dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
983
share rss forum feed


jbob
Reach Out and Touch Someone
Premium
join:2004-04-26
Little Rock, AR

2 edits

[Signals] Help me interpret my signal levels

Running a Cisco DPC3000 modem. Don't think I really have any connection issues so to speak but am unsure about the numbers I am seeing on the modem diagnostic page.

For a quick rundown I had some connection issues several months back. In checking I found my the RG-6 running from my 3 way outside to my modem was bad so replaced it. After I installed the new line my signal levels at the modem were all showing around 0 dBmv. Great. About a month later there was an area outage. When it came back on my signal levels had jumped to the 6 to 8 range.

So fast forward.....every once in a while there appears to be an outage so the first thing I always do is check my signal levels. Right now they're running around 8 to 10 dBmv.

I just bypassed my house wiring again and connected the modem directly to the drop prior to the 3-way into my home. The 4 downstream power levels read:
15.6, 14,7, 14.5 and 13.9.

So my question is what should the power levels be at the drop? 10.0 to -10.0? The higher the number in the negative range the better correct? Within reason of course! Loss of signal levels should drive the number more to the positive from the negative...correct?

If so doesn't mine seem backwards?

And in case you're wondering here are the signal levels(SNR) at the drop....
36.2, 35.8, 36.0 and 36.0. The upstream power level is 36.3.


andyross
Premium,MVM
join:2003-05-04
Schaumburg, IL

For downstream, higher numbers are stronger. You ideally want between +10 and -10.

I think that upstream number is not SNR, but transmit strength. It's maybe a bit low, but that's because your downstream is so strong.



jbob
Reach Out and Touch Someone
Premium
join:2004-04-26
Little Rock, AR

You're correct. The Upstream was the power level not SNR.

So when you say the higher number is stronger do you mean higher number in the positive range? I always thought the higher negative number was the higher power level. In other words a -15.0 dBmv is more power compared to 15.0 dBmv. So as the impedance increases the power levels go more toward the positive.

But if you're right then I need to change the RG-6 running from the 3-way to my cable modem from a 3.5 tap to a 7.0 tap.


rody_44
Premium
join:2004-02-20
Quakertown, PA
reply to jbob

Plus 15 is stronger than minus 15



pclover

join:2008-08-02
Santa Cruz, CA
reply to jbob

+8 to +10 is a bit high.

I would get a 6 dB Attenuaor.



Anonymous
Premium
join:2004-06-01
IA
kudos:2
reply to jbob

Most cable cos have different recommended ranges but downstream at 0 is always within that range. I like to see it at or around 0. Your upstream also depends on how your cable co set up their plant. Upstream at 36 is kinda low for most cable cos. So putting the splitter back and connecting the modem to a 7 dB port is probably a good idea.

Keep in mind as outside temp goes down your downstream will go up and as it goes up it will go down. Your upstream tends to go up when temps are going up and down when temps are going down.

PS If you don't have any issues don't mess with it.
--
I speak for myself, not my employer.



jbob
Reach Out and Touch Someone
Premium
join:2004-04-26
Little Rock, AR
reply to jbob

Thanks guys. I just swapped the RG-6 to a higher db tap on the splitter. Power levels in the 3 to 6 range now. Looks like where I live they're making a bit more power than I've grown accustomed to. It's easier to attenuate rather than having to make power!


Extide

join:2000-06-11
84129

Larger numbers are higher power and smaller numbers are lower power.

<--Lower           Higher -->
-10.....-5.....0.....+5.....+10

-10dBmv to +10dBmv is 'in spec'

This is a linear scale.

SNR is measured in dB and is a logarithmic scale. I believe you want >=35dB.



Streetlight

join:2005-11-07
Colorado Springs, CO

1 edit

PhReE5,

Both scales are log scales.

The signal level in dBmV = 10 x log(output voltage in mV/input voltage in mV)

(mV is milli Volt)

The original basis of this scale is that if the output and input voltages are both 1 mV, then the output:input ratio is 1, the log of 1 is 0 and the signal level your modem would read would be 0 dBmV. This 1 mV standard apparently comes from the observation that for early generation TV sets, an antenna signal of 1 mV was optimum for best picture quality. For modern cable systems, I'm not sure 1 mV is optimum for current cable boxes or internet modems. I get a pretty good picture on a modern TV with rabbit ears antennas, so maybe modern cable boxes/modems are designed to work best with a 1 mV signal.

If your cable company sends a signal of 1 mV and by the time it gets to your modem it's amplified to 2 mV, the ratio of the output to input signals is 2, the log of 2 is 0.3 and 10 x 0.3 is 3 dBmV. Thus a signal change of 3 dBmV corresponds to an increase in signal of 2. A 6 dBmV change is a factor of 4 in amplification. Cable signals can be negative relative to the 1 mV standard if the signal is reduced at the modem.

I'm not sure, but I would guess that if the signal, in dBmV, is greater than 15 dBmV, the modem's electronics are overloaded and won't function correctly. A 15 dBmV signal corresponds to an amplification of 32, a big value. Likewise, if the signal is -15 dBmV, this corresponds to an attenuation factor of 32 and there may not be enough signal for the modem to work.

Check out this link:

»www.therealcableguy.com/tv-signal-levels.htm

These voltages may actually represent a function of the RF power transmitted and received. Coaxial cables do have voltage on them, but is is extremely small whereas the cables actually act as wave guides, basically tubes that function to carry the electromagnetic RF waves down them. The size of the tube must match the wavelength of the waves as well as their amplitudes.

Things can get pretty complicated, but I think the description of decibels is correct.

--
There is nothing more deceptive than an obvious fact.

Sherlock Holmes in
The Boscombe Valley Mystery
A. C. Doyle
Strand Magazine, October 1891



EG
The wings of love
Premium
join:2006-11-18
Union, NJ
kudos:9

SNR is NOT a "range" at all. There is a bottom figure as a *minimum* which at the very least should be met, if not exceeded !!



Streetlight

join:2005-11-07
Colorado Springs, CO

EG,

My discussion was about the downstream/upstream power levels since PhReE5 asserted that they were a linear function. They are log functions.

SNR is also given as a log function, dB (deci Bell) but I'm not exactly sure how it's measured. In my work in RF spectroscopy, the noise is measured using its root mean square (RMS) value, but I'm not sure its done that way with modem signals. I'll look around and see if I can find anything to report on that.

I don't think I said SNR has a range. It's calculation can involve statistics in some cases. You're right, it's a particular positive value. I guess it could "range" from unmeasurable - the signal is equal to or less than the noise - to some positive value. For a small SNR using statistics one can give a confidence that the signal was observed. Take a look at the data analysis used to assert that the Higgs Boson was actually detected. I watched the live presentations of the two experiments and it was an amazing tour de force of statistical analysis all designed to assert that a signal was observed at a particular energy. Fascinating.
--
There is nothing more deceptive than an obvious fact.

Sherlock Holmes in
The Boscombe Valley Mystery
A. C. Doyle
Strand Magazine, October 1891


Extide

join:2000-06-11
84129
reply to Streetlight

said by Streetlight:

PhReE5,

Both scales are log scales.

The signal level in dBmV = 10 x log(output voltage in mV/input voltage in mV)

That is iteresting, I didnt know that. I always wondered why it was dBmV and not just mV!


gar187er
I do this for a living

join:2006-06-24
Dover, DE
kudos:4
reply to Streetlight

said by Streetlight:

PhReE5,

Things can get pretty complicated, but I think the description of decibels is correct.

nothing complicated that +10 is greater then -10

positive and negative numbers....thats what the OP was confused on.
--
I'm better than you!