@Corx, doing a little google-foo, silver has a 53.6% greater conductivity than gold, fwiw. That comes from the following site:
As for the T-3 timeouts, I wouldn't consider them to be an issue unless they were accompanied by other issues such as an MDD timeout, and obvious modem disconnects. Even then, I've seen MDD timeouts that accompany modem disconnects, and some that don't result in any drop in modem performance. You have some MDD timeouts in your logs as well. Checking my modem logs, I have 6 T-3 timeouts spaced over two weeks or more. So, not a huge issue and I don't pay any attention to the modem log unless there is some issue going on. I keep an eye on the modem signal levels and their ok. If there was an issue going on, I'd see it in my ping test results.
Let me draw your attention to the following Volpe pages. Have a look at the T-3 section on the following page:
Then, take a read thru the following page:
There's an interesting section further down the page:
T3 Timeout No RNG-RSP in 200 msec - Typically an Upstream Problem
1. Most likely cause is upstream impairments preventing the RNG-REQ from reaching the CMTS, therefore the CMTS never transmits a RNG-RSP
2. May occur because the RNG-REQs from multiple cable modems collided, again the CMTS does not receive the RNG-REQ (now from multiple modems) and fails to process multiple RNG-RSP messages resulting in multiple T3 timeouts
3. CMTS over-utilization - unable to process RNG-REQ and/or RNG-RSP within T3 time frame (200 msec)
4. Can also be a downstream problem due to plant impairments - i.e. RNG-RSP is transmitted, but corrupted in route and the cable modem never receives the message
Look for low MER at the cable modem
Looking back at your signal levels, the downstream levels are a little higher than what I'd like to see, but, their very ok. The signal to noise ratios (MERs) are fine. Your upstream is a little higher than normal, they normally run in a 36 to 40 dBmV range, with a maximum level of 51 dBmv (Rogers uses 52 dBmV). So, the upstream is a little higher than normal but well within specs. There's no reason at this point to suspect a signal level or signal to noise issue with your modem. But, you would have to call tech support and ask the CSR to run a signal check on the modem to confirm that. The specific question here is whether or not the upstream channels are within spec when they arrive at the neighbourhood node. If they aren't, then a tech visit would be required to check the external cable and connectors. If they are in spec, as I suspect they would be, then the answer is no's 2 and 3. The CMTS is busy and just doesn't have time to respond to all of the Ranging Requests at that particular moment in time. So, the modem raises a flag, logs the event and carries on with the Re-registration process. At some point, yet another Ranging Request goes out, either successful or unsuccessful.
Hopefully that explains what you're seeing. I really don't pay much attention to my modem logs. If's there's an issue, then I'll look at it. I caution users from paying too much attention to the logs and attempting to figure out what every entry means. You can drive yourself crazy trying to do that. Take a glance at it every once in a while, just to know whats normal and what isn't, and let it do its job. If something comes up that affects the modem performance, look at both signal level table and the modem logs to see whats up and post a query if you're not sure that the problem might be.
Looking back at your signal levels, and taking into consideration that you're using a CGN3ACSMR, I don't believe that swapping the connector would have had much effect if any, but that's with the frequecies listed in your post. If the modem had at some point attempted to use the channels that are higher than 750 Mhz, then you might have run into problems. That's hard to tell without more signal data from before and after. As it is now, you shouldn't run into any issues with the connector and use of the higher channels.
This does raise one question however. Can you have a look in your basement, where the house cables converge and see if you can determine what cable type you have in your home, RG-59 or RG-6. That will be printed on the cable jacket. RG-59 is an older cable and has a higher loss at higher frequencies. RG-6 is typically used/installed in homes these days. It all depends on how old your home is, but, just thinking about your gold F-81 connector, if that's been on the wallplate for a while, I wonder what cable type is installed? Hers's a reference page for RG-59 versus RG-6:
@Datalink , thank you so much for this detailed response. It's reassuring that you're getting T3 errors and not paying attention to them. I was just concerned as I get one every 5-6 hours, but I suppose that's just the "normal" for where I'm at...
A tech stopped by today when my internet cut out and he couldn't get the modem to reconnect until he did some tinkering with the coax running from the wall outlet to the actual modem (the cable Rogers supplied). The modem then reconnected, but was showing lots of interference and he is a bit concerned about my upstream signal strength. Mine was at 48dBmV while the neighbours were around 40-44. Do you know what might be causing this increased signal strength and interference for me?
Shortly after he left, I had noticeable spikes in ping. I contact tech support, they ran a signal check and advised me that my upstream signal to noise ratio is out of range and had two spikes in signals today. They've asked me to exchange my modem tomorrow (I'm going to pick up a CODA modem - I think these are better?).
The tech seems to think my issues (interference) might be because of the optical terminal that changes the fiber into coax in the garage. He's going to come back tomorrow to replace it as well as redo all the fiber connections in the house and at the box on the side of the house (I'm in a townhouse).
Do you think, though, that all of this could be caused by a bad fiber cable running from the outside box to my drop in the house? From what I've been told any issues with fiber would be noticeable right away (i.e., the internet not working at all), but thought I would ask you as you're very knowledgable!
To answer your question, I have RG-6 cable in the house. It's a brand new house (moved in a few weeks ago).
@Corx , ok, you have fibre to the home (FTTH) which does change things. The way that Rogers uses FTTH is to replace the traditional copper cabling that runs from the neighbourhood node to the home. In the home, the fibre cable is plugged into an Optical Network Terminal (ONT) and then the ONT is connected to a traditional modem via RG-6 cable. At the other end, the fibre cable connects to the neighbourhood node. I don’t know if the output card at the neighbourhood node has fibre connector ports, or if Rogers uses a copper to fibre converter to connect a copper output port to the fibre cable. That’s a question that’s worth asking tomorrow when the tech is around, just for the sake of knowing how this system works, end to end. Ideally it would be a fibre output port at the neighbourhood node that connects to a fibre cable, that connects to a fibre modem, bypassing any middle step with the ONT and cable modem. This might be a hybrid, fibre connector port and cable at the neighbourhood node and ONT to copper modem at the home. Bell uses a fibre modem, the Home Hub 3000, which the fibre cable connects to, and then the modem supplies both Ethernet and Wifi services, essentially an all in one modem.
With that fibre system in place, I would expect your signal levels to be almost exactly the same, that is, the download levels and signal to noise ratios don’t vary, and upload signal levels don’t vary. Its possible that the downstream levels would run high, such as what was shown in your previous post. The signal to noise ratios should be almost perfect, around 40 dB. Yours are running at 38.605 or 38.983. They’re not all the same, but it’s either one or the other. Very close to 40 dB and very consistent, which is what we’ve seen with previous FTTH installations. The upstream levels are a bit of a mystery. I’d expect to see them running in the 36 to 40 db range, if not slightly lower. That’s due to the short run from the modem to the ONT. You shouldn’t have any signal drop which would cause the Cable Modem Termination System to command the modem to run at a higher output power level. So, that says one thing to me, that either there’s an issue with the RG-6 cable that runs from the ONT to the output cable port in the wall, or that there is some issue with the fibre cable itself, either at the home or neighbourhood node, or a combination of both. The tech will have to ensure that the fibre cable installation is up to spec and that any fibre cable splices are ok.
Is there a reason why the ONT is located in the garage? Would there be a better place inside the home, and, is there more than one fibre drop in the home, which would allow you to relocate the ONT? The garage is a fairly dusty environment, not that it should make any difference. Once the fibre cable is plugged into the ONT, there shouldn’t be any issue of dust getting into the optical terminal. The only other consideration is the temperature of the garage in the summer. The ONT shouldn’t have any issues with the temperatures, but, if I had any choice I’d have the ONT installed somewhere the home.
Ok, getting back to the RG-6 cables:
Ok, there’s some food for thought. Hope this helps. If you can connect the modem to the ONT directly, that would help to determine if there is any signal loss in the house cabling, or if that loss is on the fibre side of the system.
Edit: @Corx are you running any type of surge suppressor where the cable runs thru the suppressor? Or, is the modem connected directly to the cable port on the wallplate?
@Datalink , thank you again! I've only had time to do some of what you suggested, I'll try more later. What I did was plug the modem straight into the ONT. What I noticed was a huge jump in my downstream signal levels and a decrease in my upstream signal levels. Any thoughts about this?
When I plug my modem into any other coax outlet in the house, my levels remain the same as above in my other posts. I'm thinking it might then be a splitter problem in the wiring hub in the garage or the signal coming into the house isn't configured properly?
Also to answer your question re: the garage: I'm in a townhouse with no basement, so the fiber drop was placed in here .
|Port ID||Frequency (MHz)||Modulation||Signal strength (dBmV)||Channel ID||Signal noise ratio (dB)|
|Port ID||Frequency (MHz)||Modulation||Signal strength (dBmV)||Channel ID||Bandwidth|
|1||23700000||ATDMA - 64QAM||41.750||2||6400000|
|2||30596000||ATDMA - 64QAM||40.500||3||6400000|
|3||13696000||ATDMA - 64QAM||40.000||1||6400000|
@Datalink , sorry for the second reply, but I picked up a CODA modem this morning and plugged it in directly to the ONT. Here are my new signal levels (high downstream, low upstream). I'm thinking the neighbourhood node isn't configured correct, or there's something wrong with the ONT. These levels make sense when compared to my original levels when the modem is plugged into the structured wiring set up to the house. When the tech installed the internet in the first place, they put a -7.5dB splitter in (which explains why my downstream are about 5-7db lower when th emodem is plugged into the coax wall outlet and not the ONT).
|Port ID||Frequency (MHz)||Modulation||Signal strength (dBmV)||Channel ID||Signal noise ratio (dB)|
|Receiver||FFT type||Subcarr 0 Frequency(MHz)||PLC locked||NCP locked||MDC1 locked||PLC power(dBmv)|
|Port ID||Frequency (MHz)||Modulation||Signal strength (dBmV)||Channel ID||Bandwidth|
|1||13696000||ATDMA - 64QAM||33.500||1||6400000|
|2||30596000||ATDMA - 64QAM||34.250||3||6400000|
|3||23700000||ATDMA - 64QAM||34.250||2||6400000|
@Corx no need to apologize for any posts. The forum is here to ask questions, so, don't hesitate to ask, again, and again, and again
Your signal levels make perfect sense. The main problem is that the cable output of the ONT is too high for the modem. The DOCSIS design limits for the downstream levels for all modems are +/- 15 dBmV. So, the tech has a balancing act to perform, drop the downstream levels without driving up the upstream levels to the 51 dBmV failure point. Your original signal levels, with the splitter in place are a balance point that achieves that. The black Hitron CGN3xxx modems normally operate somewhere in the 36 to 40 dBmV range for the upstream output power levels. The white Hitron CODA-4582 operates in the 30 to 33 dBmV range for the upstream output power levels. In both cases, when you plug the modem into the wallplate cable port (with the splitter in place), the downstream levels will drop in comparison to the ONT output, and the upstream levels will rise, in comparison to the ONT input. That's the result of the splitter, and at the signal levels that you're seeing, they're fine.
There are two ways to accomplish the signal reduction, one is to use a splitter, the other is to use a Forward Path Attenuator. The splitter will result in a drop in the downstream levels, but also result in an increase in the upstream levels. The Forward Path Attenuator (FPA) will drop the downstream levels and leave the upstream levels untouched. I don't believe that Rogers uses any FPAs, simply resorting to splitters when necessary. Only caveat is that any unused ports on the splitter should have a 75 ohm terminator installed. That's to prevent any signal reflection on those ports. Again, I don't believe that Rogers installs the terminators.
Here's a couple of links for the FPA and terminator's:
So, where your signal levels will end up, with the 4582 installed will be fine. I wouldn't worry about them at all. Only item to consider would be to order a pack of terminators and install them on any unused ports on the splitter or on any connected wallplates that are also not in use.
I would say that any issues that you see in the modem log will be the result of issues with the fibre cabling, if any, or with the neighbourhood node or CMTS. As I indicated before, I don't look at the log all that often so I'm not really concerned by its contents. The DOCSIS system is pretty robust will rerun the failed event as part of a regular schedule, if not every minute, then every few minutes. The term that I would use is "cranky", when it can't carry out its normal task, so it complains but carries on, normally without any effect on modem performance.
What you might want to do, if you're not the individual that attached the cable connectors on the internal cabling, is to unscrew the cable connector from the backside of the wallplate F-81 connectors, and have a look at the end of the connector, looking end on, so that you can see inside the connector. What you should see is that the center copper connector is basically flush with the end of the outside connector, ie: the same length or perhaps just slightly longer. We're talking a couple of milli-meters or so. You should be able to see the dielectric that surrounds the copper conductor. Around that, and not visible will be the outside aluminum shielding. That could be a braided aluminum shield, or it could be an aluminum foil. You might see that wrapping back along the outside nylon cable cover, extending slightly beyond the end of the connector, down the cable. What you don't want to see is any aluminum braid, or foil contacting the inner copper conductor, as that will ground out the cable and cause signal loss for that particular cable. As a new home owner, its worth the piece of mind to have quick look at those cable ends so that you know their ok and good to use at some point down the road.
Ok, that should do it for now. Please let me know if you have further questions
@Datalink , sorry for the late response to this. All connectors look fine in the house. Things are really strange. My WAN hasn't cut out for a few days now, but what I'm finding is my internet seems rather sluggish while browsing the most basic of websites (even trying to send emails is difficult sometimes). When I run speedtests (using speedtest.net and fast.com) my download speed is fine (300Mbps, what I'm paying for), but my upload seems iffy. Sometimes it'll peak and stay constant at 20Mbps, but other times it'll peak at around 15-16Mbps then immediately drop off to 1-2Mbps, or even to 0Mbps. It has dropped to 0 while on the phone with a tech.
Just for the record, when I run a speedtest on fast.com, my "unloaded" ping is around 10ms and my "loaded" ping is around 500-600ms.
I was told by the tech on the phone that there is noise on line 1 of my node. I had a support ticket submitted to have this resolved, but got a response back from the backend techs telling me that no issue could be identified. I called back and the next tech I spoke with told me the interference/noise had subsided and was only spiking periodically and wasn't constant anymore - my internet still feels really sluggish. Every time I speak with a tech on the phone they identify a problem, but then my ticket comes back suggesting no problem can be identified. Ugh, frustrating!
I've had a supervisor tech come out and check my lines - he noticed a bit of interference but said it was minimal and shouldn't affect anything. Everything just seems weird though as there is very clearly and upload issue nobody is capable of fixing at the moment. Any thoughts?
Edit: For what it's worth, I've been running ping tests on google's DNS, and I get spikes up to 1000ms. If I connect to my neighbours wifi (who kindly let me run some tests) who uses Bell, my max ping is only around 100ms for a test of the same duration (roughly 5 mins or so).
@Corx , ok, there's something going on here that doesn't make sense. Only thing I can think off is a congestion issue at the CMTS. We'll get to that. As for the interference, you're on fibre to the home, so that leaves two sources of interference, at the home, or at the neighbourhood node. There might be a mid point amplifier in the works, don't know. At the present time, I don't understand where any interference might be entering the system due to the fibre optics.
So, for now, when you have some time available, I'd like to tackle the possibility of packet loss. Can you download HRPing from: https://www.cfos.de/en/ping/ping.htm
When you've downloaded and parked the program somewhere, run a trace to anywhere, www.google.ca for example. The first hop should be the modem's IP address, the second IP address should be the Cable Modem Termination System (CMTS).
Now, to run a continuous pingtest at 250 milli-second intervals to the CMTS, use the following command:
hrping -t -T -s 250 xxx.xxx.xxx.xxx where xxx.xxx.xxx.xxx is the CMTS IP address
Use Ctrl + C to bail out of the test. The -s variable is in milli-seconds. The default is 500 ms.
To dump the results to a file, run: hrping -t -T -s 200 -F pingtest.txt xxx.xxx.xxx.xxx
That will create and write to the pingtest.txt file in the same directory as the HRPing executable, although you can point that to any directory and use any file title that you prefer. Use filename.txt, with the .txt file extension so that it can be opened automatically with notepad. Large 24 hour file results should be opened with Notepad++, which easily handles large file sizes. The time intervals can be adjusted up or down.
To run that for an hour, which is what I'd like you to do, you can use:
hrping -n 14400 -T -s 250 xxx.xxx.xxx.xxx
What I'd like to see is the bottom results group when the test is done, just to see what turns up. What you could consider doing is a 24 hour ping test to the CMTS, looking for packet loss. You can run that with a normal ping command which will run with 1 second intervals. Again, I'd be interested in the bottom results after the test completes. If you prefer to run the 24 hour test, skip the HRPing download, unless of course running a high speed IPV4 ping test might be useful for you.
To copy the bottom results you can open the pingtest.txt file and copy the bottom results lines. To copy the command window, right click on the title bar of the command window, select Edit .... Select All. Then right click again on the top title bar, select Edit ..... Copy. Then you can paste that somewhere, trim it down to the bottom test results and then post the results.
You shouldn't have any packet loss, but, I'd like to confirm that. A one hour quick test is just a snapshot, I'd like to see a 24 hour test if you could run that.
After that, I'll get you to run a ping test to the Rogers DNS, both ICMP and TCP. With a little more info on hand, hopefully we'll have some idea of what the problem might be.
@Datalink , sorry but how do I get hrPING to work? I've downloaded the zip file to my desktop, extracted it, clicked the executable went through the license agreement, then the window closed. When I try to open the exectuable again a command prompt quickly opens then closes (I'm trying as administrator on Windows 10).
I took the second IP (IPv6) from my tracert and did a ping test to it for a bit - maybe 15 minutes or so (not using hrPING as I haven't figured that out yet). Here's the results:
Ping statistics for xxxx:xxxx:xxx:xx::x:
Packets: Sent = 969, Received = 963, Lost = 6 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 7ms, Maximum = 2309ms, Average = 51ms
Hi @Corx no problem. You need to use a command prompt to run the program. To do that, go to START .... PROGRAMS .... WINDOWS SYSTEM. Right click on the Command Prompt and select Run as Administrator. You will need to run the Command for HRPing as an Administrator for the first run in order to acknowledge the licence agreement. After that you can just select the command prompt to run normally.
With the command box up, navigate to the directory where you parked the HRPing files. You will need to use the DOS change directory command, ie: cd xxx where xxx is the directory that you want to go to. Sometimes you have to go there in steps if you can't access the directory directly. When your in that directory you can then run the command to ping the CMTS. Type in, or paste in:
hrping -n 14400 -T -s 250 xxx.xxx.xxx.xxx where xxx.xxx.xxx.xxx is the CMTS IP address
You need the command box up in order to run the trace to obtain the CMTS IP address. Type in, or paste in:
tracert -4 www.google.ca
That will run an IPV4 trace to www.google.ca so that you can obtain Hop #2s IP address for the ping test. HRPing is only IPV4, unfortunately.
With that command box up, you can right click on the top title bar of the box and then use the Edit .... Select, Edit .... Copy commands in order to copy the contents of the command box and paste it elsewhere.
Please let me know if you have any problems with any of the steps.
Edit: Looking at your 15 min test, that's not a great result. You shouldn't have any packet loss and your return times shouldn't be beyond 80 or 90 milli-seconds. The normal times should be in the 8 to 13 ms range with high times in the 80 to 90 ms range due to the timing issue within the modem. When you look at the data, is the high time recorded at the very first ping, which can happen, or are you seeing ping times in the hundreds or thousands scattered in the test results?