Read the forum guidelines
Ever since my downstream channel was moved from docsys 3.0 to 3.1 my speeds refuse to go over 250mbps down. I have the 300/30 package and would average over 400bps down. Using the arris docsys 3.1 modem
Have you tried a factory reset? Most modems will have a hole/button on the back you can push with a toothpick/paperclip to factory reset the modem. This clears the configuration file, which holds the memory of what downstream channels the modem should use. After the factory reset, you will often have to power cycle your router. If a factory reset doesn't fix the issue, perhaps a moderator here can re-provision your modem.
What model modem do you have BTW?
I have the sb8200 and reset it a few times already. I've done a refresh but not sure of ive done a provisioning. My network at home is fairly beefy. Full cisco meraki setup with a ubiquiti mesh wifi network. The only thing that changed through all of these was the docsys 3.1 light turning on for downstream. I read about a firmware compatibility somewhere. Trying to avoid calling support as they will most likely just want to send someone out for something that can probably be fixed remotely
Yes I have a Mac mini hardwired. It still tops out at 250. Very odd
Which version Mac Mini? See here on how to tell.
Also, what test are you using? I suggest DSLReports speed test.
It's a late 2014 model with the core i5 processor
If your wireless is faster then your ethernet, then something is bottle necking the ethernet. Does it show a gigabit connection between the Mac and router?
Also, why do you keep not mentioning if the Reddit post is you?
Sorry thought i answered that above. That reddit post is not me. Both speed tests are basically identical hardwired or wireless. I used to break over 500mbps downloads on docsys 3.0. So in the last 8 months thats the only thing thats changed. Well except my genius son was downloading torrents and i had to call to have my service turned back on. Ive blocked bit torrent in the meraki now so good luck to him.
sschmelcke said:That reddit post is not me.
Oooh,OK. Must be a odd coincidence? I happen to watch the Cox Reddit thread too and saw that pop up right around the same time. That person seemed to be having serious security issues, so I thought it was important to make sure that wasn't you. Apologies if it took the thread off course.
sschmelcke said: I used to break over 500mbps downloads on docsys 3.0.
Do you still have the old D3.0 modem? I would be curious is your still get the speed now. I think other changes happened, that run parallel but seperate, from the change from D3.0 to D3.1. When a area upgrades, it usually also opens up additional downstream channels. This would usually increase download speeds, not decrease, but perhaps the change in downstream channels shifted you over to some that were more congested. .
To that end, could you post your signal levels? They are available on the modem's 192.168.100.1 web page. See here for more info. I am looking for downstream channels that are lower/higher then the other channels around it.
Finally, I think there is a separate issue going on with the Mac Mini. I heard the latest OSX did some screwy things with the ethernet drivers. If your getting the speed via wireless, it's getting to the modem. The question is what is causing the spikes of slowness, which the signal levels will help answer.
Sadly its long gone. I bought this modem with the intent of going gigblast when it was available. I had all green lights showing docsys3.0 and the modem was nice and fast. Got home and the downstream channel turned blue one day. Got excited and ran a test only to see what im trying to figure out now. Ive debated on trying another docsys 3.1 modem but they are still extremely pricey. If they were as cheap as as a docsys 3 modem id go for it
The only other thing I can see that would be at play is that for the D3.1 to work on that modem, Cox had to load a special configuration file. The file would have changed over night when the shift to D3.1 went live. Maybe that file changed how the throttle works to define your speed, and the test your using is more sensitive to it. It's a stretch, but I am grasping at straws at this point. Sorry I don't have a better answer for you.