How sensitive is MoCA to cable run length, couplers, and splitters?

[sch9171]

In the past when I've used coax for actual cable TV and DOCSIS internet, I have gone to significant effort to minimize the length of cable runs, eliminate as many splitters and connectors as possible, re-terminate old cables with better connectors, etc. I'm wondering if this fanatical optimization is really necessary for MoCA or not.

How much (if at all) does MoCA performance suffer from a 3-way vs. 2-way splitter, or another ~15 feet of coax in the line? As one concrete example, I have a spot where I could probably trim anywhere from 5-20 feet of cable out of the line by re-routing things. But it would be a decent amount of work, so I'd only want to do it if it actually helps. And obviously there's no way for me to test it without doing the work... In a similar (but I guess easier to test) area, I have another spot where it would be convenient to have the cable be longer, but that would involve adding a coupler and connecting another piece.

Even in places where I can theoretically test it, I'm worried that there might be reliability issues that don't surface immediately. So I'd love to hear the wisdom of the crowd here about whether I'm over-thinking it, or if I should put real effort into optimizing these things.

 

[abailey]

I use MOCA extensively in my house. I have found it to be very durable and reliable. Most of my cables are old RG59. My max cable length is probably 120 feet or so. I see no difference (in speed) between that and shorter cable lengths. Most of my cables use straight barrel connections, but one run uses a two way splitter. Once again I have no problems. My splitter is MOCA rated, however, as I did have trouble with an older non-MOCA rated splitter. As far as the MOCA not working at all, it would have to be pretty bad. Most of the time the MOCA just runs slower if the conditions are not the best.
That's my experience, your mileage may vary.

 

[degrub]

MOCA modems have built in amplifiers to help maintain adequate signal level. Best way to test is get a pair of Actiontec 6200’s and check the path on the diagnostic page. That will give you an idea of how much power budget is being used and whether improving the cable is worthwhile.
If you are using a “star” point to point isolated layout for each modem pair, i would guess no issues. If you are using a multi-drop layout, then maybe, maybe not. Each connection is 3dB from the budget and splitters can have more on each port unless low loss type or with built in amp.

cable length is probably less important than number of connections, splitters, and damaged cable or cable terminations.

any unused splitter ports should be terminated with 75 ohm caps to avoid creating reflection noise.

any splitters or amps need to be moca2 rated. Search the threads here for specific recommended brands.

is your cable RG6 ? I assume so, since you mentioned Docsis in your post.

 

[sch9171]

is your cable RG6 ? I assume so, since you mentioned Docsis in your post.

This is a new house (well, actually a very old house, but new to me), so I have never had DOCSIS service here, but all of the cable I am using is RG6.

Thanks for the info. I was just trying to get to the diagnostic page, and it appears that both of my adapters are using the same IP address (192.168.144.30), which is causing some weirdness, but I will mess with it more tomorrow...

 

[System Error Message]

without reading the answer is very very sensitive. Its copper or a cable which uses a metal conductor which is susceptible to emf and other interference.

So how sensitive depends on the cable use, the insulation, grounding and isolations, the frequency and resistance of the cable.

The terminations and other connections do matter too including the adapters used as well.

If its too much, consider powerline with the best and newest standard.

Running coax through a wall actually adds a layer of shielding to it, as long as it is run at 90 degrees to any power cables and emf sources and not in parallel.