Problem with cable modem - Page 3

Ignitionnet · 25-04-2010, 09:29

Modem lost downstream signal, as you would expect given the extremely low power level.

---------- Post added at 10:29 ---------- Previous post was at 10:18 ----------

Quote:

Originally Posted by Sephiroth

Sounds like a load of rubbish to me. A "bigger cable"? They would need to ensure that the impedance between your wall box (or indeed your cable modem) and the street box is 75 ohms. So that might require a re-pull of the cable to your wall box thriugh the green pipe you mentioned. They'll use the old cable to pull through the new.

HTH.

Thicker cable = less attenuation / signal loss so higher power level received on the other end of the cable. This isn't related to impedance at all, all cable VM use is of the correct impedance. Where cable were of an incorrect impedance it'd indicate a fault with the cable which would ideally need replacing. The reported power level is voltage relative to 1mV RMS and impedance doesn't enter into the calculation.

Problem seems to be nothing more interesting than a damaged cable or loose connector, maybe on the outside of probabilities a totally buggered amplifier, hence the >20dBmV variations in power levels.

Sephiroth · 25-04-2010, 10:08

Of course it's related to impedance. Of course the cable is 75 ohm. But in the wider picture, there are connectors between modem and street box and the end to end impedance has to remain 75 ohms. A damaged cable, oxidisation, a loose connector all affect impedance.

The relationship between dBmv and power (watts) is impedance dependant. Relative values to 1 mV RMS are meaningful only in the context of pwer delivered (energy per symbol or bit or whatever). That's why 75 ohms has to be assured.

Ignitionnet · 25-04-2010, 10:37

By the time a cable is degraded enough for impedance mismatch to cause noticeable issues with regards to power levels reflected signals are going to cause serious issues with signal quality. SNR would be seriously degraded way before power were taken out of spec.

My comment regarding impedance calculation is perfectly accurate, it's assumed to be 75ohm in the calculation due to it being such a non-issue.

As far as I know impedance isn't even measured when installs take place and it's certainly not a reason for repulling with a higher grade coax, that is simply to reduce attenuation across the drop by using a cable with a thicker core.

Sephiroth · 25-04-2010, 11:03

Look - I'm not the one suggesting a re-pull. I'm merely putting the electrical considerations that can affect the signal. One customer, after a repull, still found water in their tap box, corroded metals and although clearly 75 ohms wasn't there any more, you're right - impedance wasn't checked.

And you know what impedance mismatch does to signal.

As for "assuming that it's a potential issue" - it is a potential issue although I'm not assuming that it is the issue.

Ignitionnet · 25-04-2010, 11:07

I reckon we just test the drops by getting the installed to pop his tongue onto the cable. Once he's done a few he should get some idea what feels right and what doesn't.

Sound good?