Participant

To Much Signal?

Spectrum came to the house and said I am boosting to much signal.  What are the effects of having to much signal?  So, what happened was my internet went out and I was wondering why.  I checked everything and found out that my amplifier in my wall shut down because the ac adapter going into the powerstrip was broke.  One of the prongs was pushed in so I had to pull it back out.  Once I did that it all worked fine again.  Anyway, I scheduled an appointment just wanting them to give me a new ac adapter for my amplifier.  The Spectrum employee comes to the house and I tell him what the problem is by phone because Im at work, but after a few tests he says I have to much signal and says he needs to take the amplifier out because I don't need it.  Now, my setup has been working good for 2 years, internet hitting the right speeds 300/25, cable working good,  all the above.  Im thinking if its not broke why fix it.    So the Techs soulution was to take my amplifier out and run a splitter, one to the modem and the other to the cable.......duh..... I wish I was home at the time but I wasen't.  I was talking with the tech on the phone while my Dad was watching him do his install.......smh!!!!  So what would be the effects of too much signal be?  BAD OR GOOD i GUESS?

 

When I get home I will update everyone.

 

Thanks! 

17 REPLIES
Expert

Re: To Much Signal?

Amplifiers are bad and should never be needed for a 2 way split.

 Please post the modems signal level page

 

Valued Participant

Re: To Much Signal?

If you have too much power, you’ll suffer timeouts/disconnects/modem won’t lock signal.

 

I’ve read here amplifiers aren’t a good thing, as MsRaye confirmed. If you have too much power coming in,  a splitter can help balance your signals. I had an issue where downstream  power coming into our condo was too hot. Bucket truck made change at box outside and tech added a three-way splitter to balance signals.

 

Post your new signals with splitter, so MsRaye can diagnose them for you.

Participant

Re: To Much Signal?

When I get home I will upload signals!

 

Thanks!

Expert

Re: To Much Signal?

DS must be in the -8 to +8 dBmv window with adjacent channels being not off by more than .5 to 1.0 dB

US needs to be in the +30 to +50 dBmv range Serious degradation occurs if +53 dBmv is reached.

if any channels level changes more than 3.0 dB, modem will reset and try to relevel

 Qam 256 DS S/N must be better than 33 dB correcteds/ uncorrected increase exponentially as s/n hits 29-30 dB and won't work reliably below that.

 Poor s/n on a couple modems on a node will degrade everyone on the node thanks to all the retries from corrected... Uncorrectors will affect speeds on the servers as data has to be resent all the way from the host and that will take an additional 100+ ms each time as well as interupt the stream of data

 

Proven Sharer

Re: To Much Signal?

Note that all 8 of the DS channel levels are running hot, above +10 dBmV, which is not going to improve user internet speeds.   That's why customers should NEVER install amplifiers to try to boost their internet service.

Highlighted
Spectrum Employee

Re: To Much Signal?

Actually... to fine-tune & correct the statements here by @MsRaye@scjr , and @karlbeckman , amplifiers can be good or bad, it just depends .

 

For this topic, I'd need specific information I could only really get from being on-site.

Any maybe the expected values measured at the tap would differ from where this customer is located versus where I am located, I'm not sure.  But I don't think it'll vary by much because the CPE (modems & cable boxes) is the same used nationwide.

 

I'll try to get as specific as I reasonably can, for the benefit of any enthusiasts patient & eager enough to learn a bit more.  Apologies in advance if this is over the heads of some readers, or overly wordy for others.  I'm striving to write with accuracy.

so here we go...

 

needing a "house amp" primarily depends on:

The RF signal power needs (such as the balance of forward/Rx/receive/downstream versus return/Tx/transmit/upstream),  distance attenuation, splitter configuration, and the manufactured quality of the amp in question.

(read: does it cause ingress or egress leakage?  How much dBmV power can it take before it leaks?  Is it energy efficient?  Is it stable output?  Does it tilt the levels?  What is the effective frequency range?  etc etc etc).

 

Spoiler

 The only amps we tend to use (and often have to use if >3 outlets, are the more expensive/good-quality "unity gain" (aka. no loss, no gain) amps that neither boost nor lose signal power on the powered ports for EITHER forward or return.  (0 dB Rx, 0 dB Tx). These amps we use need to be able to also have shielding enough to contain a +15 dBmV signal with zero egress leakage when terminated properly, and 20 to 25 dBmV signal with only 30 μV/m egress leakage.  Also needs to be ok on MER & Pre-BER for -10 to +15 dBmV.  So it's some pretty specific requirements you're not going to find on some cheap-o retail store knockoff.  Just because it's good enough for an antenna signal doesn't mean it's good enough for 256 QAM or higher. 

I've also seen customers with poor quality amps causing issues for neighbors on their street, and we'll absolutely bill for damages to rental equipment (modems & cable boxes) damaged in such a way.  So it's one of those things that if you don't know, better of to just let us handle it. 

Even worse case scenario, for a $50 flat OTC for an install, you end up getting an $80 or $120 amp for no additional cost with the labor... so yeah kind of win-win if you just let Charter (Spectrum) take care of it.  And no, we should avoid re-using amps or the power supplies for amps.  They are manufactured as a pair on purpose.

So be careful!  If you're unsure, ask for tech help.  Would be happy to help on cable wiring issues.

 To answer this topic starter's primary question:

The technical answer with regards to @berryboy808's question about over-amping the Forward signal power is that it degrades the MER (modulation) and BER (post equalization bit error ratio).  

That's technically correct, but will require more explanation...

Depends on the quality of the modem, but most modems don't take kindly to Rx/Forward/Receive/Downstream signal power levels >15 dBmV.   What will typically happen when it's more out of range than the modem can handle is that the modem will try to self-reboot to establish a connection on some different DOCSIS frequencies, basically ask the CMTS for a new slot to try again.

An analogy would be to think of it as audio.  If you over-amp the volume level, it'll distort and square-off the peaks of the waves (not to mention giving hearing damage).  And yes while it's true that you can decrease the lifespan of the receiver/demodulator, the main issue is that signal quality degrades if it's too far out of range.

Spoiler
I once saw a customer wear a cable TV box tuner out with a RadioShack amp that was boosting (and also garbling up horribly the signal quality) to 58 dBmV (and that's Forward/Rx, not return).  My meter almost didn't read it, and it gave me a voltage AGC warning as well.  Thought it was a power pin out at the tap at first before I found the source.  (it was a cheap-quality forward-gain amp underneath the trailer home tucked inside a fold of the tarp)

 

It could also depend on market area, but in most of the Midwest market area (where I work (basically Kentucky, Indiana, Ohio, Wisconsin)), the cable plant outside is designed to be split for 2 to 3 outlets per tap. 

Single outlet direct off tap and you're likely to be a bit high on Rx/Forward especially if you're the first tap after the "active" (street amp or line extender), but that could be ok.

Four outlets or more and you may need an amp (if you're not at the end-of-the-line (last tap on the row)), because your Tx/Return would be too high.   More than four, you pretty much always need an amp.

 

Example:

It's basic math why...  For low frequencies used for Tx/Return:

2-way splitter loses 3.5 dB

3-way (balanced) loses 5.5 dB

4-way (balanced) loses 7 dB

So say if you start out a 45 dBmV for Tx/Return at the tap, say an extra 2 dB for loss due to footage/distance (aka. attenuation, which depends on type of wire used) and if you add a 4 way onto that you're already to where you modem has to push 45+2+7= 54 dBmV on the modulation to transmit signals upstream.  That's not good for most modems out there.

 

Common symptoms of too high Tx/Return:

Unstable Internet (especially if there's ANY noise on the line from your home or neighbors nearby), really bad Internet speeds (talking tested on Ethernet, not WiFi), unstable upstream Internet speeds (due to packet loss), and the modem (or cable box) taking WAAAY too long to boot up versus normal (if it can boot at all).

 

I also saw @MsRaye say 30 dBmV was ok (for Tx).  That's not necessarily correct either.  30 dBmV is acceptable (but kinda low) from the tap itself outside, but if it's 30 dBmV at the CPE/modem and especially if there's splitters... then that's bad...

 

Example:

30 dBmV at CPE, with a 3-way balanced splitter.  Lets say only 1 dB attenuation for footage/distance. At tap it's...

30 - 5.5 - 1 = 23.5 dBmV at tap.  Even for the lowest "value" end-of-line taps, 23.5 dBmV way WAAAY to low and means there's probably something wrong with the power supply for all the taps on that section of the street (something wrong with an "active").

 

Conclusion?

In general, it's a balance (neither too much nor too little); but if you have to choose, it's better to have too high Rx/Forward, rather than too high Tx/Return.

Most cable wiring issues are related to leaks (ingress and/or egress, mostly ingress), too low Rx/Forward, and too high Tx/Return.  Too low Tx is rare, and too high Rx is rare (but you'll see too high Rx in the coldest of winters, possibly like our current weather lately).

I see Internet issues fairly often for >52 dBmV Tx, but you'll never have Internet & On Demand issues just because of +8 dBmV (Note: there can be other reasons why Internet would be intermittent, such as low MER, ingress, etc. But a +8 or even +10 isn't going to be the root cause.  Saying it is so just shows a lack real-world application knowledge & experience -- it would be like saying 122V AC or 119V AC is bad because it's supposed to be 120V AC).   In many such cases, there's just too many darn splitters in people's homes, and a unity-gain amp is needed to bring Tx/Return down to a reasonable level and raise up Rx/Forward a little.

 

The metrics you can see fairly easily with a bit of know-how as a customer are:

Rx/Forward Power (dBmV)

Tx/Return Power (dBmV)

SNR on Cable Boxes (SNR = signal-to-noise ratio, the difference between the average amplitude of the signal, and any "noise" that isn't the signal)

Correcteds/Uncorrectables on Modems (depends on model, but most DOCSIS 3.0 modems track these like an odometer starting from the last reboot)

 

The technical metrics for signal quality are (subject to change over time):

Forward (116 channels/frequencies in hubs in my city, 50 to 750 MHz range)

Return  (4 channels/frequencies in hubs in my city, 5 to 50 MHz range)

MER (modulation error ratio, higher is better)

Pre-BER (pre-equalization bit error rate, less than 1 in a billion is acceptable, some loss will not have effect as long as FEC can correct any errors, 1.0e-9)

Post-BER (post-equalization  bit error rate , less than 1 in a billion is acceptable, any loss will cause impacting effects but depends on how severe before it becomes noticeable. Basic FEC This is expressed as an exponential, i.e. 1.0e-9)

Echo (measured in dBc. is helped by putting on terminators on open ports, especially on house amps and on taps)

Group Delay (measured in nanoseconds)

ICFR (measured in relative dB)

 

CERs/Uncorrectables are basically a more crude measurement of Post-BER.

CCERs/Correctables are basically a more crude measurement of Pre-BER.

That's not precisely mathematically correct, but for application purposes it's close enough to be synonymous.

 

 You may have noticed SNR missing from the second list?  That's because SNR is a really crude measurement, it's only the difference between the signal power and the background noise, and is not a real measurement of signal quality (or noise, really).

 

 

 


I should really do an ingress/egress leak noise topic one of these days... would you all want me to do that?

My postings on this site are my own, off-the-clock, and don’t necessarily represent TWC’s/Charter's strategies or opinions.
Expert

Re: To Much Signal?

Only problem is that most amps including those in the outside plant were designed 20+ years ago when there was little to no QAM modulation and there was a single qam data channel .

 With the recent addition of 192 mHz ortho modulation, that's only going to get more critical.

Next issue is that mostsplitters and  amps are not properly installed with open ports.

 For reverse loss and port isolation, they must be terminated with 75 ohm resistive loads. Better yet, only sized for the number of needed ports. Had numerous ones here that had open EMTA bypass ports and some had a 2 way splitter in front of those amps... found "twist on" F connectors, amps in series with amps rather than running the hardline and a 4-8 port tapoff into the building (and it had empty ports) THe biz class 8 line EMTA modem was 4 amps from the street.

 

It's going to get interesting where Spectrum is running the fwd DS into the 900 mHz range.

As for Ohio, the reason the US needs so much power is that they don't have reverse amp modules in every housing, Looks like one in every 3 or 4 here.  High US levels also mask noise, rather than diagnose  and eliminate it, just crank up the levels, lol.It took 2 bucket trucks over a day to locate one house with horrible US noise on my node. They had to open every amp and check test points.

 

And then you get new construction with 8 &10 port "structured wiring" installs done by electricians with no clue on qam... Yeah those usually work with an outside antenna.

 

 

 

Spectrum Employee

Re: To Much Signal?

Egress leaks & squirrels and their need to chew...  y'all have no idea...  Smiley Tongue

Maybe we should invest in trained hawks to hunt like what some airports do to help keep birds out of jet engine intakes on planes... Smiley LOL

 

But back on-topic and sardonic humor put aside, what I can possibly guess at what happened, was the amp was needed at @berryboy808'S address in the summer heat, when power levels were lower.

But now that we're having an extra-cold arctic air blast coming south from Canada, we're seeing systems running higher than normal signal power levels.  For this topic, the signal power was enough that the tech must have judged the amp wasn't needed.

I'll have to assume that was the right call (until facts say otherwise), but my only question would be what is the Return/Tx. 

Was the amp a forward gain amp just boosting forward at a loss of return? Or was it a unity gain amp and now the Tx is higher?  If there was say... a 36 dBmV Tx at the amp, then well yeah it's totally not needed as long as Rx is sufficient.  But it it was like a 50 dBmV at the amp, then there's really no room for a 3-way splitter or whatever.  It depends. 

 

Spoiler
We nickname this ballpark/quick mental math estimation as "cable math", and it's something we train new-hires on (or try to as much as possible anyways).  Sounds challenging but it's really kind of common sense when you think about energy flow:  splitting uses up power in both directions, an amp adds extra power but that energy has to come from somewhere (power transformer brick, plugs in a regular AC power outlet).  If you can not have to have an amp, sure great!  But if you need it, you need it.  Again, it depends.

 

All specs tables from manufacturers of amps, LEs, RG500, RG11, RG6, RG59, whatever are done based with an assumption of temperature (which is typically assuming 68F / 20C), if you lower that temperature it SHOULD be common high-school-physics-level knowledge that power levels are going to rise as the electrical systems become more efficient and attenuate less (signal loss per distance gets better and decreases).

 

Spoiler
So appointments being put in just because it's like,  "ooooh no +10 and that's RED on my software tools!", to me just shows a need for additional training for Customer Service Representatives (CSRs, phone/chat support).  No matter what business you run, training & retaining experiences employees is a key factor to success. 

 

From a customer perspective (which would apply to these forums), just please understand that being a little high on Rx/Forward, especially in very cold weather is kind of expected.


Here's my modem right now, just for the heck of it (hope the copy & paste works here)

1LockedQAM2561567 MHz11.4 dBmV39.4 dB655298143
2LockedQAM2562573 MHz11.2 dBmV39.4 dB831075903
3LockedQAM2565591 MHz11.7 dBmV39.6 dB820069854
4LockedQAM25618669 MHz11.7 dBmV39.3 dB996535644
5LockedQAM2563579 MHz11.3 dBmV39.4 dB997110405
6LockedQAM2564585 MHz11.6 dBmV39.6 dB1063652626
7LockedQAM25613639 MHz11.8 dBmV39.3 dB942517304
8LockedQAM25614645 MHz11.8 dBmV39.2 dB974693775
9LockedQAM25617663 MHz11.9 dBmV39.3 dB1094143075
10LockedQAM2566597 MHz11.8 dBmV39.7 dB1228101327
11LockedQAM25619675 MHz11.5 dBmV38.9 dB1265805395
12LockedQAM25620681 MHz11.6 dBmV39.2 dB1305081108
13LockedQAM25623699 MHz12.1 dBmV38.9 dB1143609166
14LockedQAM25622693 MHz12.2 dBmV39.1 dB1409426668
15LockedQAM25616657 MHz12.0 dBmV39.3 dB1502124491
16LockedQAM25621687 MHz11.9 dBmV39.2 dB1254546311

 

1LockedATDMA725120 Ksym/sec37.0 MHz45.5 dBmV
2LockedTDMA692560 Ksym/sec19.4 MHz46.0 dBmV
3LockedATDMA705120 Ksym/sec24.2 MHz46.0 dBmV
4LockedATDMA715120 Ksym/sec30.6 MHz46.4 dBmV

 

 

See what I mean?  "non-impacting"  (also, lol at all the Correcteds, modem's been running for awhile).

 

The other possible issue with extra cold weather (esp. for longer periods of time) is that things contract, physically.  If there's no room to give or flex, then things get pulled apart (aka. "suckout") in the cold, and that then causes egress leaks and intermittent upstream issues as the return signal goes in & out.

My postings on this site are my own, off-the-clock, and don’t necessarily represent TWC’s/Charter's strategies or opinions.
Expert

Re: To Much Signal?

If the amps are running a pilot channel for ALC there is very little change as ours do.

All bets are off if it's thermal ALC...