dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
1308

pflog
Bueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA

pflog

MVM

[Signals] Should SNR swing as much as 1dB with temperature?

Click for full size
I know dBmV are prone to swings due to temperature, but my SNR has (in the past) been fairly stable throughout the day.

I'm seeing a pattern now where the SNR is going up with colder temps and dropping when things warm up. This has me pretty concerned for the summer when the temps are 100+.

Does this indicate a wiring problem of some kind? The last tech that came out (to troubleshoot the erratic speeds after powerboost, which I knew was not a signal issue at all) did replace my connection at the POE and prior to his "work" the signal was actually better (higher SNR). I'm wondering if he broke something when he was doing that.

cielbleu
@videotron.ca

cielbleu

Anon

1 dB "swing" is nothing, you're funny! Don't worry, he broke nothing. Call back when you will get 3-4 dB SNR sudden drop.
Dumaine
join:2010-12-05
Detroit, MI

Dumaine to pflog

Member

to pflog
lol @ 1dB...........When it rains I have had anywhere from 6 to 9 dB drop (Downstream Power) here in Detroit since my install date back in November 2010. DS and US lights start flashing on the Ubee. I have called on it a few times only to hear that its normal. Till this day I have yet to lose connection though, even when the lights are flashing on the Ubee. They will flash until i reset (have let them flash for 48 hours once). Im willing to bet they have a crack somewhere up on the pole yet getting lazy line techs here to check and see is impossible.

pflog
Bueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA

pflog to cielbleu

MVM

to cielbleu
said by cielbleu :

1 dB "swing" is nothing, you're funny! Don't worry, he broke nothing. Call back when you will get 3-4 dB SNR sudden drop.

The temperature variance right now is 15 degress. When it's swinging from 55 to 100, I'm almost certain it'll be there or worse.

What concerns me is that the SNR varied hardly at all, regardless of temperatures, until this tech came out...

Johkal
Cool Cat
MVM
join:2002-11-13
Pennsyltucky

Johkal

MVM

said by pflog:

said by cielbleu :

1 dB "swing" is nothing, you're funny! Don't worry, he broke nothing. Call back when you will get 3-4 dB SNR sudden drop.

What concerns me is that the SNR varied hardly at all, regardless of temperatures, until this tech came out...

I see your point. Whereas 1 dB can be a normal swing, I would be very curious why the 1 dB swing after the Tech left. Still could be coincidence, but I would keep an eye on it.

Streetlight
join:2005-11-07
Colorado Springs, CO

Streetlight to pflog

Member

to pflog
I've never noticed a change in download SN ratio with temperature. And we have large temperature changes here in Colorado. It can be 0 deg F in the early AM and get to 40, 50 deg F or more by 3 PM. What really changes w/ temp that can affect internet is upload signal. In my experience, if it gets to 50 dB Mv and maybe higher, internet can be lost. Downstream power also changes w/ temp, but doesn't seem to be as critical as upstream signal level. YMMV.

pflog
Bueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA

pflog to Johkal

MVM

to Johkal
said by Johkal:

I see your point. Whereas 1 dB can be a normal swing, I would be very curious why the 1 dB swing after the Tech left. Still could be coincidence, but I would keep an eye on it.

Yeah, guess I'll just keep an eye on it. Come to think of it, the lower SNR overall was after the tech left. I think the fluctuations actually started early last week after an outage overnight (coincidentally after which I lost my 3rd upstream channel, but got it back after rebooting the modem).
pflog

pflog

MVM

Click for full size
Yeah, it was on the night of the 23rd (see attached).

cielbleu
@videotron.ca

cielbleu

Anon

There isn't any tech who came here, but when i look at my signal, my snr always fluctuate of about 1 dB. It's normal...

You're right, in summer with high temperature, signal loss in a trunk or distribution coaxial cable is higher and lower in winter with colder temperature. If your cable plant is build with many cascaded distribution amplifiers with long run of cable between them, the lower signal at the input of each amplifier will cause the SNR to drop of several dB (4 to 10 dB). Same thing in cold temperature, too strong signal at the input of each amplifiers means overdriving amplifiers causing distorsion and SNR drop. It's why there's AGC (Automatic Gain Control) in amplifiers, but sometimes agc can be broken or an amplifier is let in manual mode instead of agc for some reason.

It's why maintenance technicians exist, to balance cable lines so the signal stay optimal all year long, it is called maintenance.

Don't worry for a 1 db SNR variation!!! There is no issue there.

pflog
Bueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA

pflog

MVM

Right I'm not so much worried about a 1dB swing as I am thus occurring g after an outrage and seeming to correlate with temperatures, since this area had a large temperature swing in the summer, sometimes as much as 50 degrees. Perhaps the AGC got screwed up when they did whatever maintenance they were working in. I'll just monitor it and see if any sync loss or other issues arrise, thanks.
harald
join:2010-10-22
Columbus, OH

harald to pflog

Member

to pflog
No cable modem is capable of reading SNR.

What they display as SNR is a calculated value based upon the ratio of packets with correctable errors to good packets, using the formulas derived by Claude Shannon in the 1040's. The details of how this is done is proprietary to Broadcom and other chip makers. Inferences about what will cause the SNR to change are pretty shaky.

The same is true for upstream SNR. The cards in the CMTS also use similar chips.

So it is best to think of the displayed SNR as just an indication of the error rate in terms that techs are used to understanding. The estimate is also impacted by modulation errors.

NetFixer
From My Cold Dead Hands
Premium Member
join:2004-06-24
The Boro
Netgear CM500
Pace 5268AC
TRENDnet TEW-829DRU

NetFixer

Premium Member

said by harald:

No cable modem is capable of reading SNR.

What they display as SNR is a calculated value based upon the ratio of packets with correctable errors to good packets, using the formulas derived by Claude Shannon in the 1040's. The details of how this is done is proprietary to Broadcom and other chip makers. Inferences about what will cause the SNR to change are pretty shaky...

I have seen references to that as well, but that does not explain why Channel ID 6 in my SB6121 does not have a much higher DS SNR than the other three channels, instead of all of them reading 38dB.


harald
join:2010-10-22
Columbus, OH

harald

Member

What is missing is an understanding of what time frame is used for the calculation. Here is a white paper from Broadcom that explains the issues, but does not even hint at what the averaging parameters are:

»www.broadcom.com/collate ··· 01-R.pdf

beachintech
There's sand in my tool bag
Premium Member
join:2008-01-06

beachintech to pflog

Premium Member

to pflog
Your basing this off the measurement of a modem, which isn't a real reading, but rather a mathematical formula based off some other stats that it's estimating. You have no issues to worry about.
beachintech

beachintech to pflog

Premium Member

to pflog
said by pflog:

Right I'm not so much worried about a 1dB swing as I am thus occurring g after an outrage and seeming to correlate with temperatures, since this area had a large temperature swing in the summer, sometimes as much as 50 degrees. Perhaps the AGC got screwed up when they did whatever maintenance they were working in. I'll just monitor it and see if any sync loss or other issues arrise, thanks.

If that were true you would most like see a corresponding change in signal levels forward and return rather than purely a "change" in SNR calculation.
Jeremy W
join:2010-01-21

Jeremy W to pflog

Member

to pflog
Your life would be much less stressful if you didn't monitor your cable modem's stats so obsessively. As has been pointed out, SNR isn't even a real stat anyway.

pflog
Bueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA

pflog

MVM

said by Jeremy W:

Your life would be much less stressful if you didn't monitor your cable modem's stats so obsessively. As has been pointed out, SNR isn't even a real stat anyway.

I'm not obsessing, but I did notice a change after the outage, hence my question. It's been answered.
Jeremy W
join:2010-01-21

Jeremy W

Member

said by pflog:

said by Jeremy W:

Your life would be much less stressful if you didn't monitor your cable modem's stats so obsessively. As has been pointed out, SNR isn't even a real stat anyway.

I'm not obsessing, but I did notice a change after the outage, hence my question. It's been answered.

Listen, I'm a nerd too. I track this stuff just like you do, but I rarely look at it. I apologize if I came off as confrontational. I was just making a light-hearted suggestion.

pflog
Bueller? Bueller?
MVM
join:2001-09-01
El Dorado Hills, CA

pflog

MVM

said by Jeremy W:

said by pflog:

said by Jeremy W:

Your life would be much less stressful if you didn't monitor your cable modem's stats so obsessively. As has been pointed out, SNR isn't even a real stat anyway.

I'm not obsessing, but I did notice a change after the outage, hence my question. It's been answered.

Listen, I'm a nerd too. I track this stuff just like you do, but I rarely look at it. I apologize if I came off as confrontational. I was just making a light-hearted suggestion.

Understood, and I just wanted to clarify that I wasn't obsessing, but I do debug things and write code for a living, so I tend to notice when things change. I had never seen my SNR swing correlating to temps, so wasn't sure if I had loose fitting or something. Guess I'll see how things are in August