Orbi WiFi 7 RBE973
Reply

Re: Problems with CM1150V upstream on Comcast Xfinity

sleat
Tutor

Problems with CM1150V upstream on Comcast Xfinity

There are a number of past threads (now closed, or I'd be replying to them) about problems people have had with the upstream path on a CM1150V with Comcast Xfinity.  I believe I understand the underlying issue for most of these problems, after having been engaged in the same battle for several months.  There are several pieces to the puzzle, so bear with me.

 

First, note that the CM1150V uses the BGA3131 amplifier chip from NXP Semiconductors (the spinoff of Philips) to drive the upstream path.  This chip is designed to be a DOCSIS 3.1 upstream amplifier, and seems a perfectly reasonable choice.  In reading the datasheet for this chip I noticed that it is spec'ed to operate from 5MHz to 205MHz.  Hmmm... interesting numbers.  Where do they come from?  So some delving into the DOCSIS 3.1 PHY specification yielded the following (excerpted, emphasis mine):

 

7.2.3
Upstream CM Spectrum

[...]
The CM modulator MUST support upstream transmissions from 5 to at least 204 MHz and agile placement of the OFDMA channels within that range.
[...]
The CM MUST be capable of transmitting 192 MHz of active channels when operating with the 204 MHz upstream upper band edge.
In DOCSIS 3.1 upstream mode the CM MUST be capable of transmitting OFDMA channels and legacy SC-QAM channels at the same time (as controlled by the CMTS). In all cases the CM is not required to transmit legacy SC- QAM channels above a frequency of 85 MHz.


7.2.4
Upstream CMTS Spectrum

[...]
The CMTS MUST support upstream transmissions from 5 to at least 204 MHz and agile placement of the OFDMA blocks within that range.
The CMTS MAY support additional spectrum beyond 204 MHz for the upstream.
The CMTS MUST capable of receiving 192 MHz of active channels when operating with the 204 MHz upstream upper band edge. In DOCSIS 3.1 upstream mode the CM is capable of transmitting OFDMA channels and legacy SC-QAM channels at the same time (as controlled by the CMTS). In all cases, the CMTS MUST NOT configure the
CM to transmit legacy SC-QAM channels above a frequency of 85 MHz.

-----------------------------------------------------------------------------------------------

So it makes sense that the BGA3131 is designed to operate up to 205MHz.  However, notice the sentences I've highlighted.  In my case, Comcast is running four upstream QAM channels (no upstream OFDM) at frequencies of 173MHz, 273MHz, 303MHz, and 367MHz.  So not only are they asking the CM1150V do do something it was never designed to do, they are blatantly violating the DOCSIS 3.1 spec!  In fact, it would probably be just fine running the QAM channels up to 205MHz, so the lowest channel should be ok.  It's not at all surprising that the other channels are failing, especially since the CMTS is calling for transmit levels between 46.8dBmV and 48.0dBmV. (That's with no splitters in the path, so I can't use my settop box.  If I add a splitter to be able to watch cable TV, the CMTS will call for another 3.5dBmV or so to make up for the splitter loss and I lose my internet service completely.)  What's amazing is that it is working at all, so kudos to NXP for exceeding their own performance specs.

 

I've discovered that the high frequency performance of the BGA3131 improves with temperature, resulting in some rather odd behavior.  Before I understood the root cause, I observed that if the ambient room temperature was lower than about 60deg F, the problems were worse.  More interestingly, upstream traffic will cause the chip to heat up, so there's a sort of bootstrap situation where you have to warm the chip up by generating some traffic, and as long as you keep a moderate level of upstream traffic going, things continue to work.  If you let it cool down, you have to start over.  At present, in addition to removing all splitters, I can only keep it working by wrapping the modem in several layers of bubble wrap and plastic foam.  Probably not great for the other components in the system.

 

It begins to make sense that people are reporting quite a variety of symptoms.  If you are in a warm environment and you are lucky enough with the Comcast infrastructure that you don't need high transmit levels, you could be running a CM1150V and never have any problems.  Or, you could have every level of flakiness between that and failing completely.

 

So this is fundamentally Comcast's fault.  However, I bought the CM1150V (and paid to have the first one shipped back, only to find that the replacement was slightly worse) on the strength of Netgear's promise that it was "certified" to work with Comcast Xfinity.  So this is not simply an engineering problem, but a corporate-level business problem.  If Netgear is going to make promises of their products working with Comcast services, they bloody well ought to be talking to Comcast and holding their feet to the fire to operate their plant within the DOCSIS specifications that Netgear designed to.  The advertising copy suggests that the CM1150V was designed explicitly to work with Comcast Xfinity, in which case somebody screwed up big time.  To quote that famous line from Cool Hand Luke, "What we got here is a failure to communicate!"  (Both on the hardware level and on the corporate level.)

 

As of Nov. 25, 2020 I'm still pursuing this with both Comcast and Netgear.  Somehow I'm not optimistic about getting Comcast to clean up their act.  Failing that, I'm hoping that Netgear will be sufficiently embarassed by their false advertising that they will refund my money.  Time will tell.

Message 1 of 14

Accepted Solutions
decosse
Apprentice

Re: Problems with CM1150V upstream on Comcast Xfinity


@sleat wrote:

So it makes sense that the BGA3131 is designed to operate up to 205MHz.  However, notice the sentences I've highlighted.  In my case, Comcast is running four upstream QAM channels (no upstream OFDM) at frequencies of 173MHz, 273MHz, 303MHz, and 367MHz.  So not only are they asking the CM1150V do do something it was never designed to do, they are blatantly violating the DOCSIS 3.1 spec! 


@sleat I've read this thread multiple times (having my own Upstream Issues!) and took this at face value;

but only just spotted this fundamental error in your analysis ..........

The Comcast Upstream frequencies are not as you indicate, you are a factor of 10 off 

i.e. they are 17.3MHz, 27.3MHz, 30.3MHz and 16.7MHz 

 

i.e. 17300000 Hz (5 zeros) = 17,300,000Hz or 17.3MHz 

 

So actually operating well within the DOCSIS 3.1 Spec and well within the design capability of the Chipset. 

 

As I've said I'm having my own Upstream Issues with my CM1150V (and actually got a replacement from Netgear) and there is definitely a problem - but it's not because it's being asked to operate outside its spec. 

View solution in original post

Message 7 of 14

All Replies
sleat
Tutor

Re: Problems with CM1150V upstream on Comcast Xfinity

Postscript:

It's also curious that I'm paying for 17Mbps upstream (and that's about what I get when everything is working), but Comcast is running four QAM channels at 5120 symbols/sec each.  Even if they're only running QAM16 (i.e., 4 bits/symbol), that's 20.48Mbps raw signalling rate per channel, so 17Mbps throughput seems about right for one channel.  If they would just give me one QAM16 channel below 205MHz, I'd probably be happy.

Message 2 of 14
FURRYe38
Guru

Re: Problems with CM1150V upstream on Comcast Xfinity

Have the ISP check the signal and line quality UP to the modem. Be sure the ISP provisions the modem correctly.
Be sure there are no coax cable line splitters in the between the modem and ISP service box. 
Be sure your using good quality RG6 coax cable up to the modem. 
Be sure to power OFF the modem for 1 minute them back ON.
https://community.netgear.com/t5/Cable-Modems-Routers/General-info-and-Troubleshooting-for-Cable-Mod...

 

Keep on them. 

Message 3 of 14
Acnas
Aspirant

Re: Problems with CM1150V upstream on Comcast Xfinity

Exact same problem I am having with mine. Just added tv service and had to add a splitter to main drop as house is not wired for cable.  Internet went to crap.  Major packet loss.  Got the Xfinity modem today in the meantime as I dont want to buy another recommended modem and have the same thing happen.  Now have a modem that is a paper weight.

Message 4 of 14
FURRYe38
Guru

Re: Problems with CM1150V upstream on Comcast Xfinity

Have the ISP check the signal and line quality UP to the modem. Be sure the ISP provisions the modem correctly.
Be sure there are no coax cable line splitters in the between the modem and ISP service box. 
Be sure your using good quality RG6 coax cable up to the modem. 
Be sure to power OFF the modem for 1 minute them back ON.
https://community.netgear.com/t5/Cable-Modems-Routers/General-info-and-Troubleshooting-for-Cable-Mod...

Contact NG support for additonal help and information. 


@Acnas wrote:

Exact same problem I am having with mine. Just added tv service and had to add a splitter to main drop as house is not wired for cable.  Internet went to crap.  Major packet loss.  Got the Xfinity modem today in the meantime as I dont want to buy another recommended modem and have the same thing happen.  Now have a modem that is a paper weight.




Message 5 of 14
decosse
Apprentice

Re: Problems with CM1150V upstream on Comcast Xfinity


@sleat wrote:

.... especially since the CMTS is calling for transmit levels between 46.8dBmV and 48.0dBmV. (That's with no splitters in the path, so I can't use my settop box.  If I add a splitter to be able to watch cable TV, the CMTS will call for another 3.5dBmV or so to make up for the splitter loss and I lose my internet service completely.)  ...

 

 Before I understood the root cause, I observed that if the ambient room temperature was lower than about 60deg F, the problems were worse.  More interestingly, upstream traffic will cause the chip to heat up, so there's a sort of bootstrap situation where you have to warm the chip up by generating some traffic, and as long as you keep a moderate level of upstream traffic going, things continue to work.  If you let it cool down, you have to start over.  At present, in addition to removing all splitters ... 

 

 


@sleat thanks for this research 

The one thing I am unlcear about is whether this only manifests itself in the Upload performance, or whether because of that upstream issue, it subsequently also cause download speed to suffer? 

 

I'm wondering if this is the issue I was having 

It had been working perfectly for almost a year 

Then my download speed would drop off to a crawl and it was always literally completely gone EVERY morning - my first indication as I was coming down stairs was that I could see my wireless doorbell camera controller was off-line before I even got to anything on a PC.  I found a power cycle on the modem would instantly bring it back only for the problem to repeat over the course of the day and ALWAYS every morning (overnight) it would be completely gone.  

The part about keeping traffic up helping (coupled with fact that overnight it was bad - my home has different HVAC systems for Living and bedrooms, so modem location is cooler during night and obviously internet not utilized either) - Aside from the power cycle, I would find a curiousity that if I forced several back to back speed tests it would come back (hence increasing the traffic component of your findings) 

Now - I had not seen your post at this stage, called Xfinity out and they checked all the lines to the street and all was good, they installed an Xfinity Modem and service has been perfect for a month now - so conclusively the Netgear Modem (as relates to my service) was a problem. After the first week of confirmed flawless performance  (vs before replacement of continuous problems EVERY day for two weeks) I contacted Netgear who were reluctant to exchange without having an online test done - I went through multiple layers of support, stamping my feet and refusing to go through the hurdles involved of having to have the Netgear re-activated by Xfinity when I had conclusive evidence that the Xfinity Modem worked and the Netgear did not. I finally got someone to agree and they generated RMA for exchange. 

Roll forward to having just received my replacement modem and I installed it yesterday 

Here are the numbers I am getting on the Upstream on the new modem (sorry I have none for the old one) 

It is running just over the range you indicated (all over 48, up to 49.5 for the highest one) - is this potential for issue? 

One additional thing I should add - after my modem was replaced with Xfinity even though all was working fine, I did some more investigation on the wiring - I have two TV boxes and there was a 3-way splitter installed; now it was an unbalanced splitter and found that the primary TV box had been connected to the -3.5 split (so the modem was on a -7 split); it did not give me any perfromance issues with the Xfinity modem, but I swapped them regardless so the Modem Cable is now on the -3.5 leg. 

So previously, it had been on the -7db connection (but had been from the outset) but the numbers below are reflected in a -3.5db connection.

Unfortunately I did not take note of the signal levels on the original Netgear modem - honestly did no diagnostics on the modem as felt it was a signal issue and called out Xfinity tech who 'resolved' the issue by replacing the modem (which indeed confirmed it WAS a modem issue - and the replacement - although only been on for 24 hrs at this point - also seems to be perfect) Incidentally even with the -7db splitter (which was in place during the Xfinity service call) the Xfinity tech deemed the signal level at the modem to be excellent. 

 

So my main question is 

Are those Upstream power levels in my new modem in acceptable range? I'm getting perfect performance comparible with my Xfinity Modem (it seems like with 3.5 splitter now, 7 before, those numbers would likely have been another 3.5dbmV higher than currently reporting, on my old configuration?) 

And generally, does the Upstream Issue outlined here manifest itself most visibly in download connection speed, not only upload? 

I guess time will tell on this replacement, but acknowledge I have undoubtedly improved the connection with only a 3.5 db split attenuation vs the 7 it had before. 

 

CM1150V_Upstream.png

 

 

Message 6 of 14
decosse
Apprentice

Re: Problems with CM1150V upstream on Comcast Xfinity


@sleat wrote:

So it makes sense that the BGA3131 is designed to operate up to 205MHz.  However, notice the sentences I've highlighted.  In my case, Comcast is running four upstream QAM channels (no upstream OFDM) at frequencies of 173MHz, 273MHz, 303MHz, and 367MHz.  So not only are they asking the CM1150V do do something it was never designed to do, they are blatantly violating the DOCSIS 3.1 spec! 


@sleat I've read this thread multiple times (having my own Upstream Issues!) and took this at face value;

but only just spotted this fundamental error in your analysis ..........

The Comcast Upstream frequencies are not as you indicate, you are a factor of 10 off 

i.e. they are 17.3MHz, 27.3MHz, 30.3MHz and 16.7MHz 

 

i.e. 17300000 Hz (5 zeros) = 17,300,000Hz or 17.3MHz 

 

So actually operating well within the DOCSIS 3.1 Spec and well within the design capability of the Chipset. 

 

As I've said I'm having my own Upstream Issues with my CM1150V (and actually got a replacement from Netgear) and there is definitely a problem - but it's not because it's being asked to operate outside its spec. 

Message 7 of 14
sleat
Tutor

Problems with CM1150V upstream on Comcast Xfinity

In a previous post, two years ago, I presented an analysis of problems with the CM1150V. (https://community.netgear.com/t5/Cable-Modems-Routers/Problems-with-CM1150V-upstream-on-Comcast-Xfin...)

 

I just returned to that post and saw the reply by @decosse pointing out a fundamental error in what I said.  That topic is closed, so I can't reply there, hence this.

 

Yes, I realized the error sometime after making that post, much to my utterly cringeworthy and mortifying embarrassment.  It's an object lesson in dyslexia combined with falling prey to the seduction of beautiful but misguided theories.  I'll never live that one down, at least in my own mind.  Not to excuse my own sloppiness, but I'd argue that Netgear is being more than a little silly in displaying the upstream frequency in Hz, not MHz.  I guarantee those last few digits are completely meaningless.

 

Nevertheless, I'm on my third CM1150V and still having problems with the modem. When the second modem I received exhibited the same problems as the first, I spent hours dealing with Netgear customer support and eventually sent them a long video demonstrating that there was a very sharp temperature threshold below which the upstream amplifier would fail.  (This was about 62 degrees F at the surface of the metal shield can over the front end components.)  This was sufficient to get them to send me the third modem, even though this was long past the warranty and support period.  Of course, the third modem had the same problem.  At this point, having other things to do in my life, I just wrapped the modem up in a blanket of bubble-wrap to keep it warm via its own self-heating and moved on.

 

It's now December and even though I'm in Northern California, it's cold.  The room that the modem is in regularly gets down to the low 50's F, and the modem is failing again.  I've ordered some replacement NXP BGA3131 upstream amplifier chips, on the chance that there's a bad batch that Netgear got into.  Incidentally, that same chip is used in the XB6-T (aka Arris TB3482G) modem that I rented from Comcast for a while during all of this.  That modem performed fine, and I think I cooled the amplifier explicitly to test this, though that's a bit lost in the mists of time so I'd have to rent another one and repeat the experiment to be absolutely sure. If that was a valid test, it suggests that either Netgear got some bad (counterfeit?) chips, or their design is somehow abusing the chip.  I'll post something here if I do indeed get around to replacing the chip.

 

BTW, @decosse, if you're listening, I apologize for not seeing and responding to your post of 2020-12-15 12:46 PM.  To answer your questions if you still care, yes those upstream power numbers look fine, and yes, the most obvious symptom is poor download speeds.  There's no easy way to see if your upstream packets are making it back to the head end, but if you're requesting downstream packets and your requests never make it upstream, then of course you won't get your data.  This is one of the things that made it so hard to understand early on.  The other notable symptom will be a large number of T3 Timeout errors in the log file.  However there seems to be a bug in their logging code in that most of the time I see it simply overwrite the last entry rather than accumulating errors, making the whole thing even more confusing.

 

 

Message 8 of 14
FURRYe38
Guru

Re: Problems with CM1150V upstream on Comcast Xfinity

So is this a heat or cooling problem? 

Is the modem failing if it gets too cold or hot? 

 

Can you please post a copy and paste of the modems connection status and event log page.

Message 9 of 14
decosse
Apprentice

Re: Problems with CM1150V upstream on Comcast Xfinity

Don't beat yourself up too bad @sleat - it's a simple error, that was all too easy to make and you've done a lot of other good analysis to attempt to resolve for yourself and others. 

 

I gave up on the Netgear - never could get to the bottom of what was making them incompatible with my local Comcast service - two modems, same inconsistent issues (would work perfectly for a while and then have issues) I had no issues with the Comcast supplied modem, even with splitters restored and reluctantly just stayed with that. 
I believe that there was some marginality over my local service that the Netgear was unable to accommodate, while the Comcast modem could, comfortably. 

Honestly the whole support saga and general warranty policy has completely soured me of Netgear as a whole and I won't be back. Have moved to Asus for my Router (I did not have any issues with my Netgear Router system - left it behind as part of my whole smart home setup - it's just that warranty/support thing that will preclude me from ever buying another of their products ) and now looking at Arris or Motorola for the modem at this point to save the exorbitant Comcast rental fees. (I'm now in WA, no longer at original address where I had reported on this) 

Message 10 of 14
sleat
Tutor

Re: Problems with CM1150V upstream on Comcast Xfinity

The modem fails when the BCM3131 upstream amplifier gets too cold, which is somewhere between 50F and 60F.  You can dig through the whole saga in the past posts if you want to know more, just ignore the whole aspect where I went down the rabbit hole regarding Comcast upstream frequencies.  That was just a complicated distraction resulting from a simple mistake.

Message 11 of 14
FURRYe38
Guru

Re: Problems with CM1150V upstream on Comcast Xfinity

Interesting. Didn't know that cold temp would be a cause of that failure. 

 

Is the modem in a room that is always cold? I suppose one work around would be to have a heater on it if you need this modem to be in a cold room. Kind of a corner case issue here though. Most rooms on average are 65-70F I would guess and the modem running normally would be warmer then that internally. Usually the modems run fairly warm on there own so shouldn't be a cooling issue. Very interesting though. 

Message 12 of 14
ocadle
Aspirant

Re: Problems with CM1150V upstream on Comcast Xfinity

We moved to our current home just over 2 years ago. For the first year, our modem was connected to a Comcast drop in our finished basement which has a near-constant temperature of 72F. No issues then. In the Fall of 2022, I moved the modem to my office above the garage. This room has electric baseboard radiators for heat and they are only on for part of the day. In the dead of winter, this room can get down to the low 40s. Almost immediately after moving the modem to the office, we started having problems with the Internet connection going down almost every morning and definitely every weekend. It took me weeks to attribute the issue to the cold temperature. I thought I was crazy thinking that a room temp in the 50s would cause the modem to lose its upstream connection. Seems I'm not crazy after all.

Message 13 of 14
FURRYe38
Guru

Re: Problems with CM1150V upstream on Comcast Xfinity

Thanks for the feedback. 

I guess for those that have there models in cold place, get them moved to warmer places or maybe have a small cubicle heater that could be placed near the modem to keep it warmer in this colder conditions. 


Good Luck.

Message 14 of 14
Top Contributors
Discussion stats
  • 13 replies
  • 3335 views
  • 1 kudo
  • 5 in conversation
Announcements

Orbi WiFi 7