dslreports logo
 
    All Forums Hot Topics Gallery
spc
Search similar:


uniqs
8493
richemont86
join:2013-03-11
Seattle, WA

richemont86

Member

Low Upstream Power Levels

My upstream power levels are ~25 dBmV and I just had a Comcast tech come out to try and resolve the issue unsuccessfully. He added an attenuator to the back of my modem, but from what I understand that only makes the signal 'look prettier' than is actually being received by the CTMS. Regardless, with it attached I get ~33 dBmV with my downstream levels being affected as well (reducing them).

Signals with attenuator:



Without attenuator:



As you can see, I also get a T4 timeout on the 37mhz upstream channel. I also get frequent T3's.

The tech mentioned that I am the last tap on the line and no other houses are connected to it besides mine. He also said that the level at the pole was 22dBmV, which was lower than normal. He didn't say that he was going to escalate the issue though, but that he would talk to some other techs about it and gave me his card.

I also have a direct connection (no splitters) to the pole. I've tested several modems, but am currently using an SB6141, all modems had the same issue.

I'm not really sure where to go from here.

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

EG

Premium Member

Hello again !
richemont86
join:2013-03-11
Seattle, WA

richemont86

Member

Haha. Hey man. You're just all over the place, huh? Any idea what should be my next step?

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

EG

Premium Member

As stated over there, the premise techs need to escalate to maintenance.

PeteC2
Got Mouse?
MVM
join:2002-01-20
Bristol, CT

PeteC2 to richemont86

MVM

to richemont86
Well...it needs to be elevated. There is nothing that can be done at your end to improve your signal, simple as that. If they can not provide a better signal, then perhaps cable is not the solution for your internet service.

I agree, the attenuator is of no real help...kind of a cable-placebo...
CableLife2
join:2013-03-11

CableLife2

Member

In layman’s terms, your upstream power level is how loud or quiet your modems upstream voice is being received by the CMTS. In my system, upstream power levels of 38-50 is operational range.

(remember, this is all in layman’s analogies to make it more understandable)

A lower level effectively means your modem is screaming at the CMTS and the CMTS is lowering its listening voice to hear your modem, achieving a level below acceptable range of 33.

Each passive device in-line between you and the nearest plant amplifier will reduce your modems upstream voice. Once it reaches that amp, its power level is then combined with the power level the amp is tuned to.

For example, working backwards from the plant amplifier=
Plant return amp is set to +13dBmV
+3 for two taps between you and amp
+4 for 200 feet of plant coax between you and that plant amp
+10 for the 10v4port amp servicing you

In this scenario, your modem would read an upstream power level of 30. Since you’re reading 25dBmV without the attenuator, it sounds like the return needs to be balanced on the amps in cascade from the node to your amp. That sounds pretty low for any plant design scenario can conger up.

None the less, the attenuator he installed has a value to it. 3,6,9, and 12 are the most common. It is simply just another passive inline between your modem and that nearest plant amp to lower the modems voice. It is a perfectly fine application and are used quite often to tune the return power levels into the sweet spot. My concern is, why he didn’t use a larger value attenuator, a 12 would have put you in range for my system. My other concern is, a power level of 25 at your modem suggests the plant is pretty far out of balance.

I would call for another service call, armed with a little more information this time and see what you can achieve

Good luck

richemont86
join:2013-03-11
Seattle, WA

richemont86

Member

Thanks for the feedback, guys. Comcast's automated system called me to get feedback on the service tech and whether my issue was resolved, so I chose that it hadn't. According to the review system, they'll call me back to find out why it wasn't resolved so I'll try to escalate it then. If that doesn't work then I'll call this number the tech gave me with an ID# that is supposed to help get attention to the issue.

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

EG to CableLife2

Premium Member

to CableLife2
Putting attenuation at the back of the modem really accomplishes nothing for the upstream / return unless there is noise ingress that needs to be overcome right there at the premises area. And this is not the proper way to fix that. The possible noise should be located and eliminated.

A "larger" value attenuator may cause the forward power level to be out of spec.
CableLife2
join:2013-03-11

CableLife2

Member

I'm not sure where you get your information EG but almost nothing you just said is true. An attenuator increases the power level by the value written in the nomenclature on the side of it. A 6 will increase the upstream power from 40 to 46. They are used to move a modem into spec based on the tuned power level of the nearest return amp. Simple as that, there is nothing more to that conversation if its being used correctly.

It will have an effect on noise and I can go into that but it's unnecessary because using it for that purpose is as ridiculous as explaining it in detail would be

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

1 edit

EG

Premium Member

WOW ! I think this is a matter of us not being on the same page. I will bow out of this one as to not throw the OP's thread off course by arguing this.

AnonMan
@comcast.net

AnonMan

Anon

I am with EG on this one.

The sad part is I have seen many techs and even some maint people think otherwise due to improper training or technical understanding.

That said an attenuator is generally only good to bring too much signal down to help get rid of too much RF, OR to help eliminate some noise, for example if between a customer and the ped is subject to a lot of noise putting the attenuator at the ped would help as it would cause the modem to yell louder and strengthen the upstream signal between it and the ped. Once the signal hits the attenuator at the ped it goes back to the value as without it thus it would still be the same to the rest of the run. (keep in mind we are talking US only here as this could also hurt DS so the application of said attenuator and size vary)

Of course any equipment will report the modem as sending higher because of the attenuator but it really is only higher up to the attenuator because the attenuator is adding resistance to raise the level. Once it is past that the resistance is gone.

Take this:

Modem -------- (Attenuator)--------Ped------->to head end.
US pwr is 25 -3db USpwr25

The US will become and read 28 on the modem and actually be transmitting at 28 but once it hits the attenuator the -3db does the same for the other direction and results in it really being 25 but reports 28 due to the added resistance of the last little hop. Only the points between what is being attenuated and the attenuator are technically changed. The modem yelled louder because of the 3db resistance but the 3db resistance quiets it down.

Putting the attenuator at the modem does nothing unless DS levels are too high making RF noisem but give a false impression of something as only between the attenuator and the modem will have the increased signal because once it hits the attenuator it goes back to the normal value.

This is why forward path attenuators exist for example. They are generally better to put at the customers end. Regular attenuators are generally better at the ped or amp.

In this case no attenuator will help. The plant/amp or design is out of spec. The signal will be just as weak at 25 once it is past the attenuator as attenuators reduce levels so that 28 it made the modem yell at because it added more resistance turns into 25 once on the other side because it is reducing it.

Again this is all only talking about US, DS is a bit different but same principle. You can lower DS level at the modem but you are not lowering it for anything except between attenuator and the modem. This DOES help if DS is high because too much power is just as bad as not enough. So the run all the way to the attenuator again remains the same and only between it and the modem change which for DS shaping is the desired result.

This is not made to offend anyone's knowledge or skill level and regardless of what they do for a living as again I have seen maint techs swear up and down on things that they are wrong on. Attenuators have specific purposes and in this case they do not apply to the issue properly.

To remain on topic though this has screwed away thanks to me too now lol the problem is likely an amp cranked up too high.

It is not likely the entire run that needs to be worked but just the amp that feeds you/your ped needs to be adjusted.

And for the one who said why didn't that add a 12dB one, well because that would have put his DS out of kill.

tshirt
Premium Member
join:2004-07-11
Snohomish, WA

1 recommendation

tshirt to richemont86

Premium Member

to richemont86
OP ignore what CableLife2 See Profile said, what EG See Profile andPeteC2 See Profile have said is correct.
In this case the attenuation only makes the modem side power look better and has no positive effect in the return path.
You need a tech (either the one who was just there or a new one) to put a ticket in with line maintainace to change the tap levels by adjusting something between you and the node.

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

EG to AnonMan

Premium Member

to AnonMan
BINGO !! Very well stated sir !! Thank you for having the patience that I did not !
EG

EG to tshirt

Premium Member

to tshirt
said by tshirt:

, what EG See Profile andPeteC2 See Profile have said is correct.

AnonMan stated it best !

I've seen some fine techs otherwise not understand this concept...
CableLife2
join:2013-03-11

CableLife2 to AnonMan

Member

to AnonMan
Interesting, as EG mentioned previously, I think we are not on the same page as well. We effectively said the same thing only I emphasized I was keeping it in simple with laymans analogies. I don't disagree with anything you've said and can't quite see how it disagrees with anything I've said.

Ok maybe 2 things i disagree with. A 12 db sub band will effectively quiet the modems upstream voice by adding 12 db of resistance so it is at an acceptable Tx level when it reaches and is pulled by that first return amp.
A simple and effective solution to keep a modem operational until plant return levels can be adjusted.

And second, I don't now how a system could assure service quality by spot balancing. An HFC network is designed under the concept of unity gain. How can one assure unity gain if the inputs/outputs of all amps in cascade to a customers address are not swept. So that's number 2. I also disagree that this issue would be resolved by visiting the last amplifier

AnonMan
@comcast.net

AnonMan

Anon

It very well may not be just the last amp, but in a good amount of cases it is otherwise a significant number of more customers would have issues and thus be noticed faster and corrected.

It could also be the ped feeding the OP is an AMP.

Many things "could" or "could not" be the issue. Simple fact is no attenuator is going to fix this short term or long term short of being an attenuator jumper on an amp itself that is bad or wrong etc.

Your analogy of using a 12db one temporary to keep the modem operation until plant level is adjusted is flawed by the originally stated fact that it has no actual impact on the connection by the time the signal hits the CMTS.

The only thing an attenuator is good for on customers end is if DS is HOT and in that case a forward path attenuator "should" be used (given a fat splitter is not used if multiple TV outlets instead etc.). The only other time an attenuator should be used is in the ped (which a FWD Path one can be placed if not wanting on customer side or regular one depending on the goal if US noise or hot DS or both).

That said OP needs to request a tech. If any of the techs come out and claim no issue tell them to run a full test on the nice little expensive meter they have. It will fail the DOCIS test and that requires them to put in a ticket. I would assume if this was beyond just the last ped or amp a good deal more of people would have issues and this have been caught but as not it of course makes it harder for you to fight and get someone to open eyes and do the darn job that needs to be done.

If all else fails contact corp and they will open a ticket via the national center whom will directly contact local maint and force them to go out regardless of any reporting or records.

PeteC2
Got Mouse?
MVM
join:2002-01-20
Bristol, CT

PeteC2

MVM

...and so, richemont86, after all the various analysis...you can see that this is what everyone agrees with: Call 'em back and have them escalate your issue. This won't get corrected without a visit.
CableLife2
join:2013-03-11

CableLife2 to AnonMan

Member

to AnonMan
I know nothing of Comcast topology as i have never lived in a comcast service area. I dont know what goodies they utilize to manipulate their HFC network into reliability. I do, however, know cable with regards to HFC topology. I just happened across this question while eating lunch today and thought I would help the OP out.

In the system in my service area, your examples prove false. In this system, a DOCSIS modem will typically fail to block sinc if upstream power levels are less than 32dBmV at the modem. The CMTS will simply reject anything under 32 because, for ease of understanding, it doesnt want to be screamed at. Slapping a sub band 6 somewhere inline between the modem and the tap, (the back of the modem proves as successful as anywhere else such as dmark or wallplate). The modem Tx will now read 38dBmV and it will stay online, baring no other issues, for the duration of that cable companies natural born life. Stated simply, 32dBmv directly to modem - no worky. 32dBmV + Sub band 6=Worky just fine. Now this is in a system im familiar with. If thats not true with comcast then this debate is fruitless.

I suspect your experience is very different than mine because you allude to standards and ideas that seem very hazardous to maintaining service quality.

1. Suggesting that an attenuator should be used to overcome upstream noise = mathematically logical but reprehensible in practice

The upstream noise should be fixed, not combated with an attenuator. That's just poor tradesman-ship.

2. the possibility that he is serviced from a pedestal with an amp yet is showing 25tx at the modem? Thats either lack of understanding or just throwing items against a wall to see what else might stick. The lowest value tap directly off an amplifier one should see is 17. Adding 3 for drop loss, that would mean that amplifier is transmitting at 5. Not impossible but I have a some beach front property to sell you in Nebraska if you want to believe it to be a logical point of discussion.

3. tuning an amp without considering what amp is before it and what amp is after it is spot balancing and will get the techs fired in the system out here. Ive met one personally and he now works for the Electric utility here. Regardless of who shows effect and where, levels issues are resolved by balancing forward and return levels from node to amp servicing the customer - not by cherry picking the amp in between that is feeding customer complaints.

4. The numerical value of the return pad in an amp is the value of resistance added against the return amplifier thus creating the tx level of that amp. The level produced is going to be set at the amp and increase as you get further out in taps and other passives from the amp. A service tech doesn't typically check, or have a need to check the plant map before resolving a house issue. That said, he wouldn't typically know what level of Tx he should expect out of the tap. The rule of thumb ive heard them say is, tap value + 15, give or take 3. Which translates to, on a 17 value tap, they assume 33 +/- 3 is about right. Anything within that range, a maintenance tech will tell them to quit crying and get the job done.

And this is absolutely true. It is the service techs job to amplify or attenuate forward and return levels as they distribute to the premise equipment. What the are given off the tap, baring no MER/BER issues, is often perfectly fine to work with. Your suggesting that this job WILL require a maintenance ticket is completely flawed. If he is on an end of line 4 value tap, 25 tx with no passives is Absolutely logical. You should be seeing that if you know this stuff like your first post suggested. Your logic has become progressively flawed since that first post however.

Anyway, i think the OP gets the idea.
richemont86
join:2013-03-11
Seattle, WA

richemont86

Member

No real need for debate..

"In physics, attenuation (in some contexts also called extinction) is the gradual loss in intensity of any kind of flux through a medium. For instance, sunlight is attenuated by dark glasses, X-rays are attenuated by lead, and light and sound are attenuated by water.
In electrical engineering and telecommunications, attenuation affects the propagation of waves and signals in electrical circuits, in optical fibers, as well as in air (radio waves)."
Source: »en.wikipedia.org/wiki/At ··· enuation

PeteC2
Got Mouse?
MVM
join:2002-01-20
Bristol, CT

PeteC2

MVM

said by richemont86:

No real need for debate..

True. Your signal sucked (no big tech terms required)...it still is not good. You will more likely than not continue to have connectivity issues based on what you posted.

EG gave the best advice long before this: "As stated over there, the premise techs need to escalate to maintenance."
CableLife2
join:2013-03-11

1 edit

CableLife2

Member

:headslap: somehow in all this posting, it still isnt clear so let me try again.

"EG gave the best advice long before this: "As stated over there, the premise techs need to escalate to maintenance.""

NOT NECESSARILY

Unless there is something youre seeing that we havent discussed, 25 off tap is not singularly cause for alarm. We dont know know the plant design scenario creating this output. It could be perfectly logical or it could be crap. He most definitely needs a service tech who knows his job and that determination can be made on site. I just dont think its right to create expectations before a premise tech has made an assessment. I think the OP is armed with enough information now to assure the premise tech does his job.
richemont86
join:2013-03-11
Seattle, WA

richemont86

Member

UPDATE: Just got done arguing with a local tech supervisor about this whole attenuator vs no attenutator business. I told her regardless of the signal strength, with or without it, I still get a T4 timeout on the other upstream channel, so the problem is not solved. She's reluctantly agreed to send someone out to look at the tap I guess.

I get the sense I'll be having the attenuator conversation again soon. And that Comcast really hates to escalate things to line maint.

AnonMan
@comcast.net

AnonMan

Anon

Have you had someone tell you what the US SNR is? Just for giggles find that out. If it's low that's your problem.

The next thing is to ask to have the maint supervisor contact you.
I have one for my area and also 4 of his guys as well. When I had issues I could call one of them and they could see it happening to the entire node but because it was so intermittent and didn't last long no one else would ever pay attention but they did and fixed it right up. Just took having someone on phone while it happened who had access to see it.

The other option is they can put your modem on a watch list. After my original issue was fixed a new one creaped up and was caught only after being monitored as the US SNR again was dropping but happening for only seconds so the entire node didn't always drop. This turned out to be another issue caused by someone else lol.

We know it's a pain. It took me 5 months of hassle to get good service but now I have good contacts and am fairly happy, minus prices lol. It sucks, yes but sometimes you have to fight a little or do work on your own to get things done right.
richemont86
join:2013-03-11
Seattle, WA

richemont86

Member

Problem solved! Line maintenance came out and swapped out the tap, which he said had significant corrosion, and tweaked one of the amplifiers down the line that wasn't operating like it should. My signals at the moment (without attenuator):



tshirt
Premium Member
join:2004-07-11
Snohomish, WA

tshirt

Premium Member

That looks excellent.
You should be trouble free now, and just in time for a big speed boost coming in the next 10 days.

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

EG to richemont86

Premium Member

to richemont86
Happy surfing !

AnonMan
@comcast.net

AnonMan

Anon

Sounds like the main issue was the amp... though I am sure the tap probably need to be swapped still or they wouldn't have done.

Anyway all looks good now. Wish mine could look like that...

I am +18 DS and 38.50 US... lol good thing for fwd path attenuators and in my case maint can't do nothing. First amped ped on the line and has to be hot due to next hop. My SNR is great though lol 42.. Only thing a pain to deal with is US SNR, you can't see it and is annoying to catch. Hopefully yours is good though.

If you have any issues still find out what the US SNR is and try to get from them on phone during or around time of issue also but I suspect you are good now after the amp tweak. Enjoy your internet!
richemont86
join:2013-03-11
Seattle, WA

richemont86

Member

Thanks for the help and getting me informed guys!

PeteC2
Got Mouse?
MVM
join:2002-01-20
Bristol, CT

PeteC2

MVM

said by richemont86:

Thanks for the help and getting me informed guys!

Glad that you got it successfully resolved!

Just a note: Not to be argumentive with CableLife...I get that one can get overly whacked out over modem-page stats, and signal levels by no means have to be "perfect" in order to have good service.

In your case however, you were noticing problems...when you combine that with what the modem stats had shown, then it is clear that things are not "ok" as they are, and assistance is justified.

As you saw, even though Comcast was "reluctant" to send a tech out...it was the right thing to do and the only way you would have gotten a resolution.