dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
676

train_wreck
slow this bird down
join:2013-10-04
Antioch, TN

train_wreck

Member

[Signals] why are sub-20MHz channels "noisy"?

have seen this mentioned, why are they "noisy" ('im guessing meaning worse signal-to-noise ratios)?

DarkLogix
Texan and Proud
Premium Member
join:2008-10-23
Baytown, TX

DarkLogix

Premium Member

Likely because that range wasn't important before cable internet. so back then they didn't worry about issues that would effect it so those issues built up more than issues that effect the higher ranges.

Tobin
join:2003-09-21
Burlingame, CA

2 edits

2 recommendations

Tobin to train_wreck

Member

to train_wreck

Re: [Signals] why are sub-20MHz channels "noisy"?

Notice that porous mesh in front of your microwave oven? It keeps you safe from 2.45GHz microwaves. It is considerably less effective against a 700MHz radio signal or a 20MHz one. The same applies to the braid and foil on your RG6 cable.

Remember when everyone said T-Mobile sucked in urban areas with tall office buildings or hilly areas? It's because they were using the 1900MHz band while carriers like AT&T used 850MHz.

Ever wonder why 5GHz 802.11a/n/ac has lower range than 2.4GHz? Because it's that much harder for that signal to penetrate rebar, wood, drywall, even the atmosphere itself.

Speaking of our atmosphere, it shields us from most of that higher frequency radiation that would otherwise be deadly. It also means there's more lower frequency signals both artificial and naturally occurring, terrestrial or cosmic that worsen SNR.

TL;DR, shielding is generally less effective vs longer wavelengths.

e: I am oversimplifying things a tad. The mesh in your microwave window would also leak 60GHz signals like a sieve.
Tobin

2 edits

Tobin to train_wreck

Member

to train_wreck
DarkLogix, the reason cable internet even uses those frequencies is because every traditional cable system supports it. There occasionally have been proposals to move return-path signals to above downstream frequencies, but that would trade noise impact for a much steeper power requirement. That would also require wholesale upgrades to infrastructure.

I'd be less worried about 20MHz noise than 700-900MHz noise anyway. LTE networks are growing fast and guess what kind of signals they are blasting into the open? The need for properly terminated and shielded cable hardware is paramount now that we have an insatiable appetite for wireless bandwidth. A loose connector with a wonky ground path equals ingress. A cable connected to a splitter on one end and not connected to anything on the other end is an antenna...
Matt7
join:2001-01-02
Columbus, OH

Matt7

Member

said by Tobin:

I'd be less worried about 20MHz noise than 700-900MHz noise anyway. LTE networks are growing growing fast and guess what kind of signals they are blasting into the open? The need for properly terminated and shielded cable hardware is paramount now that we have an insatiable appetite for wireless bandwidth.

Totally agree.. The cable provider I am customer of (TWC) uses 705 Mhz, 711 Mhz for two of the DOCSIS 3.0 downstream channels. I see a lot of interference (correctables, uncorrectables) on those two slices compared to the other slices near by.

andyross
MVM
join:2003-05-04
Aurora, IL

andyross to Tobin

MVM

to Tobin
said by Tobin:

I'd be less worried about 20MHz noise than 700-900MHz noise anyway. LTE networks are growing fast and guess what kind of signals they are blasting into the open?

Comcast and other cables companies have already been moving stuff around to avoid the 700MHz band. By me, they moved a bunch of stuff over the winter. Alot was moved into the 100-200MHz area opened up with analog reclamation.

Tobin
join:2003-09-21
Burlingame, CA

Tobin

Member

Andyross, that's true, but a temporary solution when consumer bandwidth expectations continue to grow. Properly shielded equipment/cable and terminated ports/connectors provide more than enough shielding. It's just an easy mistake to loosely hand twist an f-connector onto a port or leave one open altogether.

gar187er
I DID this for a living
join:2006-06-24
Seattle, WA

1 recommendation

gar187er to train_wreck

Member

to train_wreck
there is just a lot of garbage down that low. Aeronautical, marine, HAM, etc....its best to stay out of it, cause the only thing that will work that low would be QPSK, MAYBE a 16qam.

As was stated, certain MSOs are exploring expanding the return band all the way up to 150-200mhz. But you need to go change every diplex filter in every amp in the system to do this. and some modules you cant, so you need to swap the whole module which gets expensive, REALLY fast.

telcodad
MVM
join:2011-09-16
Lincroft, NJ

1 edit

telcodad to train_wreck

MVM

to train_wreck
said by gar187er:

there is just a lot of garbage down that low. Aeronautical, marine, HAM, etc....

Yes, also most random RF noise, caused either by natural sources (e.g., lightning) or man-made ones (e.g., electric motors, switching in electrical circuits, etc.), has a power spectral density that's inversely proportional to frequency (i.e., varies by 1/f), so it affects lower frequencies more than higher ones.

That's why you typically hear more static/crackling on AM broadcast radio stations that are at the lower end of the band, and used to see more picture noise/static in the old, analog, lower-VHF-band TV channels years ago.

Also check out:

Why Is Cable's Upstream Path So Darned Ugly?
By Leslie Ellis, Translation-Please.com - April 2, 2001
»www.translation-please.c ··· ed-ugly/

train_wreck
slow this bird down
join:2013-10-04
Antioch, TN
Cisco ASA 5506
Cisco DPC3939

train_wreck to Tobin

Member

to Tobin
said by Tobin:

I'd be less worried about 20MHz noise than 700-900MHz noise anyway. LTE networks are growing fast

yes, i think i may have noticed an effect of this in one of my other threads

»[Signals] ever had the SNR of 1 channel much different than the rest?

thanks for the info folks!

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 train_wreck:

yes, i think i may have noticed an effect of this in one of my other threads

I see a larger number of error codewords on my 705 & 711 mhz channels (AT&T LTE), but not enough to cause problems (this is a 44 day snapshot):



train_wreck
slow this bird down
join:2013-10-04
Antioch, TN
Cisco ASA 5506
Cisco DPC3939

train_wreck

Member

Click for full size
word, don't really see that here.
(uptime is ~2 weeks)

gar187er
I DID this for a living
join:2006-06-24
Seattle, WA

gar187er to NetFixer

Member

to NetFixer
thats not a large amount. did you divide to see the percentage?

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 gar187er:

thats not a large amount. did you divide to see the percentage?

As I said, I don't consider it a problem -- I was just noting that there did appear to be a measurable difference for the two frequencies I am currently using that are shared by AT&T LTE.