@rmckayfleming Log into your modem and check "Gateway > Connection > Rogers Network". (I presume that you are an Ignite Internet customer.). If you don't see excessive uncorrectable codewords errors (the downstream OFDM channel should have very few, if any uncorrectable errors) then your network connection is probably fine.
It also could be your web browser... possibly an extension-related issues... or if you have DNS-over-HTTPs enabled, name resolution could be slow at times with some DoH servers. Sometimes it could be an issue with the browser itself. e.g. With Vivaldi, there are times where opening a link in a background tab can (randomly and very intermittently) take a few seconds. The same thing can also happen with other browsers too, depending on the configuration.
Personal opinion, "if I say power cycle my modem things, might be fine for awhile,", that indicates that you have a cable signal level problem, most likely due to the external cable that runs from your house to the local tap.
While your logged into the modem looking at the codewords, please copy and paste both the signal level table and codeword table and paste them into a post. It would be useful to see those signal levels after a modem reboot, and when you are experiencing issues with your internet service.
To copy the tables, park your curser just ahead of the first character in the table. Hold the shift key down and scroll down and to the right, until the curser sits just after the last character in the table. Release the shift key and right click .... Copy. In a new post, right click .... Paste. The table should paste in just as it appears in the modem's UI. You can also use the arrow keys on the keyboard to scroll down and to the right.
Do the same for both signal levels, and codewords.
Now, just thinking about this, didn't Comcast, and therefor Rogers shut down the web access to the modem, or was that just the online access? Don't remember the details on that one. If you can access the modem directly via web browser, you should be able to copy and paste the table data.
After the reboot, the signal levels should look pretty good. When they degrade, that's the interesting point. Do they degrade to the point where Rogers will actually do anything about it? That is the question of the day.
@rmckayfleming your downstream QAM channels (1 to 32) are a little low in terms of the signal level, but, that won't cause any issues. The target is 0 dBmV with a Signal to Noise Ratio between 36 to 40 dB. So the QAM levels aren't too bad, and the Orthogonal Frequency Division Multiplex (OFDM) channel should be ok. Only the moderators and Level II techs have access to the OFDM MIB data which is a breakdown of the OFDM channel into smaller range sub-channels.
The upstream is a different story. For some reason your QAM upstream channels are all running higher than they should be. The target is 36 to 40 dBmV. The max power output power level, according to DOCSIS spec is 51 dBmV. Rogers uses 52 dBmV for some reason. By then I'd expect the modem to start shutting down channels, one by one in order to maintain communications with the Cable Modem Termination System (CMTS). If you were in that situation you would know it as your data rates would drop.
Ok, so, something is definitely amiss. My thinking here is that there is such a large discrepancy between the downstream and upstream signal levels, that it might be caused by a signal attenuator installed somewhere in your cable system. That would have been installed by a tech if the incoming signal level was too high when the modem was installed, so the tech may have installed a signal attenuator to drop the downstream signal levels. That attenuator also drops the outgoing signal as well, and as a result, the CMTS will command the modem to increase its signal level output. That would explain the high upstream signal levels. I'm wondering if you're getting occasions where one or more of the upstream channels is exceeding 52 dBmV, which should cause the modem to drop one or more of the upstream channels. That would definitely cause problems such as what you're currently experiencing. Using any signal attenuator to drop the downstream level without forcing the upstream levels into failure range is a delicate balancing act. There are Forward Path Attenuators which drop the downstream levels without affecting the return upstream levels, but, I've never seen Rogers use these, which is unfortunate as it would very simply address these situations.
So, can you have a look at the back of the modem and see if there is a signal attenuator attached to the modem. The attenuator will look like or very similar to the following:
https://www.antronix.com/products/other/attenuators
The attenuators have different signal drop levels, 3 dB, 6 dB, 9 dB and possibly 12 dB. If there is a signal attenuator connected to the modem, can you determine what digit is printed on the side of the attenuator. That will designate the signal drop thru the attenuator, in both directions.
The other possibility is that there is a signal attenuator connected to the incoming cable, where that cable meets the house internal cabling, or that a tech has installed an attenuator in the outside Network Interface Device (NID) which is the grey Rogers box that sits outside of the home, or at the local tap, which that NID connects to. The local tap is essentially an overgrown splitter that is located in a nearby pedestal, or on a nearby utility pole. My bet is that its installed on the modem, but, there are other possibilities.
There is also the possibility of a splitter connected to the incoming cable to act as a signal attenuator. I'd expect that to be a large splitter which usually provides a large signal drop. Can you check for a splitter connected to the incoming cable and determine what number is printed on the port that the modem cable is connected to That will indicate the signal drop thru that port. At the same time, please note the model number of the splitter so that I can look up the specs. If there is a splitter which still remains connected to the incoming cable, and its has other house cabling connected to it, those cables should be disconnected and the open ports on the splitter capped with a 75 Ohm terminator to prevent any signal reflection caused by those open ports. We're not there just yet, but its worth looking at this when you go hunting for the splitter, if its installed.
Here's what the splitters might look like:
https://www.antronix.com/products/drop-passives/splitters
So, my bet, at the current time is that there was an initial high downstream signal level that required attention. If there is an attenuator or splitter installed, can you remove it, reboot the modem and then copy and paste the signal levels after the reboot. That would show the signal levels as they are currently running. It would be useful to see what they are running at, without any signal attenuation.
The other explanation, beyond the signal level issue is degradation of the external cable, which is causing signal drop outs. That can be detected by pinging the CMTS, looking for packet loss. That would be the next step, to quantify any packet loss due to the external cabling.
Do you have underground cabling in your neighbourhood, or is the cabling run from a utility pole?
I've attached two images of the setup in the basement below. Looks like it's a fibreoptic connection to the house (this house was built in 2020). There's no attenuator or anything directly on the modem (a Technicolor CGM4140COM hardware revision 2.2), only thing seems to be the big splitter in the basement. I'll have to figure out which cable heads to the modem.
@rmckayfleming Thanks for the picture! That fibre optic thingy is an RFoG ONU; it converts the DOCSIS cable Internet service signals from fibre to coax. On the RFoG ONU, disconnect the coax cable that leads to the splitter and connect the coax cable leading to your modem's wall plate. That should bring your modem's signals back to normal.
You are not getting a huge number of Uncorrectable Codewords on your downstream D3.1 OFDM channel but the modem stats to not provide any visibility into the upstream performance. Hopefully, removing the splitter will resolve at least some of the issues that you are seeing.
@rmckayfleming removing the splitter probably won't solve all of the potential issues. That Radio Fibre over Glass (RFoG) Optical Network Unit (ONU) runs hot. That is to say, that the coax signal level output is too high for the connected modems. It appears to be standard practice for the techs to install an 8 port splitter which drops the inbound signal level by 11 dB. The unfortunate side effect is to push up the outbound signal levels by 11 dB, which is why your upstream signal levels are so high. I wouldn't expect to see any signal level drift, upwards or downwards. The ONU should hold the signal levels pretty steady, in both directions.
If you remove the splitter, that will result in the downstream signal levels returning to the output level of the ONU. You can expect to see levels reaching + 9 dBmV, which is too high for the modem. In theory the modem has a range of +/- 15 dBmV. In practice, on the Rogers network, its typical to see issues with internet performance when the levels are outside of +/- 6/7 dBmV.
The upstream levels would drop down to 38/39 dBmV, which is within the typical 36 to 40 dBmV range.
The temporary solution would be to use a 6 dB attenuator or splitter that can provide that signal drop. Do you happen to have any attenuators or splitters around that you could use in place of the 8 port splitter?
The ideal solution, at this point in time is a Multicom – MUL-FPA85 – Forward Path Attenuator. That's an attenuator which drops the signal levels between 300 to 1200 Mhz and leaves the lower frequencies alone. That attenuator would be acceptable if Rogers were to switch your neighbourhood node to a mid-split configuration which means that the upstream frequencies would run from 5 to 85 Mhz instead of the current configuration of 5 to 42 Mhz. The ONU can run a mid-split configuration, so, its just a question of when Rogers decides to use that configuration for your neighbouhood node. Here's the splitter:
https://www.multicominc.com/product/multicom-mul-fpa85-forward-path-attenuator/
Here's the spec sheet:
https://www.multicominc.com/wp-content/uploads/MUL-FPA85-Forward-Path-Attenuator-Spec-Sheet-3_21.pdf
That attenuator comes in a 3 and 6 db drop version.
So, thats the target, but, in the mean time, if you have another attenuator or splitter around, I'd give that a go.
Does the modem fan run continuously by any chance? I'm wondering if it is, given the upstream signal level output?
@Datalink wrote:
removing the splitter probably won't solve all of the potential issues. That Radio Fibre over Glass (RFoG) Optical Network Unit (ONU) runs hot. That is to say, that the coax signal level output is too high for the connected modems. It appears to be standard practice for the techs to install an 8 port splitter which drops the inbound signal level by 11 dB. The unfortunate side effect is to push up the outbound signal levels by 11 dB, which is why your upstream signal levels are so high. I wouldn't expect to see any signal level drift, upwards or downwards. The ONU should hold the signal levels pretty steady, in both directions.
I don't disagree with anything that you said... but I think that it would take both that splitter and a long-enough coax run to drive the upstream power to 50 dBmV. Removing the splitter would definitely make the signal to the modem rather hot... but if we remove the splitter and observe the change, we could then also provide a more prescriptive solution. For any remediation, I would prefer to stick with parts that Rogers provides.
@rmckayfleming I experienced a similar problem in my parents' home where, after cleaning up splitters, etc. from the old, previous install, the incoming signal was too strong... and during the COVID lockdown, techs would not enter homes. I got the Rogers Install tech to give me a 6 dB attenuator, I installed it, then the tech ran tests from the driveway to confirm that the signal levels were within norms.
I would send a private message to @CommunityHelps and see if they can dispatch a tech to resolve your signal issues, who can provide you with whatever parts will be needed.
Removing the splitter and connecting the modem directly to the ONU should confirm my thoughts on the signal levels with just the ONU running. That splitter drops the signal levels by 11 db in both directions, which is a considerable amount. Typically the RG6 cable run to the modem isn't terribly long and its internal to the home. That doesn't guarantee that there isn't an issue with that cable, but, the chances of a problem with it are significantly reduced compared to running an external RG6 cable back to the local tap.
The compromise solution at this point would be a straight 6 dB signal attenuator which still drops the signal level by 6 db in both directions. Not ideal, but at least a Rogers tech might have one on hand, or possibly have a couple of 3 dB attenuators that can be used back to back. Failing that, an Antronix CMC4003BH or CMC4003BU splitter could be used as its has 6 dB drops on its output ports. Again not ideal, but, perhaps a Rogers tech might have one on hand.
Ideally, you would drop the downstream signal level by 9 dB and leave the upstream signal level alone. That's where the Multicom attenuator comes into play. Unless there's been a very recent change in the techs attenuator stocks, I doubt that the contractor techs carry any Forward Path Attenuators. A Senior Tech (real Rogers tech) might have these on hand, but, don't quote me on that one.
There are numerous parts to this problem:
1. the fibre cabling
2. the ONU
3. the splitter
4. the modem
5. the signal level output from the ONU
The easiest part of this is the signal level issue. Installing a straight 6 dB attenuator would drop the upstream signal levels to a manageable level and should allow some room for any upwards drift. I asked previously about the modem fan, given that the modem is running near the top of the output signal level range. Does that fan run continuously, or just when the modem is hot enough to enable the fan? If its not supposed to run continuously and it actually does, that indicates that the modem is running hot for some reason. Just thinking aloud, I would blame that on the upstream signal level. Yes, you could probably run these modems at 50 or 51 dBmV on a continuous basis, but, just because you can, that doesn't mean that you should. Its just like any other piece of equipment, yes, you can run anything near its max output level, but, that doesn't mean that you should.
So, first step would be to take care of the signal levels, next step might be to replace the modem, with an appropriate attenuator in place. If those two steps don't cut it, then its time for a tech to look at the ONU and beyond, up to and including the neighbourhood node.
@rmckayfleming can you comment on whether or not the fan is running continuously, and, do you have underground or overhead cabling in your neighbourhood?
@rmckayfleming wrote:
It sounds like the fan is running, but not loudly. The modem is quite warm though (and it's elevated with space around it, so not cramped or anything). All of the cabling is underground.
I haven't gotten around to trying directly connecting the modem to the ONU yet. I'll try that today.
The Ignite gateways do have a tendency to run warm (https://communityforums.rogers.com/t5/Internet/Ignite-Gateway-Modem-or-Easy-Bake-Oven/m-p/495643) but this is apparently normal. (Mine is in a cool basement and mounted so that it has open airflow, and the only thing warm is the air coming out the top; the enclosure is cool to the touch.). The fans on the XB7 are also very quiet, and the gateway should shut itself down when its internal temperature gets critically high.
I'm not sure why @Datalink asked whether you have underground or overhead cabling coming into your home. I seriously doubt that the fibre coming into your home would need to be replaced. If you had coax coming into your home from outside, your signal could vary with temperature and this could vary even more with overhead cabling. If a coax cable is damaged, you could also be affected by moisture or noise ingress. You can also run into weird RF-engineering challenges to resolve if you are at the far end of a street and/or have realign weird signal levels to deal with. Also, with coax infrastructure in your neighbourhood, all the equipment, taps, connectors, etc. along the entire signal path must be in good shape, end-to-end, and performance can degrade over time. However, you have fibre coming into your home and an RFoG ONU generating the signal on the coax inside your home. Nothing complicated. Any Rogers tech should have the required parts on hand to optimize your installation for Ignite TV/Internet, where the only coax-connected device in your home is your Ignite gateway.
One other thing: while you still have the splitter installed, make sure that any unused cable outlets have terminating caps installed.
@rmckayfleming Those stats look fine to me. Ideally, you would want the downstream power levels, across all channels to average 0 dBmV but yours are below 10dBmV and well within norms... and your upstream power levels are totally fine and your error stats are clean. If I had this in my home, I would be totally happy.
If your Internet service remains stable and your error stats remain clean, no further action needs to be taken.
As I pointed out in my initial reply, there are other factors that could account for slow response times from time to time, but those would be unrelated to packet loss due to wonky signal levels.
@rmckayfleming wrote:
Alright, yesterday I disconnected the other two lines (they're just blank lines, no cap at then end just a cut cable). But network requests were being annoying again this morning, power level on the upstream channels was 50-51.5dBmV (yesterday was more like 49.5-50).
I don't know the specifics of the circumstances where "network requests were being annoying again" for you, but I'll share another story: One of my side-projects is being an alpha tester for a software developer. I have a script, that I wrote, which uses curl to fetch the latest software build. The content is hosted overseas and is fronted by Cloudflare, and I am usually the first person on this continent that downloads a new daily build. When I fire up my script, the download is SLOW... but if I terminate the download and restart it, the new build downloads immediately. This slowness has nothing to do with Rogers; CDNs, caches, load balancers, etc. can cause all sorts of weird issues in some cases.
Just pointing this out because your current problem could be application-related or due to service/server-side issues, or related to how Rogers routes its traffic, and may not necessarily have anything to do with your modem/gateway's connection to the Rogers network.
-
5053 views
-
7 Likes
-
-
