dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
1049
perfectimage
join:2013-07-25
North York, ON

1 edit

perfectimage

Member

[Cable] New service and 476 modem.Are my stats ok ?

Just got connected today and hooked up my new Technicolor DCM476 STAC.02.50.
Could someone please comment on the stats of my modem as I haven't a clue as to good or bad?
And if someone could point me to where I could learn what these stats actually mean.Newbie.
Thanks in advance.

Status Code: Operational
Software Version: STAC.02.50
Software Model: a81a
Bootloader: 2.3.1

Cable Signal Details
Forward Path:
Channel Frequency Power SNR BER Modulation
1 621.0 MHz 6.4 dBmV 41.4 dB 0.000 % 256 QAM
2 615.0 MHz 6.2 dBmV 41.4 dB 0.000 % 256 QAM
3 633.0 MHz 6.8 dBmV 41.3 dB 0.000 % 256 QAM
4 639.0 MHz 6.9 dBmV 41.3 dB 0.000 % 256 QAM
5 645.0 MHz 6.9 dBmV 41.6 dB 0.000 % 256 QAM
6 651.0 MHz 6.9 dBmV 41.1 dB 0.000 % 256 QAM
7 657.0 MHz 6.9 dBmV 41.3 dB 0.000 % 256 QAM
8 663.0 MHz 6.9 dBmV 41.3 dB 0.000 % 256 QAM
Return Path:
Channel ID Frequency Power Modulation
1 31.3 MHz 41.5 dBmV 16 QAM
2 25.3 MHz 41.5 dBmV 16 QAM
3 22.1 MHz 41.5 dBmV 16 QAM
4 38.6 MHz 43.0 dBmV 16 QAM
Data Service Details
Provisioned Address: Yes
Provisioned Time: Yes
Provisioned Configuration: Yes
Registered: Yes
BPI: Enabled

sbrook
Mod
join:2001-12-14
Ottawa

sbrook

Mod

That's fine.
perfectimage
join:2013-07-25
North York, ON

perfectimage

Member

Thank you.
DigitalRain
join:2013-03-16

2 edits

DigitalRain

Member

I'm far from a broadband engineer or an applied mathematician, but I'll do my best.

Forward path power:

The forward path power level, or downstream power, is the strength of the signal that your modem receives from the Cable Modem Termination System (CMTS), which you may also know as the head end, the point of interface (POI), or the cable/RF plant. This is the point where the signal leaves the local cable provider's network and journeys on to the Internet.

The ideal value is 0.0 dBmV, which represents a 1 millivolt signal traveling along a coaxial cable with a 75 ohm resistance. Values away from this ideal have a logarithmic (base 10) relationship with it, such that going from 0 dBmV to 10 dBmV represents a 10 fold increase in signal strength, while going from -10 dBmB to 10 dBmB is 100-fold increase in signal strength. Some modems, such as your DCM 476, also use 0.0 as a placeholder value for a channel that is not receiving signal. This practice should be avoided, in my opinion, because it creates misperceptions among end-users and among front line technical support representatives as to what 0 actually represents. My Motorola SB6120 does not show dead channels at all, which imho is the better way to do things.

The DOCSIS 3.0 specification used by modern cable modems and the networks that they operate on requires that modems be able to operate within a range of -15 to 15, which means that the weakest signal they are able to use stably is about 1,000 times stronger than the strongest. Some cable providers, such as Rogers, only guarantee stable operation between -11 and 11. Values outside of the acceptable range will result in packet loss, slow speeds, and intermittent outages, with more severe divergences preventing the end user from being able to get online at all. Unacceptable downstream signal levels can be caused by problems anywhere on the cable network, from the CMTS to the client-side modem.

Overly strong signals can often be resolved by the end user using a splitter, an attenuator, or other equipment capable of either weakening or dividing the signal. Just keep in mind that anything that manipulates the signal may have negative side effects on the quality of the signal, which I'll describe in more detail in the section below. A signal that is overly hot when it enters your home to begin with may need to be looked at by a cable technician, however.

Weak signals can be caused by problems anywhere between the cable plant and the client modem. Client-side weak signals tend to be caused by an overly-divided signal (i.e. too many splitters, bad splitters, or too many output ends on a splitter), bad cabling, bad connections/connectors, or -- less commonly -- a bum modem. An overly weak signal that cannot easily be rectified by addressing problems with client-side equipment generally indicates a more widespread Internet outage or other more serious infrastructural problem that needs to be looked at by a cable technician and/or the cable company employing them. If ever it comes to this, call your ISP and get 'er done! Keep in mind that even an extremely localized issue with signal strength can indicate a serious problem requiring the intervention of a technician.

Two last thing to bear in mind: First, adding any piece of equipment between your home cable run and the modem (particularly an amplifier or a splitter with a built-in amplifier) can result in signal degradation, which can become an issue any time the signal entering your home is noisy to begin with. Second, because a CMTS has to serve thousands of customers at once, it is difficult for individual clients to receive an ideal signal without some degree of client-side signal strength manipulation.

So long as you can do so without adding too much noise to the signal, do your best on your end to make sure that your modem is receiving a signal as close to 0 dBmV as possible. While useless for dealing with severe outages caused by cable cuts or other infrastructural problems, it will act as insurance against anything that may cause a mild disruption in the signal sent from the CMTS to your modem.

Which brings us to...

Forward path SNR (downstream signal-to-noise-ratio):

Whereas power measures the strength of the signal, SNR measures its quality or its fidelity, expressed in terms in terms of the ratio of usable signal to noise. An SNR of 0 dB represents an equal proportion of usable signal to noise (or a 1:1 ratio), which is pathetic. In the real world, you'll only ever see 0 as a placeholder value when a Thompson modem is not receiving a signal on a given channel. Conversely, an SNR of 40 dB indicates that there is a 10000:1 ratio of signal to noise.

The SNR required depends of the type of modulation used on the downstream signal. Since a 256 QAM downstream signal is ubiquitous on modern cable networks, whether the client-side modem is DOCSIS 2 or DOCSIS 3, 32 SNR is generally considered an acceptable minimum, with values of 35 or higher being ideal. Values of around 42 - 43 DB are typically the highest you'll see. It is possible to have a noisy signal of close to ideal strength, in which case you need to take a look at your home cable set-up or have a cable technician take a look at your line, as I mentioned above. Conversely, it is also possible to have a signal that's either too weak or too strong, but that is neither noisy nor dirty, in which case you'll need to go back to the section above.

Return path power (upstream power):

Whereas downstream power refers to the strength that that the cable modem receives from the CMTS and SNR refers to the quality of said signal, upstream signal indicates the strength of the signal send by the cable modem to the CMTS. The strength of the signal emitted by the cable modem on any given channel is determined by a process known as ranging, whereby the client-side modem, which starts by emitting a signal of 8 dBmV, increases the strength of the signal it emits until the CMTS tells it that it's reached an optimal value, usually the upstream signal strength required for the CMTS to receive a signal of 0 dBmV, or, at minimum, as close to 0 dBmV as possible without forcing the cable modem out of specification. Cable companies and ISPs aim for an upstream power of between 33 and 54 on DOCSIS 3.0 (with some favouring a slightly more narrow range), and you generally want this to be on the lower end of that range, say in the 30s or low-40s. Like virtually any piece of electronics, a cable modem that works less hard tends to run cooler and live longer.

Values that are too low are not particularly common, but can be a sign that the cable modem is having trouble emitting a sufficiently strong signal to accommodate the CMTS (which generally means that the modem's not doing so hot) or that something is up with the cable network that is resulting in the cable signal being insufficiently attenuated between the modem and the CMTS. As the latter can cause problems for some users, cable companies and ISPs generally like to have this issue looked at.

Values that are too high are much more common, and generally mean that something on either the client side (a bad splitter, for instance) or on the infrastructure side (something that overly attenuates the signal between the modem and the CMTS) is forcing the modem to work harder than it is meant to. This will eventually causes the modem to start its ranging over from 8 dBmV or to give up all together Somebody else will have to fill in the blanks on this one.

Since one CMTS has to provision many modems while each individual modem is free to adjust its own signal to please the CMTS, it is far easier for a modem to accommodate the needs of the CMTS than it is for the CMTS to accommodate the modems that it provisions. As a results, modems are more tolerant of suboptimal signal than is the CMTS.

Return path (upstream) SNR:

Who cares? The signal leaving your modem will have at most a trivial amount of noise.
souss
join:2010-12-15
Mississauga, ON

souss to sbrook

Member

to sbrook
I might as well ask you instead of making a new thread about this.

Status Code: Operational
Software Version: STAC.02.50
Software Model: a81a
Bootloader: 2.3.1

Cable Signal Details
Forward Path:
Channel Frequency Power SNR BER Modulation
1 621.0 MHz 8.1 dBmV 40.2 dB 0.000 % 256 QAM
2 615.0 MHz 7.8 dBmV 40.0 dB 0.000 % 256 QAM
3 633.0 MHz 8.6 dBmV 40.7 dB 0.000 % 256 QAM
4 639.0 MHz 8.8 dBmV 40.6 dB 0.000 % 256 QAM
5 645.0 MHz 8.7 dBmV 40.3 dB 0.000 % 256 QAM
6 651.0 MHz 8.5 dBmV 40.3 dB 0.000 % 256 QAM
7 657.0 MHz 8.5 dBmV 40.2 dB 0.000 % 256 QAM
8 663.0 MHz 8.4 dBmV 40.0 dB 0.000 % 256 QAM

does that look fine? I get random pingspikes (especially when gaming) at times and want to make sure there's nothing wrong with my modem. I got a new router recently which almost fixed the issue but now it seems like it's back.

Here's a tracert from said ping spikes:
Tracing route to www.google.ca [74.125.226.151]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.1.1
2 793 ms 850 ms 852 ms 10.125.23.1
3 890 ms 891 ms 928 ms 24.156.137.233
4 953 ms 972 ms 981 ms 69.165.168.158
5 999 ms 1023 ms 1014 ms 24.52.255.121
6 1054 ms 1068 ms 1089 ms 69.196.136.36
7 1071 ms 1128 ms 1158 ms 72.14.212.134
8 1170 ms 1194 ms 1196 ms 209.85.255.232
9 1205 ms 1258 ms 1256 ms 209.85.250.7
10 1319 ms 1314 ms 1339 ms 74.125.226.151

Trace complete.
videonerd
join:2007-01-21

videonerd

Member

From what I've been told, the values can be vary quite wildly but with acceptable results.

Forward path: 0 dBmV, >32dB SNR
Return path: 30-50 dBmV

Your pings are definitely way high. I had a similar issue several months ago for a good month with a lot of Teksavvy/Rogers back and forth with no resolution, then it just magically went away.

My stats right now (Rogers is mucking with the POI the last few days...) but I have a 3-way splitter with a -3dB pad on my cable modem.

Forward Path:
Channel Frequency Power SNR BER Modulation
1 591.0 MHz -2.0 dBmV 40.3 dB 0.000 % 256 QAM
2 597.0 MHz -1.6 dBmV 40.8 dB 0.000 % 256 QAM
3 603.0 MHz -1.3 dBmV 41.3 dB 0.000 % 256 QAM
4 609.0 MHz -1.6 dBmV 41.3 dB 0.000 % 256 QAM
5 615.0 MHz -1.1 dBmV 41.3 dB 0.000 % 256 QAM
6 621.0 MHz -1.2 dBmV 41.5 dB 0.000 % 256 QAM
7 633.0 MHz -1.4 dBmV 41.2 dB 0.000 % 256 QAM
8 639.0 MHz -1.7 dBmV 40.0 dB 0.000 % 256 QAM
Return Path:
Channel ID Frequency Power Modulation
5 38.6 MHz 38.5 dBmV 64 QAM
6 30.6 MHz 37.0 dBmV 64 QAM
7 23.7 MHz 36.5 dBmV 64 QAM
0 0.0 MHz 0.0 dBmV QPSK
DigitalRain
join:2013-03-16

DigitalRain to souss

Member

to souss
Something serious is going on here. You're not only getting tons of packet loss between your home network and the CMTS gateway, but your latency is also growing immensely with each subsequent hop. Something stinks.

How does the upstream look? What kind of packet loss are your getting when you ping 50 times to Google? Are the problems less severe during off-peak hours? Have you tried another ethernet cable between the router and the modem? Have you tried with the router?

Here's what my trace route looks like, pardon my French.

Détermination de l'itinéraire vers google.ca [74.125.226.159]
avec un maximum de 30 sauts :

1 1 ms 1 ms 1 ms router.asus.com [192.168.1.1]
2 19 ms 8 ms 9 ms 10.125.72.129
3 14 ms 18 ms 14 ms 69.63.255.177
4 17 ms 18 ms 14 ms 69.63.253.170
5 17 ms 17 ms 15 ms 69.196.136.132
6 17 ms 17 ms 16 ms 72.14.212.134
7 15 ms 14 ms 14 ms 209.85.255.232
8 16 ms 18 ms 16 ms 209.85.250.7
9 19 ms 14 ms 15 ms yyz08s14-in-f31.1e100.net [74.125.226.159]

Itinéraire déterminé.

sbrook
Mod
join:2001-12-14
Ottawa

sbrook to DigitalRain

Mod

to DigitalRain
Digitalrain ... actually 0.0 dBmV is NOT optimal. It is simply the reference level for the dBmV scale when presented to a load of 75 ohms.

The optimal level is actually as high as reasonable within the range and depends on the interaction of signals between subscribers on the tap from the trunk.

Many cable operators aim for about -1 dBmV.

Signal to noise ratio ... that's the noise at the receiver and may originate anywhere on the transmission path. It can come from the electronics, the optics, the connectors, ingress from unterminated connectors etc.

So, the upstream SNR does matter, but because it's measured at the receiving end (i.e. at the CMTS at the cable co's head end) it's not reported to the sending modem. Techs with their tools to look at the head end can determine the upstream SNR. The CMTS adjusts the cable modem's sending power not just to get aim for optimal signal strength at the head end, but also the SNR. By raising the signal strength, the SNR will hopefully improve.

It's actually quite important since this the upstream frequencies suffer the most from ingress on the cable system.
souss
join:2010-12-15
Mississauga, ON

souss to DigitalRain

Member

to DigitalRain
I've been getting this issue on and off lately, it used to be a daily occurrence a few months ago, and now it seems like it has resumed since earlier this week.

At first I thought it was my router, which I ended up replacing with an Asus RT-N66u.

0 packetloss, but look at that insane ping.
Ping statistics for 173.194.43.119:
Packets: Sent = 181, Received = 181, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 7ms, Maximum = 1457ms, Average = 706ms
Control-C
souss

souss

Member

And now it's back to normal.

I'm going to connect directly into my modem and see if the issue remains.

Tracing route to www.google.ca [173.194.43.120]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.1.1
2 9 ms 7 ms 7 ms 10.125.23.1
3 17 ms 11 ms 10 ms 24.156.137.233
4 8 ms 8 ms 8 ms 69.165.168.158
5 11 ms 48 ms 8 ms 198.84.242.33
6 10 ms 8 ms 9 ms 69.196.136.132
7 13 ms 7 ms 7 ms 72.14.212.134
8 9 ms 8 ms 7 ms 209.85.255.232
9 8 ms 9 ms 10 ms 72.14.239.73
10 12 ms 10 ms 8 ms 173.194.43.120

Trace complete.

TypeS
join:2012-12-17
London, ON

TypeS to souss

Member

to souss
said by souss:

Here's a tracert from said ping spikes:

Tracing route to www.google.ca [74.125.226.151]
over a maximum of 30 hops:
 
  1    <1 ms    <1 ms    <1 ms  192.168.1.1
  2   793 ms   850 ms   852 ms  10.125.23.1
  3   890 ms   891 ms   928 ms  24.156.137.233
  4   953 ms   972 ms   981 ms  69.165.168.158
  5   999 ms  1023 ms  1014 ms  24.52.255.121
  6  1054 ms  1068 ms  1089 ms  69.196.136.36
  7  1071 ms  1128 ms  1158 ms  72.14.212.134
  8  1170 ms  1194 ms  1196 ms  209.85.255.232
  9  1205 ms  1258 ms  1256 ms  209.85.250.7
 10  1319 ms  1314 ms  1339 ms  74.125.226.151
 
Trace complete.
 

I would say you should check to see if anyone is uploading something long to youtube, uploading a large stack of images to Facebook or any other photo sharing network (Instagram, Flickr, Picasa, etc) or torrenting with high upload.

If you're using a router to share your connection with others, I'd disconnect and direct connect your computer to the modem for 24 hours to see if it happens when only you're using the internet connection. That way you can control what exactly is leaving and entering your modem's WAN interface.

If disconnecting the router is an unfeasible thing to do in your home (you have VoIP or someone needs access for something important), go around to their computers see whats running, ask if they upload videos or large amounts of images every now and then. Might want to consider QoS or implementing browser extensions to control upload speeds to media sharing sites like Facebook, YouTube, Instagram, etc.
DigitalRain
join:2013-03-16

DigitalRain to sbrook

Member

to sbrook
That's interesting. So in other words, it's possible for everything to look beautiful in the modem statistics, but for the CMTS to be receiving a signal that is still marginal or even unusable. Generally speaking, does this tend to result in issues that are more localized or in more widespread outages and signal degradation? Or is it something so dependent on the source of the ingress that it's difficult to establish a rule of thumb? What kind of ways (if any) would it be possible to identify this problem by looking at the modem diagnostics that are available to the end user?

sbrook
Mod
join:2001-12-14
Ottawa

2 recommendations

sbrook

Mod

The problem can be anywhere. If it's between the segment and the head end it's basically all people on that segment. If it's on a user's drop from the tap into the house, then it's probably only 1 user. If it's at a node level, or CMTS level, then they're pretty widespread.

Taking noise measurements at various points on a trunk can help determine location ... but the problem is it takes an interested and experienced tech as opposed to the hit and run techs who are used to simply stripping off and installing new connectors and cable!
souss
join:2010-12-15
Mississauga, ON

souss to TypeS

Member

to TypeS
It could be the issue, but nobody here is a 'power user' of the connection other than me. I rarely ever torrent, and if I do its overnight. I do have VoIP (Vonage) but it is rarely used.

I'm going to try connecting directly to modem tonight and see what comes about. I remember doing the same thing a few months ago and didn't experience any ping spikes. I assumed it was my old router which I immediately replaced.

I will look into QoS again, I was planning on getting Tomato on my router but it seemed like the issue went away after I replaced my old router.

TypeS
join:2012-12-17
London, ON

TypeS

Member

My dad is far from being a power user, he is actually quite incompetent when it comes to using a computer.

One day I noticed loading web pages came to painful slow crawl. Ran a traceroute to sites I was trying to get to, everything after first hop (my router) had huge latency. I ask dad what he is doing, he was uploading 100 or so images from his DSLR camera to Facebook (he is not pro photographer either, he just spend stupid money on a camera).

So you never know, maybe someone in your home went photo crazy lately like my father.
souss
join:2010-12-15
Mississauga, ON

souss

Member

My brother will torrent the occasional album or two but I always tell him to close/stop everything when he's done. I'll usually check after just to make sure. My sister will stream a movie every few days, youtube clips here and there, and mostly Facebook I guess. She could have some sort of malware on her laptop since she isn't that tech savvy. I'll have to check on that.

Everybody is at work/school during the day so if I experience this it's usually between 5pm-12am(ish).
DigitalRain
join:2013-03-16

DigitalRain to sbrook

Member

to sbrook
Conversely, though, couldn't the use of some extremely basic statistics limit the necessity of a tech being that diligent? Without needing to do all this spade work, it shouldn't be that difficult to figure out where a problem lies simply by virtue of figuring out who's affected.

sbrook
Mod
join:2001-12-14
Ottawa

sbrook

Mod

that require a level of cooperation with the telephone support staff and the networking staff and their call handling systems that simply doesnt exist. Heck, 20 people in a neighbourhood can call in to report a failure and the next call the last tech doesnt know ther are a heap of calls from that neighbourhood reporting. It is quite pathetic really!
perfectimage
join:2013-07-25
North York, ON

perfectimage to DigitalRain

Member

to DigitalRain
said by DigitalRain:

I'm far from a broadband engineer or an applied mathematician, but I'll do my best.

Forward path power:

The forward path power level, or downstream power, is the strength of the signal that your modem receives from the Cable Modem Termination System (CMTS), which you may also know as the head end, the point of interface (POI), or the cable/RF plant. This is the point where the signal leaves the local cable provider's network and journeys on to the Internet.

The ideal value is 0.0 dBmV, which represents a 1 millivolt signal traveling along a coaxial cable with a 75 ohm resistance. Values away from this ideal have a logarithmic (base 10) relationship with it, such that going from 0 dBmV to 10 dBmV represents a 10 fold increase in signal strength, while going from -10 dBmB to 10 dBmB is 100-fold increase in signal strength. Some modems, such as your DCM 476, also use 0.0 as a placeholder value for a channel that is not receiving signal. This practice should be avoided, in my opinion, because it creates misperceptions among end-users and among front line technical support representatives as to what 0 actually represents. My Motorola SB6120 does not show dead channels at all, which imho is the better way to do things.

The DOCSIS 3.0 specification used by modern cable modems and the networks that they operate on requires that modems be able to operate within a range of -15 to 15, which means that the weakest signal they are able to use stably is about 1,000 times stronger than the strongest. Some cable providers, such as Rogers, only guarantee stable operation between -11 and 11. Values outside of the acceptable range will result in packet loss, slow speeds, and intermittent outages, with more severe divergences preventing the end user from being able to get online at all. Unacceptable downstream signal levels can be caused by problems anywhere on the cable network, from the CMTS to the client-side modem.

Overly strong signals can often be resolved by the end user using a splitter, an attenuator, or other equipment capable of either weakening or dividing the signal. Just keep in mind that anything that manipulates the signal may have negative side effects on the quality of the signal, which I'll describe in more detail in the section below. A signal that is overly hot when it enters your home to begin with may need to be looked at by a cable technician, however.

Weak signals can be caused by problems anywhere between the cable plant and the client modem. Client-side weak signals tend to be caused by an overly-divided signal (i.e. too many splitters, bad splitters, or too many output ends on a splitter), bad cabling, bad connections/connectors, or -- less commonly -- a bum modem. An overly weak signal that cannot easily be rectified by addressing problems with client-side equipment generally indicates a more widespread Internet outage or other more serious infrastructural problem that needs to be looked at by a cable technician and/or the cable company employing them. If ever it comes to this, call your ISP and get 'er done! Keep in mind that even an extremely localized issue with signal strength can indicate a serious problem requiring the intervention of a technician.

Two last thing to bear in mind: First, adding any piece of equipment between your home cable run and the modem (particularly an amplifier or a splitter with a built-in amplifier) can result in signal degradation, which can become an issue any time the signal entering your home is noisy to begin with. Second, because a CMTS has to serve thousands of customers at once, it is difficult for individual clients to receive an ideal signal without some degree of client-side signal strength manipulation.

So long as you can do so without adding too much noise to the signal, do your best on your end to make sure that your modem is receiving a signal as close to 0 dBmV as possible. While useless for dealing with severe outages caused by cable cuts or other infrastructural problems, it will act as insurance against anything that may cause a mild disruption in the signal sent from the CMTS to your modem.

Which brings us to...

Forward path SNR (downstream signal-to-noise-ratio):

Whereas power measures the strength of the signal, SNR measures its quality or its fidelity, expressed in terms in terms of the ratio of usable signal to noise. An SNR of 0 dB represents an equal proportion of usable signal to noise (or a 1:1 ratio), which is pathetic. In the real world, you'll only ever see 0 as a placeholder value when a Thompson modem is not receiving a signal on a given channel. Conversely, an SNR of 40 dB indicates that there is a 10000:1 ratio of signal to noise.

The SNR required depends of the type of modulation used on the downstream signal. Since a 256 QAM downstream signal is ubiquitous on modern cable networks, whether the client-side modem is DOCSIS 2 or DOCSIS 3, 32 SNR is generally considered an acceptable minimum, with values of 35 or higher being ideal. Values of around 42 - 43 DB are typically the highest you'll see. It is possible to have a noisy signal of close to ideal strength, in which case you need to take a look at your home cable set-up or have a cable technician take a look at your line, as I mentioned above. Conversely, it is also possible to have a signal that's either too weak or too strong, but that is neither noisy nor dirty, in which case you'll need to go back to the section above.

Return path power (upstream power):

Whereas downstream power refers to the strength that that the cable modem receives from the CMTS and SNR refers to the quality of said signal, upstream signal indicates the strength of the signal send by the cable modem to the CMTS. The strength of the signal emitted by the cable modem on any given channel is determined by a process known as ranging, whereby the client-side modem, which starts by emitting a signal of 8 dBmV, increases the strength of the signal it emits until the CMTS tells it that it's reached an optimal value, usually the upstream signal strength required for the CMTS to receive a signal of 0 dBmV, or, at minimum, as close to 0 dBmV as possible without forcing the cable modem out of specification. Cable companies and ISPs aim for an upstream power of between 33 and 54 on DOCSIS 3.0 (with some favouring a slightly more narrow range), and you generally want this to be on the lower end of that range, say in the 30s or low-40s. Like virtually any piece of electronics, a cable modem that works less hard tends to run cooler and live longer.

Values that are too low are not particularly common, but can be a sign that the cable modem is having trouble emitting a sufficiently strong signal to accommodate the CMTS (which generally means that the modem's not doing so hot) or that something is up with the cable network that is resulting in the cable signal being insufficiently attenuated between the modem and the CMTS. As the latter can cause problems for some users, cable companies and ISPs generally like to have this issue looked at.

Values that are too high are much more common, and generally mean that something on either the client side (a bad splitter, for instance) or on the infrastructure side (something that overly attenuates the signal between the modem and the CMTS) is forcing the modem to work harder than it is meant to. This will eventually causes the modem to start its ranging over from 8 dBmV or to give up all together Somebody else will have to fill in the blanks on this one.

Since one CMTS has to provision many modems while each individual modem is free to adjust its own signal to please the CMTS, it is far easier for a modem to accommodate the needs of the CMTS than it is for the CMTS to accommodate the modems that it provisions. As a results, modems are more tolerant of suboptimal signal than is the CMTS.

Return path (upstream) SNR:

Who cares? The signal leaving your modem will have at most a trivial amount of noise.

Thanks for your time to explain to me.This stuff is over my old head.So from looking at my stats they seem fine ?
I probably will never fully understand this but just would like to know in simple terms what numbers would be cause for concern?Frequency,power and snr.ie what are good or acceptable ranges to be in?Sorry if Im' asking stupid questions but just interested.

sbrook
Mod
join:2001-12-14
Ottawa

sbrook

Mod

These are not the official ranges, but the ones where to be outside usually means problems

If you have a DOCSIS 2 modem (i.e. one channel up, one channel down) then your signals should look like

Downstream (aka Forward Path)

Power -10 to +10 dBmV Note that 0.0dBmV it's a reference level, so - means below reference and + means above.

SNR should be higher than 32 dB

Upstream (aka Return Path)

Power should be less than 52 dBmV

(No SNR normally available)

If you have a DOCSIS 3 modem (i.e. multiple channels up and down), use the same numbers, but the numbers for the various channels should ideally be within 1 to 2 dB of each other.

rodjames
Premium Member
join:2010-06-19

rodjames to DigitalRain

Premium Member

to DigitalRain
Did you really type all that?

oceros37
join:2013-07-20
St Thomas, ON

oceros37

Member

said by rodjames:

Did you really type all that?

There is this new feature that has taken the computing world by storm. It's called copy, and paste. It tastes good too.

rodjames
Premium Member
join:2010-06-19

rodjames

Premium Member

So, you expect us to read a bunch of stuff that you didn't actually write, and then we know stuff? You're weird.