Re: Cable Broadband Internet Signal loss from serial RG59 cabling - RF amp recommendations?

drydem <walter_lee@xxxxxxxxxxx> writes:

On Jan 22, 12:17 pm, comph...@xxxxxxxxx (Todd H.) wrote:
drydem <walter_...@xxxxxxxxxxx> writes:
Thanks for the tip.

I went to the local Radio Shack store and I purchased
a Bi-Directional Cable TV 10db Amplifier (Radio Shack Catalog
Number 15-2505) which is designed to allow for digital cable
modems, digital TV, and pay for view services.  I replaced
the older RF 20db Amplifer with the newer Bi Directional
RF 10db Amplifier - while the cable channel signals were still
usable/powerful --  the cable modem was still unable to
get an internet broadband signal in the office.  Why didn't
it work? I'm not sure.

There are a few possibilities.  You're (admittedly) using the wrong
cable, you're crimping your own connections, using a Radio Shack
amplifier and a Home Depot 3-way splitter.  It's amazing you get TV
much less cable modem signal.  :-)

May I suggest a more methodical approach?

Let's start by plugging the cable modem in where it last worked.  Look
at the web page athttp://192.168.100.1where most DOCSIS modems will
have a signal page somewhere... and record upstream power level,
downstream power level, and signal/noise ratio (SNR).  Post those
results.  Then we'll know how much margin you have to play with.  For
example, right now I'm looking at -8.5dBmV downstream (a bit whimpy
but ok), 49.25dBmV upstream (pretty high... this is how much the cable
modem has to crank up the gain to get a signal to the head end), and
34dB SNR (solid).  This trio is on the edge of reliability, in my
experience.  Ideally I'd like downstream a little higher, upstream a
little lower, and SNR the same or, heck, higher is always good for
SNR.

That's an excellent idea. Thanks.
Since this network cable is a serial - I can measure
the signal from the external wall source to each connection
point to determine if there is any radical loss of power.

using the Webstar dPC2100 cable modem as a
network diagonistic tool is an excellent idea
thanks. The Webstar dpc2100 cable modem status
on ip address 192.168.100.001 has a receive
power and transmit power level. Because my old
old 20db RF amplifier is a unidirectional I
remove it for the following reading. Reading
from the source signal down the line I found
that by the time I get to third floor in the
master bedroom I have a dramatic loss in my
ablity to transmit a signal. I suppose I could
redo this wall socket and see if that helps...

here are my readings:

1) Original Comcast Cable Signal(RG6 cable via compressed Fconnector)
Recieve Power -05.2 dBmV
Transmit Power +42.0 dBmV

2) Signal passes through Ideal 2Ghz 3-way digital/Satellite Splitter
Recieve Power -02.2 dBmV
Transmit Power 47.0 dBMv

This reading doesn't make sense. It's 3dB higher than the original.
Which leaves us wondering how we got 3dB of gain by going through a
3-way splitter. Something is wonky.

3) Signal passes through Ideal 1Ghz 2-way splitter in Basement
Receive Power -07.3 dBmV
Transmit Power 52.0 dBMV

Yer on the edge of the cable modem working without packet loss at this
point.

4) Signal passes through Ideal 1Ghz 2-way splitter to Living Room
Cable TV outlet
Recieve Power -13.5 dBmV
Transmit Power 55.5 dBmV

And transmit is probably maxed out at this point. Cable modem
probably wont' sync here.

5) Signal passes through Ideal 1Ghz 2-way splitter to Master Bedroom
Cable TV Outlet
Recieve Power -17.7 dBmV
Transmit Power 008.3 dBmv --- ouch!

Transmit opwer obviously something strange here. but with 55 being
the max most modems will try to push, weird readings here don't pahse
me On the receive end, which is all you care about because it's a TV
line for ya, you're seeing the same 6dB drop you've seen after prior
2-way splitters. 3.5dB is the nominal insertion loss of a 2-way
splitter, but the cable modem may be testing at a relatively high
ferquency that sees more loss than nominal. Cable losses also
contribute and are frequency dependent.


The last reading is really pathetic - give to drastic drop in
signal the office or guest room connection must be getting zilch.

Nah, it's not really that bad. Does the TV work there? Of course the
modem work as it's out of the upstream gain range.

I was experimenting with those fancy twist on RG59
f-connectors in the master bedroom... it looks like I am
going to have to redo that wall socket as a crimp f-connector
instead...

The twist on's yeah... I seemt o recall them being a trouble spot.


--
Todd H.
http://www.toddh.net/
.

Relevant Pages
Loading