Upstream bonding channel ranging status aborted after series of Abort Response and T3 time-out log message pairs
Hi there,
I'm having an issue where I'm losing at least one upstream bonding channel's ranging about once per day, and I'm wondering if there's been any line work going on in my area that could be causing it. This started over the weekend, I believe. My connection has been pretty much rock-solid for the last several years up until this point, minus the typical reboot here and there.
My logs at 192.168.100.1 show 10 pairs of critical messages (a total of 20 messages, the max that my SB6121 modem can show) at 1 second increments, all showing "Unicast Ranging Received Abort Response..." and "No Ranging Response received - T3 time-out..." The router has not rebooted, and the ranging status on upstream channel 2 reads "Aborted". This has happened at approximately 12:15 am, 12:30 pm, 7:20 am, and 8:20 am over the last four days.
See this forum post (here) and this forum post (2nd set of logs, far down on the page) for logs that match exactly what I am seeing.
My signal levels and SNR all appear to be fine (I'd post screenshots but I'm not at home at the moment), and my internet connection is still alive.
Downstream Values:
SNR: 37 dB for all 4 channels
Power Level: 1 dBmV for 3 channels, 2 dBmV for the fourth channel.
Upstream Values (in order, channel IDs 1-4):
Power Level: 46 dBmV, 46 dBmV, 48 dBmV, 49 dBmV
Ranging Status: Success, Aborted, Success, Success
I have not noted any spikes in upstream power level or downstream SNR.
I did change one connection over the weekend: I removed a power amplifier that had been attached to the line for several years. I no longer needed the amplifier as I had already removed all extra coax splitters between the cable entry point to my home and my modem. The amplifier had previously been pushing the downstream power levels above the specified +15 dBmV, but the modem was a champ and never had an issue before. The modem is now the only device on the line. The cable between the cox entry point and the modem is new within the last six months. I have a splitter joining my modem cable and the cox entry cable, with a termination on the unused side. This problem has occurred both with and without the splitter (replaced with a splice connector at one point).
The one thing I can think of is that when I removed the power amplifier from the line (it had been there since I purchased the house several years ago), I may not have noticed some corrosion on the center conductor of the cable coming from the cox side of things. This could be the case because I know that cable is at least five years old, and sits outside. However, this might not be the case since the connection only has issues for those 10 seconds when I get a burst of T3 time-outs, a modem reboot alleviates the issue, my SNR or signal levels don't jump around, and the modem goes many consecutive hours without logging any error messages between these major events.
Tonight I will clean the connectors on both ends of my cable, and hope that helps. In the mean time, is there anyone who is having a similar experience, or is there work going on near me, or is there something I'm missing?
Thanks for the help!
Thank you for your reply, Chris.
i think the problem has been solved, and I want to document it in case anyone else stumbles across this post with the same issues.
Symptom: good signal levels, but losing bonding status on upstream channels. General weird seemingly bad internet connections that would drop altogether at times. Modem logs getting spammed with T3 timeouts and other assorted ranging errors.
Problem: inadequate or failing power supply for my SB6121 arris/Motorola surfboard modem.
Solution: replace the power supply with a new one. Use a 12VDC, minimum 1A power supply with a 5.5mm X 2.5mm barrel power connector. I used a power supply for a monitor I purchased from Amazon (ASIN# B003Z6ZR5O).
The symptoms described in my original post continued through last weekend, and I finally had an opportunity to call Cox tech support while I was having issues. I left the modem plugged in (knowing that rebooting the router would alleviate the symptom) since I wanted the tech to see what was going on (if that was going to make a difference). I didn't want the first reply I got to be "signals look great from here!" I navigated through the automated reboot-your-router phone support recording (<begin rant> look, I know that phone guide is there and built the way it is for a reason, but is there a way to bypass it? I seriously had to answer at one point "Are all the lights on your modem solid green?" With "No, two of them are solid BLUE", then almost threw my phone and laptop against the wall when I was instructed to reboot my computer. It's a good thing for Cox that they have a monopoly on decent broadband in my area, and that I still haven't found better tech support anywhere else. Ok. <end rant>).
After the phone recording, I got to speak to a professional and cheery tech who was able to send an SNMP reboot message to my modem, and see some messages or information that I don't have access too. I don't remember the exact terms she used, but she asked if my modem was plugged in via a surge protector. I replied that yes, the power supply is on a surge protector (note that the coax is not). She then suggested that I move the power supply from the surge protector direct to a wall socket, because sometimes the circuitry of the protection in the surge protector doesn't play nicely with any regulation in the power supply. I said I'd give it a shot and test over the next few days.
I'm thinking that a power conflict or resonance causing bad power for my modem is technically possible under a certain set of rare circumstances, but it's far more likely that my wall wart power supply was failing after several years of constant use and was no longer providing clean power or sufficient power to my modem. I replaced the power supply.
Since replacing the power supply, I have not received a single fault, warning, critical error, T3 thing, or problem with my modem. Internet has been rock solid at my speeds I pay for. It's possible something else in the network upstream changed at the same time, but I doubt it. I usually chalk up weird electrical things to gremlins, but apparently the new power supply doesn't have any.
Thanks again for the follow up, Chris. Cox's tech support is tremendously improved over the last several years... Except for that awful automated phone support.