CM1000V2 modem cannot connect computer

[NeilU]

So I changed the splitter to a two split and a different cable 

Now the errors are only on 3 channels and the levels are significantly lower than originally.

I think I will leave it like this for now (unless those channels 25,27 and 28 would be important).

[plemans]

You're doing a juggling game. Because the power levels are still out of spec. All the channels should be within 3.5dbmv of each other or they'll have bonding issues. Yours go from 9.2dbmv to 1.2dbmv. a 8dbmv difference. So still out of spec. 

It can bit a pita to do things like this. Its why signal attenuators were made. You can stick them in line and you don't have multiple connections that can be causing issues. Its why I also only used a single x4 block for mine and capped the other ends. 

[Kitsap]

---

@NeilU wrote:

So I changed the splitter to a two split and a different cable 

Now the errors are only on 3 channels and the levels are significantly lower than originally.

I think I will leave it like this for now (unless those channels 25,27 and 28 would be important).

---

All of the channels that are shown as locked are important.  None more so or none less so.

If there are problems with the signal levels on specific channels, that is the responsibility of the ISP to resolve.

The attenuator pads @plemans mentioned are available from Amazon here:  https://a.co/d/26Df1Ll.  They come in different sizes ( 3 db, 6 db, 10 db).  Make sure your buy those that cover the standard frequency range for cable TV internet systems (5 Mhz to 3 Ghz).

The connection caps @plemans mentioned are also available from Amazon here: https://a.co/d/5ELepHl.  They are important to prevent leakage that adds noise to your upstream signal.

[NeilU]

I am beginning to understand. I thought we were talking about putting a splitter but an attenuator is something totally different.

I am now getting all kinds of errors so I am understanding - I will get a set of the three sizes and a set of caps. 

Thank you 

[Kitsap]

---

@NeilU wrote:

I am beginning to understand. I thought we were talking about putting a splitter but an attenuator is something totally different.

I am now getting all kinds of errors so I am understanding - I will get a set of the three sizes and a set of caps. 

Thank you 

---

By design, passive (non-amplifying) splitters add a certain amount of attenuation.  Unless the label was removed, the output connections are usually marked as -3.5 db or -7 db. 

[plemans]

Thanks for the links @Kitsap 

Like Kitsap says, a passive splitter acts as an attenuator. But the problem comes down to it adds more connections/more chances for issues. And they can be poorly made. Like I said, the only reason I used one was I had it sitting around. 

Best bet? 

direct connect the modem right to the cable coming into the home. That should give you your power levels to know what attenuator to buy. If you're getting 12-13 buy a 6db attenuator. I've found its better to be a little hot with the signal (5-7dbmv) versus to low. 

Get the attenuator and connect it directly to where the cable comes in the home and then connect the modem right off that. With zero splitters. 

Then check signal levels/errors and use it for a bit to see if it stays stable. 

If it doesn't, the line outside the home needs to be checked. 

[NeilU]

I have a direct cable ready to put in place - probably this evening when we are not online. Then I can see just how that helps. 

The attenuators and caps are from Amazon and I will see if I need any of that - if not they are easy to return. 

I will be able to see how things progress since I have a bunch of screenshots over the last day for power and errors. 

[NeilU]

Here is an update. Late yesterday I got the 10 db and installed it. After restarting it cleared up all the codes and lowered the power. I have watched this for a full day and this is the levels and codes after a full day - correctable had been down to 0 except for the first and last few. Power had been down to -2.8 for the first channel and now is down to -6

Others are in a tight spread. 

I just received the 3db and tomorrow the 6db arrives. 

I can replace the attenuator with the 6 to see if that is better. 

[plemans]

I'm betting they'll need to check the line. You've still got channels going from -6 to +3.0. So a 9dbmv difference. Again, it should be within 3.5dbmv of each other. 

so something is causing issues with the power levels. When they come, I'd pull out the attenuator. 

[Kitsap]

I will second what @plemans says about pulling attenuator.

When you are keeping screen shots of cable connection for records, be sure and include the system uptime at the bottom.  The correctables/uncorrectables values accumulate since the last modem boot time.  Knowing the time frame the numbers accumulated over is big part of the evaluation.  If the numbers are big and you do not know whether they accumulated over a few hours, a few days, or a few months, then it is difficult to put them in perspective.

There are definitely some issues with the channels in the higher frequencies.

[NeilU]

The screen shots are time stamped and I identified which one was when the system got restarted and essentially all set back to zero. 

I didn't look at how the channels are increasing by frequency. Yes the problems are at the high frequencies - there is a definite distribution of errors at the 603 MH level with the primary channel (if I assume the one listed as 1 is such - id 28) 

Here is a question - now I see the "Frequency Start Value" which I can enter any value. 

What is an appropriate one to be putting there or does it matter? 

I changed the number and it immediately went to reboot. whoops 

Well It now restarted and everything is set to zero on correctable and uncorrectable and the value I tried as starting frequency was ignored. So we have a new beginning set which I can track again. 

So now I need to understand what the implication of issues at different frequencies. 

[Kitsap]

---

@NeilU wrote:

The screen shots are time stamped and I identified which one was when the system got restarted and essentially all set back to zero. Including the System Up time in the screen shot makes this simpler.

I didn't look at how the channels are increasing by frequency. Yes the problems are at the high frequencies - there is a definite distribution of errors at the 603 MH level with the primary channel (if I assume the one listed as 1 is such - id 28) 

Here is a question - now I see the "Frequency Start Value" which I can enter any value. 

What is an appropriate one to be putting there or does it matter?  Best to leave that alone.  Does not change the end information.

I changed the number and it immediately went to reboot. whoops  See above.

Well It now restarted and everything is set to zero on correctable and uncorrectable and the value I tried as starting frequency was ignored. So we have a new beginning set which I can track again. 

So now I need to understand what the implication of issues at different frequencies. 

---

The laws of physics control that happens at the higher frequencies.  The power levels will naturally decrease as the frequency increases.  This where your ISP comes in to play and where they have responsibility.  There are amplifiers along the path in the ISP distribution system.  The amplifiers are grouped to cover four or five channels at a time.  The gain on the amplifiers can be adjust by the ISP.  This happens remotely from the location where the CMTS (Cable Modem Termination System) is located.  In your situation, it looks like a higher frequency amplifier in the path is bad.  There can be several of these amplifiers in the path depending on how far you are located from the CMTS.

I did no go back and reread the whole thread.  Your coax cable should be RG6.  It might show RG6 or series 6 stamped along the edge.  If it is older coax, it should be replaced.

[NeilU]

Thanks for all the input previously.

I installed a new RG6 cable all the way from input to the modem. 

I first installed the 6db - this was a week ago. That yielded a couple of +2 and above with many at about +6 to 7

Later I installed the 10db and now the values were from about -4 to +3.5 after restart. No uncorrectable for almost a day and then on 6th I rebooted and have tracked ever since.

Within a day it started to get errors which would sit for some length of time and then every so often it appeared to do some sort of power shifting and then hold with errors for up to a day or so. The primary highest channel would also jump from maybe -6 to -3.9 to -1.3 and each time maybe hold for half a day or so while the number of uncorrectable codewords would hold and then jump. 

The latest attachment is since March 6. 

I am going to vacationing for the next week so I won't be able to watch this now and may do a reboot before we leave and revisit in a week. 

This is the latest screen shot - The total number of unerrored seems doesn't make sense over time. For a number of days the total codewords were constant across all the channels. Suddenly Channels 27 and 28 were much lower than all the rest which then remained the same continuing until now. 

[Kitsap]

---

@NeilU wrote:

Thanks for all the input previously.

I installed a new RG6 cable all the way from input to the modem. 

I first installed the 6db - this was a week ago. That yielded a couple of +2 and above with many at about +6 to 7

Later I installed the 10db and now the values were from about -4 to +3.5 after restart. No uncorrectable for almost a day and then on 6th I rebooted and have tracked ever since.

Within a day it started to get errors which would sit for some length of time and then every so often it appeared to do some sort of power shifting and then hold with errors for up to a day or so. The primary highest channel would also jump from maybe -6 to -3.9 to -1.3 and each time maybe hold for half a day or so while the number of uncorrectable codewords would hold and then jump. 

The latest attachment is since March 6. 

I am going to vacationing for the next week so I won't be able to watch this now and may do a reboot before we leave and revisit in a week. 

This is the latest screen shot - The total number of unerrored seems doesn't make sense over time. For a number of days the total codewords were constant across all the channels. Suddenly Channels 27 and 28 were much lower than all the rest which then remained the same continuing until now. 

---

You are making good progress on the signal power levels.  There are still a lot of uncorrectable line errors over several days.

The line errors should be the responsibility of the ISP for wiring outside your house.

[plemans]

If you've done what you can, the isp needs to check it.