XG1v4 DVR / XG2v2 STB and Moca 2.0 home network setup advice

[krkaufman]

this is the moca PoE I have separating the two coaxial networks.

and this is the type of splitter

note they make them able to split more devices if needed.

If correctly recommending the Holland GHS-PRO-M series splitters for their MoCA compatibility, stick with them; they have models with 2 to 8 outputs. (specs PDF, one source)

 

[krkaufman]

coax > splitter > Xfinity modem (XB6)|moca adapter > XG1v4 DVR.

I'm unclear on exactly how these devices are coax-connected. Is this a 2-way splitter or 3-way? Are the splitter outputs connected to the modem and MoCA adapter? Is the XG1v4 connected via the "Device"/"TV-STB Out" port of the MoCA adapter?

The RF pass-through port of the MoCA adapter should not be used to connect any MoCA-enabled device such as an Xfinity X1 DVR or client box, as the MoCA adapter uses a diplexer internally and MoCA signals are severely attenuated via the port. You'd need an additional coax split, or possibly try connecting the modem via the MoCA adapter's pass-through port. If unused, the MoCA adapter's unused pass-through port should be capped with a 75-ohm terminator.

Also in use are two Xfinity XG2v2 set top boxes that allow playback of content from the XG1v4. The second moca adapter is plugged in upstairs where one of the XG2v2 is in use and setup as follows... coax > moca adapter > XG2v2.

Same situation as above, the XG2v2 should not be connected via the pass-through port of the MoCA adapter. A MoCA-compatible splitter is required to connect the MoCA adapter and Xfinity client box separately to the coax. The MoCA adapter's unused pass-through port should be capped with a 75-ohm terminator.

Moca is turned OFF on the Xfinity modem.

Even with MoCA disabled on the Xfinity gateway, I prefer to install a separate MoCA filter in these circumstances to prevent interference should MoCA be accidentally re-enabled on the gateway.

 

[krkaufman]

In the room where the AT&T is, do you need to use the coax for any cable functions? If not, the solution is actually super simple--connect the motorola in that room and then in the attic, disconnect that particular coax feeding that room and connect the other motorola to that cable. This way, you have nothing on that cable except the motorolas and you should hit 1Gb no problem.

And this would be necessary for the bonded MoCA 2.0 adapters to hit their maximum throughput, up to 1 Gbps, achieving "TURBO" mode when no other MoCA nodes are present on the coax.

Relative throughputs for each MoCA spec:

• MoCA 1.1 w/ GigE ports can achieve up to 150 Mbps.
• Standard MoCA 2.0 up to 400 Mbps; 500 Mbps TURBO
• Bonded MoCA 2.0 up to 800 Mbps; 1000 Mbps TURBO
• MoCA 2.5 w/ GigE ports up to 1000 Mbps

Given the above, only the goCoax MoCA 2.5 adapters would likely be capable of hitting your required 1 Gbps requirement in a setup with 3+ MoCA devices on the same coax plant (i.e. precluding TURBO mode for MoCA 2.0).

Oh my. I don't think you're going to have much luck with 2 moca networks on the same coax. It doesn't work well with powerline technology and I don't think it will with moca either.

It's theoretically possible; whether the specific hardware (and its software) allows for the supported configuration is another question. Per pg. 6 of the MoCA 2.0 specs doc (and pg. 8 from the 2.5 specs):

The MoCA 2.0 frequency plan defines, within the new extended band D, two sub-bands for independent network operation. These sub-bands comprise the D-low and D-high, as follows:

Sub-band D-Low (DL): 1125 to 1225 MHz edge to edge (100 MHz wide)

Sub-band D-High (DH): 1350 to 1675 MHz edge to edge (325 MHz wide)

Guard-band between sub-bands: 1225 to 1350 MHz (125 MHz wide

Now, if you're connecting both xfinity and att to the same lan, there may be dual dhcp server issues unless xfinity is not using one, and that can also be an issue, but would appear as stuff working 100% and then just not working at all. Hope this helps!

Right, the Motorola/ATT and Xfinity MoCA networks must not be mixed, any more than you'd connect an Ethernet cable between the ATT and Xfinity gateways. Short of dedicated coax for the "ATT" MoCA network, you'd need to get the Xfinity MoCA network operating in Band D-Low and the ATT MoCA network operating at Band D-High -- and privacy on the ATT MoCA network would be a good idea.

So, can the X1 devices be configured to only operate at Band D-Low? Then, if so, can the MM1000, ECB6200 or goCoax adapters be configured to strictly operate at Band D-High, to functionally isolate the ATT MoCA network from the Comcast network? Your adapter's vendor should be capable of assisting with this setup, or needs to be ... given Comcast is the largest cable provider in the country.

Reading the MoCA 2.5 specs, I'm not sure how this dual network setup would be affected were you to use the goCoax adapters. One would hope their backward compatibility would allow for dual operation, if reducing their maximum throughput to that of bonded MoCA 2.0.

Also, the dual network operation possibility would be affected if your Xfinity Internet connection is DOCSIS 3.1 and using frequencies above 1 GHz, consuming the space that would otherwise be dedicated to the MoCA D-Low sub-band.

 

[Michael Nunes]

If correctly recommending the Holland GHS-PRO-M series splitters for their MoCA compatibility, stick with them; they have models with 2 to 8 outputs. (specs PDF, one source)

I did a lot, I mean a lot of reading in old threads both here and on dslreports in which you’ve posted tons of invaluable information. I’ve since made the following changes...

I just had Comcast come by this morning and give me a new Commscope 9 port MOCA 0 gain amp (guy said they don’t use PPC anymore, probably because those cost more?). I read it has a moca filter on the input, but I put an extra one there just incase (can’t hurt right?). I’ve connected all my home runs to the 6 ports that are labeled to support MOCA (passive port doesn’t and for some reason I guess port 5 doesn’t either). FWIW, the equivalent PPC 9 port amp claims to do MOCA on all ports (including the passive). The only splitter I’m using is a BAMF 2 port MOCA splitter where the X1 modem and XG2v2 AS WELL AS the first MOCA adapter exist. I’m making use of the “device port” on the MM1000, which I understand has a MOCA filter inline with it. At the location with the other MOCA adapter I’m using no splitters, but again I am making use of the “device port” to give connection to a XG1v4 DVR. Magically, all my X1 boxes and still play back off the DVR despite every X1 box claiming to have MOCA enabled, but with no actual link on any of them. This is why I’m still making use of the “device port” instead of using a MOCA splitter, because it appears the X1 boxes don’t actually require MOCA to function at all.

Anyhow, I got the MOCA link working on d band high/1400mhz with security enabled. I’m not sure if it’s doing a full gigabit or anything because I don’t even have a way to test gigabit speeds locally (no wired ethernet computers).

So everything is up and working and later today the gocoax 2.5 MOCA adapters should be here and I plan on replacing the Motorolas with those and returning the Motorolas.

Can anyone think of any other ideas to further bolster my MOCA network? Before I got the new amp today I was making use of a MOCA splitter at the old amp to ensure the two coax segments that have the MOCA adapters were having their MOCA signal passed tap to tap... with the new amp it appears I’m able to get rid of this splitter at the amp... but should I do it anyhow? In my mind it’s extra insertion loss that’s no longer warranted, so best to leave it out.

Thanks guys!

FYI: this is the old thread that has plenty of valuable advice and links to even more advice... http://www.snbforums.com/threads/add-another-moca-connection.49433/

 

[Jumpstarter]

If correctly recommending the Holland GHS-PRO-M series splitters for their MoCA compatibility, stick with them; they have models with 2 to 8 outputs. (specs PDF, one source)

You are correct in stating holland for its name it stands the test of time, but ultimately the only thing that matters is the quality and the ratings on the splitter. I still would recommend holland if available to your needs though.

 

[Samir]

One thing I have noticed about generation interoperability on powerline is that while it is supposed to work in theory, it doesn't in reality due to the various implementations by the manufacturers. I suspect the same thing is the case for moca, hence why I wouldn't bet on them working smoothly.

Sounds like you've made some great progress!

 

[krkaufman]

Commscope 9 port MOCA 0 gain amp (guy said they don’t use PPC anymore, probably because those cost more?).

Heh, was going to talk about the amplifier next. Quick thoughts were ... the PPC amp isn't "designed for MoCA," which can be OK when only dealing with MoCA 1.1 (needing only 50 MHz, from 1125-1175 MHz, for operation) but becomes problematic and then prohibitive when trying to use the entire Extended Band D frequency range, as would be required for the dual networks "independent operation" solution.

Do you have the model number for the Commscope amp, to allow research?

 

[Michael Nunes]

Heh, was going to talk about the amplifier next. Quick thoughts were ... the PPC amp isn't "designed for MoCA," which can be OK when only dealing with MoCA 1.1 (needing only 50 MHz, from 1125-1175 MHz, for operation) but becomes problematic and then prohibitive when trying to use the entire Extended Band D frequency range, as would be required for the dual networks "independent operation" solution.

Do you have the model number for the Commscope amp, to allow research?

Yes, its a Commscope CSMAPDU9VPI. The new equivalent version of the PPC 9M-U/U appears to have better specs, but the Comcast guy said they don’t use them anymore. Maybe you can also give me your opinion of the two on a spec vs. spec basis and also comment as to whether only ports 2,3,4,6,7,8 on the Commscope actively pass MOCA tap to tap whereas it appears on the PPC all ports pass MOCA.

Commscope: https://www.commscope.com/catalog/connectors/pdf/part/4322|4330|86278/CSMAPDU9VPI.pdf

PPC: https://cdn2.hubspot.net/hubfs/2057289/Downloadable_docs/Spec_sheets/Drop_Amplifiers_and_Splitters/M Entry Series Specifications_V7.2.pdf

 

[krkaufman]

I’m making use of the “device port” on the MM1000, which I understand has a MOCA filter inline with it. At the location with the other MOCA adapter I’m using no splitters, but again I am making use of the “device port” to give connection to a XG1v4 DVR. Magically, all my X1 boxes and still play back off the DVR despite every X1 box claiming to have MOCA enabled, but with no actual link on any of them. This is why I’m still making use of the “device port” instead of using a MOCA splitter, because it appears the X1 boxes don’t actually require MOCA to function at all.

This would surprise me (that the X1 devices aren't using MoCA). One possibility is that the X1 MoCA gear is leveraging MoCA's built-in power boosting to power-through the sub-optimal pass-through port connections. You'd need to be able to access the X1 device's MoCA stats to determine whether they're closer to the optimum (-30 dB) or worst case (+3 dB) TX power level.

I'd also be more comfortable if someone from Comcast could ensure that the X1 devices are configured to only operate in sub-band D-Low.

 

[Jumpstarter]

One thing I have noticed about generation interoperability on powerline is that while it is supposed to work in theory, it doesn't in reality due to the various implementations by the manufacturers. I suspect the same thing is the case for moca, hence why I wouldn't bet on them working smoothly.

Sounds like you've made some great progress!

My experience with moca and powerline has both been great. My one pause about powerline though is the average consumer sees the listed specs on the box when they buy it and think they are investing in that spec, but the average consumer doesn't realize its functionality exist more in the realm of theoretics and they don't understand all the environmental factors like age of electrical wiring.

 

[Michael Nunes]

This would surprise me (that the X1 devices aren't using MoCA). One possibility is that the X1 MoCA gear is leveraging MoCA's built-in power boosting to power-through the sub-optimal pass-through port connections. You'd need to be able to access the X1 device's MoCA stats to determine whether they're closer to the optimum (-30 dB) or worst case (+3 dB) TX power level.

I'd also be more comfortable if someone from Comcast could ensure that the X1 devices are configured to only operate in sub-band D-Low.

I’ve pulled up the service menu on every X1 device and confirm that they all have MOCA active, but they all show “noLink” and they also have no MOCA transmissions in their transmission table... which is grime because they all still function properly (playback recorded content room to room and playback all live tv). They’re also all reporting to be functioning properly to Comcast.

It’s worth mentioning that prior to putting the new amp in this morning, all the X1 devices didn’t even report an active MOCA connection, the MOCA page on the X1 service menu of each box was just blank. Not the case anymore. Now they all show MOCA active, but no link. Obviously the no link is more then likely due to the DVR and XG2 both being connected to the “device port” on the relevant MOCA adapters.

Also, just to be clear... everything is working now, X1 boxes and my MOCA network link. At this point I’m just trying to bolster my MOCA link to ensure I have the best possible conditions present.

Also, my att fiber/ip network is not bridges in anyway with the xfinity network so to speak. The att network only consists of ethernet cables... except for the moca link over the coax in order to bridge the two floors of my house in which I can’t easily run new cable. There’s zero concern about my att internet and xfinity internet service being in the same network.

 

[krkaufman]

You are correct in stating holland for its name it stands the test of time, but ultimately the only thing that matters is the quality and the ratings on the splitter. I still would recommend holland if available to your needs though.

Which is why the GHS-PRO-M splitters are recommended over typical 1 Ghz cable splitters and generic splitters rated to 2 GHz. The GHS-PRO-M series was supposedly designed explicitly for OTA/cable and MoCA Extended Band D operation, with the key spec being that port-to-port isolation is reduced at MoCA frequencies, reducing the loss for MoCA port-jumping.

 

[krkaufman]

I’ve pulled up the service menu on every X1 device and confirm that they all have MOCA active, but they all show “noLink” and they also have no MOCA transmissions in their transmission table... which is grime because they all still function properly (playback recorded content room to room and playback all live tv). They’re also all reporting to be functioning properly to Comcast.

Well that's not a mystery at all.

 

[Jumpstarter]

I’ve pulled up the service menu on every X1 device and confirm that they all have MOCA active, but they all show “noLink” and they also have no MOCA transmissions in their transmission table... which is grime because they all still function properly (playback recorded content room to room and playback all live tv). They’re also all reporting to be functioning properly to Comcast.

this maybe the case if they are all setup to connect to the cloud instead of each other.

 

[Samir]

My experience with moca and powerline has both been great. My one pause about powerline though is the average consumer sees the listed specs on the box when they buy it and think they are investing in that spec, but the average consumer doesn't realize its functionality exist more in the realm if theoretics and they don't understand all the environmental factors like age of electrical wiring.

Yeah, I don't like that gimmick either. The same game is played with wireless routers too.

Reminded me of the boom boxes in India that had crazy things on them like '1700W!' and then you look at the actual power label and it's like 60W max. Shaaaadddeeeee....

 

[Jumpstarter]

Yeah, I don't like that gimmick either. The same game is played with wireless routers too.

Reminded me of the boom boxes in India that had crazy things on them like '1700W!' and then you look at the actual power label and it's like 60W max. Shaaaadddeeeee....

Well with wireless routers that gap is more quickly closing, but I don't see the gap closing for powerline unless they promises to replace house wiring , but even that will only help for a best case, and not their theoretical numbers.

 

[Michael Nunes]

this maybe the case if they are all setup to connect to the cloud instead of each other.

This may very well be the case, because my “companion boxes” are both essentially dvrless main X1 boxes... the XG2v2 both have 4 tuners, ethernet ports, the whole nine. They’re the primary boxes deployed as a standalone solution when the customer doesn’t want/need a dvr. They’re not your typical “dumb” secondary room remote box that /has/ to communicate with the main box to do anything.

 

[Michael Nunes]

Also, I see everyone in here recommends the Holland splitters and I’m just curious what everyone’s thoughts are on the BAMF splitters available on amazon? The description kinda convinced me that they may be more purpose built, but now that I’ve done some more reading the past few days it just sounds like a bunch of bs. I’m just gonna go ahead and order some Holland’s and return the BAMF, but would like to hear your guys thoughts.

BAMF 2-Way Coax Cable Splitter Bi-Directional MoCA 5-2300MHz https://www.amazon.com/dp/B0113JAN8K/?tag=snbforums-20

 

[krkaufman]

Yes, its a Commscope CSMAPDU9VPI. The new equivalent version of the PPC 9M-U/U appears to have better specs, but the Comcast guy said they don’t use them anymore. Maybe you can also give me your opinion of the two on a spec vs. spec basis and also comment as to whether only ports 2,3,4,6,7,8 on the Commscope actively pass MOCA tap to tap whereas it appears on the PPC all ports pass MOCA.

Commscope: https://www.commscope.com/catalog/connectors/pdf/part/4322|4330|86278/CSMAPDU9VPI.pdf

PPC: https://cdn2.hubspot.net/hubfs/2057289/Downloadable_docs/Spec_sheets/Drop_Amplifiers_and_Splitters/M Entry Series Specifications_V7.2.pdf

The Commscope's specs seem to indicate that MoCA should be able to communicate between all of the output ports, including the passive VOIP port -- and appear effectively the same as the PPC.

Operating Frequency Band 1125 – 1675 MHz
Band Rejection, output to input, minimum 40.00 dB
Band Rejection, VOIP to input, minimum 35.00 dB
Insertion Loss, output to output, maximum 35.00 dB
Insertion Loss, output to VOIP, maximum 35.00 dB
Insertion Loss, VOIP to output, maximum 35.00 dB

What I don't know is how a power inserter, potentially part of a coax connection to the "Power In" Output port might affect MoCA communication. One would hope that a MoCA amp would only be shipped with a power inserter spec'd for the full cable+MoCA frequency range.

 

[krkaufman]

Also, I see everyone in here recommends the Holland splitters and I’m just curious what everyone’s thoughts are on the BAMF splitters available on amazon? The description kinda convinced me that they may be more purpose built, but now that I’ve done some more reading the past few days it just sounds like a bunch of bs. I’m just gonna go ahead and order some Holland’s and return the BAMF, but would like to hear your guys thoughts.

BAMF 2-Way Coax Cable Splitter Bi-Directional MoCA 5-2300MHz https://www.amazon.com/dp/B0113JAN8K/?tag=snbforums-20

My vote is "replace them."