fine. What you need to do is watch your connection for awhile and if still seems to be lagging then you need to talk to tier 3 and ask the tech to run the show cable flap list command if they haven't already done it which they may have and haven't told you because 95% of the users would be mystified by this. This will show them if other modems near you are also having problems. If it looks that way, a good, experienced tech will probably then ping some of those modems, while they still have you on the phone, and if they see a problem then they will roll a truck as soon as possible. Some techs don't know to do this or just won't. I've had to call more than once, even to tier 3, and ask for one of the old timers by name. Several times, when the tech looked at the flap list and pinged other modems in the area sure enough just as I had said there was a problem that they were unaware of in my area until I reported it. A couple of other times though (this is over an almost 7 year period now), a tier 3 tech looked at the flap list without my even asking them to and told me everything was great and that my stats were so perfect that it was "scary

     

     

  • Asjb123Asjb123 Posts: 14
    May 17, 2019
    So none of your customer support wants to respond as to what the issue is because it’s clearly plant/cmts issues here and spectrum isn’t going to a thing about it.
  • Asjb123Asjb123 Posts: 14
    May 17, 2019

    Also if I’m reducing an incoming hot power signal why is cmts pushing my modem to increase the power signal? Which is done when time servers are modifiedAlso if I’m reducing an incoming hot power signal why is cmts pushing my modem to increase the power signal? Which is done when time servers are modified

     

  • karlbeckmankarlbeckman Posts: 2,208 ✭✭✭✭
    May 17, 2019

    First, it's more likely that the helpful field techs from earlier times that you referenced (and other Spectrum employees) don't often log into the forums to read user postings, because they already know that the forums are peer-to-peer customer conversations, not related at all to Spectrum's Tech Support. 

     

    Next, CMTS is telling your modem to increase its upstream power to make up for the loss inserted to get the downstream signal power back to the proper range of -8.5 to +10.0 dBmV.

     

    And last, it's not necessary to redact the MAC IDs that appear in the status report entries.  They belong to the maintenance link between the CMTS and your modem, not related to the privacy or security of your data traffic payload.  All you did was make it more difficult to read the status text which describes the problems being detected by your modem.

  • Asjb123Asjb123 Posts: 14
    May 17, 2019
    The upstream power is not being increased it is the downstream side that is being increased unless there is a spike that is happening on the upstream side which is purely plant/cmts issues causing the fluctuations in my downstream power levels.
  • RAIST5150RAIST5150 Posts: 667 ✭✭✭✭
    May 18, 2019

    @Asjb123 wrote:
    So none of your customer support wants to respond as to what the issue is because it’s clearly plant/cmts issues here and spectrum isn’t going to a thing about it.

    This is a peer to peer support forum... their techs don't come in here, just community managers that may be able to forward data to tech support, but mostly direct you to other contact methods (like social media accounts) that can facilitate escalation to other departments.

     

    Basically, what you have here is a group of end users familiar with either the industy as a whole, or at least TWC/Spectrum's way of doing things that can provide insight into what may be going on and possibly some end-user workarounds and/or tactics for getting things addressed by Spectrum support.

  • RAIST5150RAIST5150 Posts: 667 ✭✭✭✭
    May 18, 2019
    Actually, @karlbeckman is correct about the modem adjusting upstream power when directed by the headend to do so. The plant doesn't activelly adjust your downstream power on the fly, but it CAN direct the modem to increase upstream if the return signal quality is not up to par. Techs can adjust downstream levels along the route though... just may require escalation to a higher tier of support than the team they sent.

    If the downstream level is too high at your demarc, they ideally would be adjusting gains (or something similar) upstream instead of trying to insert losses after the fact. Overboosting can worsen SNL, and then forcibly injecting losses on top of that will degrade it further.

    The less headroom there is above the noise floor the more prone the line become to errors... you want to avoid intentionally injecting nore noise into the line.
  • Asjb123Asjb123 Posts: 14
    May 18, 2019

    This is what I’m talking about with power signal being increased  may 18thThis is what I’m talking about with power signal being increased may 18th

     

    May 18thMay 18th

     

    May 1stMay 1st

    This is the increase I’m talking about and both these snapshots are with a -12db attenuator at the drop... guess what I’m trying to say is Why is the power signal getting increased at any given time when all signals to the house are perfect ?? There’s no splitters just ground block then attenuator and the line after that goes directly to modem. 

  • karlbeckmankarlbeckman Posts: 2,208 ✭✭✭✭
    May 19, 2019

    You are confusing the downstream and upstream power levels. The modem power levels are different for upstream (US) vs. downstream (DS).  As their names imply, the signals are travelling in different directions on the cable.  Your DS level could also be called modem receive and the US side is sometimes called modem transmit. When you inserrt a 12 dB loss to bring the DS (receive side) signals to an acceptable level, the modem must transmit 12 dB stronger to overcome the loss of that 12 dB pad.

  • Asjb123Asjb123 Posts: 14
    May 19, 2019
    I’m aware of downstream signal and upstream signal. . And the forward path attenuator has little to no effect on the upstream power levels as it attenuates 120-860Mhz, Highest upstream channel is at 37Mhz so as I said before it has little to no effect on my upstream, I also never see more than .3 dB fluctuations on upstream side of things... with constant upload speeds of 40-41Mbps upload... not saying it’s not some issue on upstream channels that is causing downstream issue.

    Also my apologies it is a -10db forward-path attenuator.
  • Asjb123Asjb123 Posts: 14
    May 19, 2019

    See what I mean With the power signal 

    9FD9D8AD-001A-43A3-A564-F52817AF40DC.png

     

    4F47693B-373E-496E-BE24-31670CF83D47.png

     

  • Asjb123Asjb123 Posts: 14
    May 19, 2019

    So I went through and checked spectrums work of their last coax job they did to my modem. .After replacing their shotty workAfter replacing their shotty work

     

    F95ABE32-96F6-4594-83E8-CB8BCA870770.png

     

    Also can see again here there’s a half moon on the centerAlso can see again here there’s a half moon on the center

     

    Can see the center is chunkedCan see the center is chunked

     

    Here’s where they jammed the coax in the box and stressed the coax... also the braid sticking out of the connecterHere’s where they jammed the coax in the box and stressed the coax... also the braid sticking out of the connecter

    So I re ran RG6Q obviously new connecters and for the hell of it replaced the ground block. . And here’s where I’m sitting at now. 

  • karlbeckmankarlbeckman Posts: 2,208 ✭✭✭✭
    May 20, 2019

    Unless your "attenuator" is actually a multiple active amplifier device with bandpass filtering, the 10 dB attenuation applies in both directions, affecting both the US and DS levels.  Your Spectrum techs or any junior college electronics instructor can explain to you exactly how that works. 

  • RAIST5150RAIST5150 Posts: 667 ✭✭✭✭
    May 21, 2019
    If it is something rated at -10db on your downstream only, then you likely don't need it.

    Seriously... it probably should be tested without it. Environmental conditions impacting resistances and other factors are enough of a headache to deal with as is (especially in the hottter months... some key factors that cause DS levels to vary throughout the day are temperature and moisture). Don't need to be intentionally inserting a device that may exacerbate things, especially if it likely is not needed. Risking more noise on the line, and also likely throwing upstream off slightly---for a line that appears may be within specs without it.

    The goal should always be the minimum number of devices/junctions needed, with the shortest cable run needed, and keep the line as straight as possible--and when it needs to bend, no sharp turns (ideally no sharper than roughly what it takes to go around a 2 liter bottle).

    You want to minimize as much insertion of noise/loss/reflection/etc. as possible.
This discussion has been closed.