Community Forum

Paying for Gigabit but Daily Drops to 30 Meg

Regular Visitor

Paying for Gigabit but Daily Drops to 30 Meg

I've had the Gigabit tier service since moving to my new apartment in April, 2020. Generally, I get just under 900mbps down, and about 42 mbps up.


Every day (or more), I have a "slowdown", where my speeds (BOTH upload and download) drop to about 25-32mbps. The only solution is a modem reboot. I am a network technician for a prestigious college, so I know my way around a network and troubleshooting steps. "Have you tried ____?" Yes, yes I have tried that already.


My physical setup is as follows:


8-way tap block on utility pole → Xfinity demarc → 8-way amplifier → modem → router → network



My router is a Dell PowerEdge 1950 III running a Linux-based system called "pfsense". Before everyone jumps to the conclusion it must be that, I have already tried using my Negear NightHawk X6 R8000 router, which provides the same results. I can run the "speedtest" command, which I'll attach at the end. Same results.


My modem is a Netgear NightHawk CM1150V (yes I need the telephony modem, as I have two telephone lines). "Oh this sounds like a modem issue", well I already tried that. It's less than 3 months old, so I was able to get a warranty replacement CM1150V. Same results.


"Well what about that amplifier?" I have tried both the amplified and non-amplified ports. Amplified port gives me power ratings somewhere around +8dBmV for all downstream channels. The non-amplified ("VoIP") port (which I am currently using in the attached screenshots) gets between +0.2 to +3.7dBmV depending on the channel.



"What about your Wi-Fi?" I am using hardwired connections. The result is the same even from the "router" itself. My wired network is as follows:


Modem → pfsense server via two 1gbps copper Ethernet (using 802.3ad LACP - and yes, I get the same results with LACP disabled, as well as from the Netgear router, which does not support LACP, so I've ruled out LACP as a cause).

pfsense server → Cisco main network switch via 1gbps fiber

Cisco main switch → Cisco local switch via 10gbps fiber

Cisco local switch → desktop PC


Again, same results occur with LACP disabled and running the test directly from the pfsense server.


"Well maybe it's your cabling" I've proven that not to be the case. I ran all the interior connections myself, and have two separate feeds from the amplifier to the server room. Using either feed produces the same results (and the same signal tests, etc., and of course I have tried both feeds using both the amplified and the non-amplified port). I also have a Sadelco cable TV signal tester, which shows the signal strength on digital/analog cable TV lines. The signals were beautiful.


From the amplifier in the basement, I ended up running a new coax between the demarc and the input. Same results.

A Comcast technician came out to my place and he ran some sort of "new" test (according to him) where he can run signal/cable test against my modem and it responds to his equipment. He said it can even tell when a coax cable had been bent 90° and bent back in the past. It also shows fluctuations in the signals. He said it was perfect, and was impressed by my cabling.


I asked the technician if he would please run a new drop from the tap block. We use the same hardline tap blocks where I work inside wiring closets, and I know they can sometimes have issues. He used a different port on the tap block, connected a new PoE filter, ran a brand new RG6 down to the demarc, and connected it to a different coupler inside the demarcation box. Same results.


At this point, everything between the output from the tap block on the pole all the way to a different modem and different router, has been swapped out. It's not my equipment, it's not my wiring. I have now proven this to be the case, that the issue must exist upstream of the tap block.


I would prefer Comcast doesn't waste my time, and would please PROVE to me if they try to say there are no issues on their end. Because at the end of the day, I am paying for gigabit service, and I am getting 30 megs. There's obviously an issue, and I've ruled out every single piece of equipment and wiring on my side of that equation.



Attached info:


Current speed test, 2020-08-02:


My usual speed test, after rebooting the modem:


Current speed test from pfsense:



Cable modem event log - the two "Time not established" and the three events above it are always at the time I reboot the modem. There are never any new events between the reboot - at which point everything is working fine for a while, but you see the new events then - and the slowdown. The 3-5 events are always right after rebooting.


I do also get these errors sometimes:


And of course, the signal page (un-cropped, since it squeezes it onto a smaller iframe...):


If someone from Comcast can actually check and PROVE whether or not the equipment upstream from my demarcation point is having issues, I would really appreciate that. I've seen forum posts from people with the same issues as mine, some were never resolved, and one I recall, they were told it was their modem, they replaced the modem three times, and after several months of this, Comcast ultimately found out there was an issue with some node upstream from the customer, replaced the node, and the customer's issues went away. I would really not like to play the blame game here. I've proven beyond a reasonable doubt that it's not my equipment. I would now like Comcast to do the same.


If anyone else has had this issue, and you have gotten it resolved, PLEASE let me know what you did! I see "Closed" topics on this where people never say what the root cause ultimately was...


Thank you,