Sign in with
Sign up | Sign in
Your question
Solved

Motorola SB6121 self rebooting

Last response: in Networking
Share
December 25, 2012 1:38:48 AM

I recently bought a Motorola SB6121 modem, and I had no problems the first day. On the second day it started to randomly reboot by itself.

Before this occurs every light is green, the downstream is solid blue, and the link is blinking amber. After that the link becomes a solid amber or the link LED just goes off, then around 10 seconds later it will reboot itself.

I was wondering if anyone knows why this is happening?
December 25, 2012 3:22:23 AM

Kingsdown77 said:
I recently bought a Motorola SB6121 modem, and I had no problems the first day. On the second day it started to randomly reboot by itself.

Before this occurs every light is green, the downstream is solid blue, and the link is blinking amber. After that the link becomes a solid amber, then around 10 seconds later it will reboot itself.

I was wondering if anyone knows why this is happening?



Verify power supply matches output with modem requirement, its not uncommon for a company like moto to package the wrong power adapter and say give you a 12v 750 mA one from the old SB51XX series when meaning to give you say a 12v 1A one, not sure what the SB61XX series uses but I seen to recall them needed a 1 amp versus a 750 mA dc converter.

Next story what do the logs say at 192.168.100.1?
Firmware updates? It may be doing these one or two times but that's the end of it, not consistently for along time. That's usually a signal impairment or problem with the modem. If you see a lot of T3 errors in the log [not one or two those happen when its locking in normally] then you may have a issue with noise either in your local coax network or on the plant outside causing the modem to rescan and lock trying to find better frequencies.

Possible causes are limitless, but if you have a fan try pointing it at it, maybe your location is too hot or has poor air circulation.
Make sure fittings are tight by wiggling the cable near the fitting with one hand and turning it with the other, unbinds threads when you do not have the hand torque to do it. You can also use pliers and the weight of your arm or a 7/16 wrench but do not use the tools and use muscle the ports can break easily.
m
0
l
Related resources
December 26, 2012 5:02:03 PM

So would it be wise to get a 12v 1 amp dc converter and see it that solves the problem? Or do the logs show a different problem?
m
0
l
December 26, 2012 5:09:29 PM

Open a browser and surf to 192.168.100.1, choose the 'signal' page. What is the upstream level? It should be in the mid 30's. If it's in the 40's or 50's then you have a cable side problem. Could be too many splitters, bad splitter/s, pinched cable, poorly crimped fittings, bad drop, etc. Call the cable company and have a tech sent out to check it out.
m
0
l
December 27, 2012 12:39:49 AM

Yeah it says 47 dBmV. There is a tech coming tomorrow.
m
0
l

Best solution

December 27, 2012 1:23:34 AM

Kingsdown77 said:
The power adapter is 12v 750mA.

http://chris999.imgur.com/all/

Includes logs, signal, configuration, and model number/firmware information.


From here everything looks normal, who is your provider?
Share
December 29, 2012 3:26:11 AM

Well it just ended up being the modem itself, just a bad modem. Got a new one and everything is working just fine.

Thanks anyways.
m
0
l
December 29, 2012 3:26:57 AM

Best answer selected by Kingsdown77.
m
0
l
December 30, 2012 12:37:58 AM

ex_bubblehead said:
Open a browser and surf to 192.168.100.1, choose the 'signal' page. What is the upstream level? It should be in the mid 30's. If it's in the 40's or 50's then you have a cable side problem. Could be too many splitters, bad splitter/s, pinched cable, poorly crimped fittings, bad drop, etc. Call the cable company and have a tech sent out to check it out.



I realize the issue is fixed but I want to address this misinformation. This is simply wrong. I do not know who told you 40 or 50 is high for transmit but they need to go look up the DOCSIS standards and read them.

The way the return works on cable plant is you have a specific input value at the amps [be it Line extender, Mini Bridger, or Node which in cable is where you go from fiber to coax not to be confused with nodes in like tracert or anything] in my FFO that value 18 to 20 depending on when the plant was designed and what freq it was spaced for [440Mhz, 750Mhz, 860Mhz, etc...].

Given that I need 20 at an amp, at a 23 value tap my return right out of that tap before it hits ANY customer splitters, it will be 43. In comparison, later down the line where its at the end, it will have a 4 value tap, at that tap I would estimate 24, in reality it will be about 28 due to insertion loss between taps and what not, but 24/28 is VERY LOW, and if you run a line for just internet from that tap you will be forced to raise the return levels with either a PAD or an equalizer as too low of a level is just as bad as too high on the return. But in the end what really matters is the SNR. If you have 54 and can account for why its that high through plant design and passive loss at the house it should actually work fine. Comcast currently want it between 35 and 51 for installs, and the reason why is cable fluctuates with the changes to the physics of the world from moment to moment. This leaves some play room for when that happens.

Summary: 40 would be high at a 4 value tap [though there is a way to have it legitimately in plant design], but it wold be excellent at a 23 value tap, so the value you have depends entirety on where you are connected in a run, how much passive loss you have in your coax home network, and what the plant is designed for at the amplifiers. Install specs from Comcast current states 35 to 51 for return, DOCSIS standard says 30 to 54 for multiple upstream channels, 30 to 57 for single channel. More about the standard can be found at http://cablelabs.com/.
m
0
l
December 30, 2012 1:05:55 AM

Sounds like you don't work the subscriber side of things. The numbers I gave are the numbers used by Cox in my area. The Motorola transmitter maxes at +52dBmV, at which point the modem is literally shouting at the top of its lungs to be heard. Cox tries to keep this in the mid 30's. They also try to keep downstream at the modem between 0dBmV and +16dBmV.
m
0
l
December 30, 2012 2:05:29 AM

http://www.dslreports.com/faq/coxhsifaq
http://www.dslreports.com/faq/3692

For non-DOCSIS 3.0 modems:

Downstream power level (Receive): -15dbmV to +15dbmV
A value of -15 or worse indicates a poor downstream signal path. A tech would aim for a value close to the optimal 0 dBmV, but a good cable modem should be capable of working within the broader range of -15 to +15 dBmV, provided the downstream Signal to Noise Ratio remains good enough.

Upstream (Transmit): 35-52 dBmV. A value within the range +35 to +52 dBmV is within spec with the low to mid 40's the most common. If the cable modem is going offline, and the upstream signal strength is at or above +52dBmV, then a poor upstream path is probably the problem.

Carrier/Noise Ratio: 35 db and above The downstream Signal to Noise Ratio must be 35 dB or higher. The lower ratio the more noise and the poorer the performance. The Cable Modem will have to keep requesting retransmissions of packets with uncorrectable errors.

DOCSIS 3.0 modems:

Downstream power level (Receive): -15dbmV to +15dbmV
A value of -10 or worse indicates a poor downstream signal path. A tech would aim for a value close to the optimal 0 dBmV, but a good cable modem should be capable of working within the broader range of -10 to +10 dBmV, provided the downstream Signal to Noise Ratio remains good enough. Additionally all downstream channels should be within a +/-2db window with respect to each other.

Upstream (Transmit): 40-52 dBmV. A value within the range +40 to +52 dBmV is within spec with the low to mid 40's the most common. If the cable modem is going offline, and the upstream signal strength is at or above +52dBmV, then a poor upstream path is probably the problem.

Carrier/Noise Ratio: 35 db and above The downstream Signal to Noise Ratio must be 35 dB or higher on each channel. The lower ratio the more noise and the poorer the performance. The Cable Modem will have to keep requesting retransmissions of packets with uncorrectable errors.

===========================

What you said
It should be in the mid 30's. If it's in the 40's or 50's then you have a cable side problem.

Also what you said
Sounds like you don't work the subscriber side of things.

I am a comcast level 4 install repair tech who also knows how to work on plant.
The cox tech who told you you have to be in the 30's may have meant your specific house due to the tap value you run off, or maybe
you misunderstood and he/she was talking about SNR which for down you want 35+ and up you want 30+, whatever the case, the info
as you posted it is incorrect for DOCSIS specifications, which are published and certified on equipment as meeting them.

DOCSIS 3.0 specification - its public for the world to see
http://www.cablelabs.com/cablemodem/specifications/spec...

The specific part we want is here
http://www.cablelabs.com/specifications/CM-SP-PHYv3_0_1...

page 93 [their numbering not necessarily acrobats], you are concerned with qam64 upstream at this day and age, you will find 57 for multichannel mode disabled, and 54 for it enabled. That is the MODEM's actual capabilities, in order to be certified as a D3 modem, it is also the cable systems capabilities. Unfortunately that only works in a void, real world physics make the techs have to lower that range, hence why cox says 40 to 52, I am guessing their plant must run 23 to 14 value taps and then insert an LE or they EQ the heck out of it after that, which is a overall good design for managing noise/ingress. In either case, their actual install specs appear to exclude the range you said it should be in completely by starting at 40 and not including anything in the 30 range at all.

Now would being in 30's be that bad, that depends on you and your neighbors, but home owners tend to limp wrist cables when they put them on, and so they generate noise for the system, low value taps remove less signal going into the plant. A well places EQ can remove the low frequency return noise and force the modem to transmit higher improving not only your SNR on the upstream but the entire node's [more signal less noise] as well as less beats from over driven low frequencies, actually I am sort of envious of plant designed for 40 to 52, that would be a hell of a lot less noise sensitive than ours is. Regardless, the values you supplied and the conclusion you drew about what I do in cable appear incorrect.

I did not mean to offend you, simply correct a misconception you might of had either from a poorly trained tech or CSR or possibly a confusing conversation.
m
0
l
December 30, 2012 3:58:36 AM

Quote:
I did not mean to offend you, simply correct a misconception you might of had either from a poorly trained tech or CSR or possibly a confusing conversation.

I contracted with both Cox and Time Warner Cable, installing modems back when two techs were required (10-15 years ago, pre docsis). The numbers I quoted are the ones in current use by Cox system wide in this area. I'm not here to get into a "whose is bigger" contest, but I've been around since 5MB hard drives were the size of a refrigerator, and a 110 baud modem was greased lightning. I understand the numbers.
m
0
l
December 30, 2012 3:47:49 PM

OK, keep telling people under 40 then. If the actual specification that is used to certify the modems and the cmts is not going to convince you, then nothing will. Have a nice day.
m
0
l
!