Tech Support > Computer Hardware > Modems > Optimizing for the warm weather...
Optimizing for the warm weather...
Posted by Eric on March 30th, 2005


Hi,

Now that the weather is becoming warmer with each day, I have noticed that
the downstream power level to my cable modem is beginning to decrease. The
warm weather also is causing my cable modem to increase it's upstream power
level as well. The SNR is also slightly decreasing with warmer weather.
During the very cold nights last winter, I was seeing SNR at around 38 and.
With the warm days now (currenty 80 degrees), the SNR is between 35 and 36.
Downstream power has gone from -3.5 and -4 dBmV to between -7 and -8 dBmV.
Upstream power has gone from 42 dBmV to 45 dBmV. Everything still looks well
within specs, but I'm concerned that it may pass the thresholds once the 90+
degree temperature arrives.

The cable modem is located right where the cable comes into the house,
connected to a wireless router. There are three pieces of additional
hardware giving attenuation to the cable line once it comes in: a surge
protector, a splitter, and a male/male coupler. The splitter is very solid,
it's frequence range is rated up to 1Ghz, and it is spec'd at giving 3.5 dB
of attenuation. As a recent test (while 80 degrees outside), I ran the
cable line coming in directly to the modem to see what the downstream was
and it was a solid +1 dBmV. I then ran it through the splitter (plus a
three foot section of my own RG6) and got -2.5 dBmV. I then added the
coupler to the splitter, and found it to give an additional 2 dB of
attenuation. So, with the splitter, coupler, and a small section of my own
RG6, the downstream power level is at -4.5 dBmV -- attenuating the original
+1 dBmV by 5.5 dB.

I then ran it through the surge protector, which seems to introduce another
2.5 dB of attenuation. The attenuation of the surge protector, however,
doesn't seem constant. It seems to jump between 2 and 2.5 dB -- giving the
total output between -7 and -8 dBmV.

Unfortunetly, I never took at look at signal level coming in during the cold
weather. I suspect that it must have been hotter than +1 dBmV though, since
I was using the surge protector, splitter, and coupler at the time and
seeing between -3.5 and -4 dBmV.

This will be the first summer that I will have the cable modem. Should I
expect much more loss as the weather begins to climb from 80 to 90+ degrees?
As I have read, the range for the downstream power level should be
between -15 and +15 dBmV, but idealy you want to keep it within -10 and +10
dBmV -- with 0 dBmV being optimum. Should I look at removing one of the
three pieces of hardware before the summer months arrive? Perhaps locating
a good splitter that has a female connector for input and males for the two
outputs (to remove the coupler)? Is the surge protector absolutetly
necessary for coax -- or does the cable modem (plus the cable HDTV STB box)
provide their own surge protection for the coax inputs? All the electrical
inputs for the cable modem, router, HDTV, STB, Dolby Digital receiver, ect
are all going through the surge protector of course.

Or, perhaps, I really don't need to be too concerned about it. From
everything that I have, it all looks well "in the green".

Is this looking pretty good?

Currently, at 76 degrees outside, and cable uses QAM 256:

Downstream power level: -6 dBmV
SNR: 36 dB

Upstream power level: 45 dBmV
(This will fall back to around 42 dBmV by evening. Without the splitter,
coupler, surge protector, it only requires around 38 dBmV to talk to the
cable company.)

Thanks!
-ES



Posted by Rick Kunath on March 31st, 2005


Thec wrote:

The cable plant losses increase as the ambient temperature rises. Cable
plants work best cold. The active devices in the cable company's plant have
correction circuitry to compensate for this. ALC (automatic level control)
holds the output level of an active device constant and ASC (automatic
slope control) holds the pre-emphasis level constant. Cable plant active
devices pre-emphasize the higher frequencies of the baseband because of the
higher loss of the coaxial hardline as frequency increases. This "slope" as
it is called allows the downstream active device to see a flat (or nearly
so) frequency response on the baseband. Since the loss of hardline changes
with increasing temperature, the ALC and ASC circuitry automatically
increases the gain and slope as appropriate to maintain a constant level to
the next downstream active device. This is adjustable, and may be operating
incorrectly in your case.

This is pretty low, though legal. I'd expect to see +15 dBm in a properly
set up and designed system at your demarcation point.

Good so far.

Describe the coupler? What kind of device is this?

Dump the surge supressor. Get yourself a Regal gas-discharge ground block,
and place this at the demarcation point.

As to cable plant signal level variations, expect some change, but the
changes you are reporting would indicate that your plant is incorrectly set
up or has incorrectly operating ASC/ALC circuitry.

Call your cable company and have them send out a line tech to take a look at
their plant.

Rick Kunath

Posted by Ed Nielsen on April 1st, 2005


Those numbers are well within the operating range of a cable modem.
Cable modems are designed to wrok with an inout window of -15 - +15dBmV,
and an output of 8 - 55dBmV. SNR is excellent -- supposed to be <35dB.

(More below)


Rick Kunath wrote:

Not low -- perfect. Cable modems operate within a window of -15 -
+15dBmV. Much above or below that and sync goes away. With +15dBmV
(15dBm would be ~+63dBmV) at the groundblock, as the levels fluctuate
connectivity would be intermittent.

Most CATV systems in the US design their systems to have 15-20dBmV at
the tap, which would typically be ~10-17dBmV at the groundblock. 64
QAM carriers run at 10dB down fron the nearest analog carrier, which
means that at a groundblock that has +11dBmV as read by a regular Signal
level meter, the cable modem would read +1dBmV. 256 QAM is 6dB down
from analog.

Even with +15dBmV at the groundblock, a 2-way splitter for the cable
modem connection, and the outlet 50' from the splitter, the cable modem
would read 2.1dBmV (256 QAM) or -2.1dBmV (64 QAM).

CIAO!

Ed N.

Posted by Dave on April 1st, 2005


I think you meant to say that the SNR should be >35 (greater than 35).


"Ed Nielsen" <egnlsn@comcast.net> wrote in message
news:eN-dnTixSIfg-9DfRVn-hg@comcast.com...


Posted by Ed Nielsen on April 2nd, 2005


My bad! Thanks for catching that.


CIAO!

Ed N.


Dave wrote:

Posted by Rick Merrill on April 2nd, 2005


Eric wrote:
....There are three pieces of additional
What surge protector? Are you sure it's not just a ground tap?


Posted by Eric on April 7th, 2005



"Rick Kunath" wrote in message
Hi,

Thanks for the informative reply! The "coupler" that I was refering to is
just a simple female-female coupler that allowed me to connected a splitter
directly to the surge protector without using another cable.

http://www.x10.com/electronics/allca...n_200-100.html

(This isn't the exact one, but best picture I could find of one similiar.)

The demarcation point, I'm assuming is the outside box? Read a post on
dlsreports.com about taking a cable modem (and laptop) outside to hook it up
directly to see what it gets. (?) I have about 30 feet of coax cable
running from the outside box to the surge protector inside. Thanks for the
tip on the ground block. I'll look into doing this and not running
through the surge protector.

Cheers,
-Eric



Posted by Warren on April 7th, 2005


Eric wrote:
What surges are you protecting your equipment from by running the coax
through the surge protector? Lightning? A little surge protector isn't going
to help. Even if the surge causes it to disconnect, a lightning hit is going
to be able to jump that tiny, tiny little gap. And the chance of some other
kind of surge is very low. And even if you beat the odds, and have a power
surge from one of these rare causes, what would you loose if it made it to
the modem? A $50 modem?

There really isn't enough of a risk involved to bother with a surge
protector on the coax cable line. Compare that to the problems it causes by
adding extra exposed, unshielded connectors, and choking-off the signal. If
conditions are well enough that adding these adverse conditions doesn't make
a difference, that's one thing. But if it's having a negative impact, and
doesn't have a positive impact to off-set that negative, the choice is easy:
Stop running the coax through the surge protector.

--
Warren H.

==========
Disclaimer: My views reflect those of myself, and not my
employer, my friends, nor (as she often tells me) my wife.
Any resemblance to the views of anybody living or dead is
coincidental. No animals were hurt in the writing of this
response -- unless you count my dog who desperately wants
to go outside now.
Care for your landscape with Black and Decker cordless tools
http://www.holzemville.com/mall/blac...ker/index.html




Posted by Eric on April 14th, 2005



"Warren" wrote in message
Its nothing special. Just a simple APC (PF11VT3) surge protector.

Picture:
http://www.amazon.com/gp/product/ima...&s=electronics

Thanks. I'm not going to worry about going through it. Only thing that
could be lost would be the cable modem (Motorola Surfboard) and cable TV
(Pace HD) STB -- both of which are owned by the cable company anway. :^)

Probably would even find it humorous to see the Pace box destroyed anyway,
but thats an whole new subject!

Cheers,
-Eric






Similar Posts