Intermittent MOCA Problems

[nrgiii2]

I am having problems with ECB6250 MOCA 2.5 adapters where the connection to the Mac Mini (see diagram) drops once or twice a day.
There are coax 2 splitters and both are MOCA capable.
The POE filter is outside the house inside the Comcast ONT.
Resetting the office MOCA adapter resolves the problem and all is well until the next drop.
All unused coax ports in the house have 75 ohm terminators (despite the old photo showing 1 empty jack).
There are no TV signals going over the coax, just Comcast DOCSIS 3.0 and MOCA.
I've attached screenshots from the office MOCA adapter (ECB6250 #1) .

Any idea what I can change to address the problem?

Thanks in advance!

Norm

 

[krkaufman]

There are coax 2 splitters and both are MOCA capable.

Except they’re not; the 6-way isn’t a model I’d consider designed for MoCA. Based on the attached diagram, I’d consider shifting to a 2-way plus 4-way splitter configuration, instead (ex: Antronix MMC1002H-B + MMC1004H-B), with a 70+ dB “PoE” MoCA filter installed directly on the 2-way’s input port to maximize the reflective benefit. The top-level 2-way would connect to the modem location and the secondary 4-way splitter.

The Picolink splitter should be fine, given it shipped with your adapters, but it would be worthwhile ordering a replacement for it, as well, as a precaution. (Order from somewhere with a good return policy.)

Separately, you could take a pair of MoCA adapters and connect them at each end of a given in-wall run to check the MoCA stats (get the PHY rates) for each coax line, to see if any of the runs demonstrate inferior quality. If so, open the wallplates and check the termination quality for the cabling; you can test with the wall outlets bypassed, as well, to rule out a coax outlet as the issue source.

p.s. The PHY rates table shown does seem to indicate an issue with MoCA adapter #1, with the rates in the vertical column associated with adapter 1 all showing well below the preferable 3500-3600 range. Pretty odd for the diagonal “GCD” rates to be so low, as well; these are typically in the 600-700 range.

 

[degrub]

at least replace the 6 way with a powered "0 loss" MOCA 2 rated splitter. That may address the #1 moca adapter path loss issue if it is not a cable/termination issue. It will also keep the signal level to the COX modem at a high level (possibly too high - you will have to check in the modem down/up link pages) . You will want to get the COX modem off of the MOCA coax path eventually as they are likely to move to DOCCIS 3.1 which will take over 2 or 3 of the MOCA 2.5 bands leaving you with the equivalent of MOCA 2.0 bandwidth to share among the nodes. So 1) isolate the COX modem from the rest of the coax if possible and 2) replace that splitter with a MOCA 2 certified splitter(s).

 

[nrgiii2]

at least replace the 6 way with a powered "0 loss" MOCA 2 rated splitter. That may address the #1 moca adapter path loss issue if it is not a cable/termination issue. It will also keep the signal level to the COX modem at a high level (possibly too high - you will have to check in the modem down/up link pages) . You will want to get the COX modem off of the MOCA coax path eventually as they are likely to move to DOCCIS 3.1 which will take over 2 or 3 of the MOCA 2.5 bands leaving you with the equivalent of MOCA 2.0 bandwidth to share among the nodes. So 1) isolate the COX modem from the rest of the coax if possible and 2) replace that splitter with a MOCA 2 certified splitter(s).

Thanks for the reply.
I'm pretty sure Comcast is already using DOCSIS 3.1 in my area as they offer GB internet. I intentionally did NOT buy a 3.1 modem to avoid infringement into the MOCA spectrum.
I was wondering if DOCSIS 3.1 traffic could causing my problems even though I don't have a 3.1 modem? Would the POE filter be expected to block DOCSIS 3.1 ? I suspect not.
I can't really separate the modem and MOCA with the way the house is wired. I will be running ethernet to my wife's office if I can't resolve the MOCA issues.

 

[nrgiii2]

Except they’re not; the 6-way isn’t a model I’d consider designed for MoCA.

Well that's bad then. I was going off the frequency range on the labels which seemed to cover MOCA. I will install new splitters as you suggest and report back. I really appreciate your help!!
Just checked my POE and it's rated -40dB. I take it that's not enough and I should have -70dB ? This is the POE is have now: Link

 

[OzarkEdge]

OE

 

[nrgiii2]

So that means MOCA and DOCSIS 3.1 cannot share the same cable right? Stringing CAT6 cable could be in my future....

 

[krkaufman]

More on DOCSIS encroachment on the MoCA frequency range, here. As described, isolation of DOCSIS signals from the MoCA-infused coax is a longer-term requirement; short-term, installing an additional MoCA filter directly on a MoCA-sensitive DOCSIS 3.1 modem is usually sufficient to address the most common hiccup. Absent a spare MoCA filter to protect the modem, a quick fix can be temporarily shifting the MoCA network to operate only in the D-High frequency range, restoring the MoCA setup to the full Extended Band D once a "prophylactic" MoCA filter is in place at the modem.

​

Would the POE filter be expected to block DOCSIS 3.1 ? I suspect not.

A "PoE" MoCA filter *would* attenuate DOCSIS signals that fall within the filter's stop-band, 1125-1675 MHz, so placement of the "PoE" MoCA filter will need to be adjusted long-term, as the coax is updated to support DOCSIS 3.1 initial or full rollout frequencies above 1002 MHz. Short-term (like now), DOCSIS *signals* are not likely your issue.

I'd just recommend updating the 6-way splitter to a MoCA-optimized configuration as described above, and reevaluate. (Along with optionally performing the MoCA pair tests for each in-wall cable, and visually inspect the terminations on each in-wall cable.)

 

[krkaufman]

So that means MOCA and DOCSIS 3.1 cannot share the same cable right?

Longer-term, no, but the DOCSIS 3.1+ signals will only need to reach the cable modem -- so isolation of the ISP/modem signal may be achievable w/ a bit of creativity.

Short-term, there aren't many providers actually using D3.1 frequencies above 1002 MHz, yet.

 

[nrgiii2]

Longer-term, no, but the DOCSIS 3.1+ signals will only need to reach the cable modem -- so isolation of the ISP/modem signal may be achievable w/ a bit of creativity.

Short-term, there aren't many providers actually using D3.1 frequencies above 1002 MHz, yet.

Thanks again!
Is it worth spending the money to upgrade the POE from -40db to -70dB ?

 

[krkaufman]

Is it worth spending the money to upgrade the POE from -40db to -70dB ?

Given that you’re having MoCA issues, I’d recommend it, as a means to ensure no interference from outside signal sources. Trivial cost relative to everything else, right?

• preferred MoCA filter: PPC GLP-1G70CWWS(Amazon US listing) … 70+ dB stop-band attenuation, spec’d for full MoCA Ext. Band D range, 1125-1675 MHz

edit: Oh, also, I meant to say that you may be able to get a 70+ dB MoCA filter from Comcast, free, if you request the added protection. I can point to some testing showing the deficiency of mere 40+ dB filters, if needed, if you look to leverage the “security concern” angle.

 

[krkaufman]

There are no TV signals going over the coax, just Comcast DOCSIS 3.0 and MOCA.

Sidebar (unrelated to your issues): In this digital age, there *are* cable TV signals flowing over your coax, amidst your DOCSIS signals; you just don’t have the authorization or gear to access them.

 

[nrgiii2]

Given that you’re having MoCA issues, I’d recommend it, as a means to ensure no interference from outside signal sources. Trivial cost relative to everything else, right?

• preferred MoCA filter: PPC GLP-1G70CWWS(Amazon US listing) … 70+ dB stop-band attenuation, spec’d for full MoCA Ext. Band D range, 1125-1675 MHz

edit: Oh, also, I meant to say that you may be able to get a 70+ dB MoCA filter from Comcast, free, if you request the added protection. I can point to some testing showing the deficiency of mere 40+ dB filters, if needed, if you look to leverage the “security concern” angle.

Will do. For $7 it's a no-brainer.

 

[nrgiii2]

Sidebar (unrelated to your issues): In this digital age, there *are* cable TV signals flowing over your coax, amidst your DOCSIS signals; you just don’t have the authorization or gear to access them.

Got it. I meant to say my attic TV antenna is on completely separate coax from the MOCA.

 

[OzarkEdge]

Thanks for the discussion... I'm learning without trying too hard. I was able to combine my broadcast TV antenna signals with MoCA2.5... works great... DOCSIS3.1 is on a separate/dedicated coax run, so no conflict... didn't think about it having cable TV signals on it, too... maybe my neighbor(s)' MoCA, too, since I did not filter the ISP cable.

OE

 

[krkaufman]

Got it. I meant to say my attic TV antenna is on completely separate coax from the MOCA.

Does the “OTA” coax run everywhere that you have MoCA adapters installed on the “cable” coax? (Basically two parallel, separate coax trees?)

‘gist: Just having the OTA coax available at your router location and one other room shared with a cable coax run would afford you the ISP/modem path isolation from MoCA that you’ll eventually require to address DOCSIS encroachment. (Skipping details since that’s a longer term issue; just thought I’d throw it out to mitigate concerns that may have sprouted since the thread’s inception.)

p.s. MoCA can also share coax with OTA signals, so if you have any OTA-only locations begging for a wired connection…

Of course, if you have OTA coax at all your MoCA locations, with some requisite prep work you could just migrate the MoCA network to the OTA coax, instead, and your ISP/modem feed could just be a direct connection, sans splitters or filters.

 

[nrgiii2]

Does the “OTA” coax run everywhere that you have MoCA adapters installed on the “cable” coax? (Basically two parallel, separate coax trees?)

‘gist: Just having the OTA coax available at your router location and one other room shared with a cable coax run would afford you the ISP/modem path isolation from MoCA that you’ll eventually require to address DOCSIS encroachment. (Skipping details since that’s a longer term issue; just thought I’d throw it out to mitigate concerns that may have sprouted since the thread’s inception.)

p.s. MoCA can also share coax with OTA signals, so if you have any OTA-only locations begging for a wired connection….

The OTA coax is dedicated and on it's own path. It only comes anywhere near the MOCA coax in the last few 10 feet or so. It's never "side by side".

 

[OzarkEdge]

p.s. MoCA can also share coax with OTA signals, so if you have any OTA-only locations begging for a wired connection….

Indeed, this is slick... I can now get OTA and 2.5GbE at any coax wallplate using the Holland diplexer and a MoCA2.5 slave adapter.

OE

 

[krkaufman]

The OTA coax is dedicated and on its own path. It only comes anywhere near the MOCA coax in the last few 10 feet or so. It's never "side by side".

All that really matters is that OTA coax outlets are present wherever you require MoCA connectivity; then it’s just a matter of whether the OTA coax plant is MoCA-hospitable (or, ideally, MoCA-optimized).

Not to worry for now; just something to keep in mind for down the road, when ISP/modem isolation from MoCA signals becomes a requirement.

 

[krkaufman]

I can now get OTA and 2.5GbE at any coax wallplate using the Holland diplexor and a MoCA2.5 slave adapter.

Good place to drop the ol’ MoCA endpoint connection alternatives graphic…