Having trouble with my Bonded MoCA 2.0 setup

[Sgtkeebler]

Hello SNB,

First time poster long time lurker. I just recently purchased the Actiontec MoCA 2.0 adapters from my local Best Buy. I went to set it up and was using the original splitter that came with the adapters for my room, and our living room. The download speeds that I was getting with the splitters that came with the adapters were around 200-300 mbps. Now there was a problem with the way this was setup. Since the splitters that came with the adapters were only two-way splitters we couldnt hook up our main tv/DVR box which caused our other tv/dvr's to stop working. To fix this I had to buy a 3 way splitter, and the 3 way splitter which is made by the same brand of splitter that came with the MoCA adapters StarDust, the new 3 way splitter has the output of 3.5db - 7.5db - 7.5db whereas the splitter that came with the MoCA 2.0 adapters were 3.5db-3.5db.

I hooked up my router to the 3.5db, the tv and MoCA adapter's to the 7.5db connections. This has cut my download speeds to 90-100 mbps. Does 4db really make that big of a difference? Why is there such a drastic cut in download speeds? I just purchased a new MoCA rated 3 way spitter where the output is 5.5db straight across in hopes to improve this. Do any of you all have any advice on how I can increase my download speeds? Would it be worth having Comcast come out to determine if we need an amplifier or anything like that?

My second question is that I understand I need a MoCA 2.0 poe filter. I took a picture of our cable connection coming from the street because I'm not sure if it already has a poe filter on it.

 

[Internet Man]

I just purchased a new MoCA rated 3 way spitter where the output is 5.5db straight across in hopes to improve this. Do any of you all have any advice on how I can increase my download speeds? Would it be worth having Comcast come out to determine if we need an amplifier or anything like that?

It is important that the splitter is MoCA rated otherwise it might not allow MoCA signals to pass through.

My second question is that I understand I need a MoCA 2.0 poe filter. I took a picture of our cable connection coming from the street because I'm not sure if it already has a poe filter on it.

Your picture does not show a POE filter, just a grounded connector. Here's an example of a MoCA POE filter. This thing should be attached to the input of your MoCA rated splitter.

 

[krkaufman]

The download speeds that I was getting with the splitters that came with the adapters were around 200-300 mbps.

1) That’s a wide window. Were the numbers that inconsistent?

2) You haven’t stated your spec’d download /upload rates?

3) What speed do you see when connected directly to your gateway/router via Ethernet?

4) What speed do you see with the adapters direct-connected to each other via coax, with one connected via Ethernet to the router and the other to your PC/laptop?

Since the splitters that came with the adapters were only two-way splitters we couldnt hook up our main tv/DVR box which caused our other tv/dvr's to stop working. ... I hooked up my router to the 3.5db, the tv and MoCA adapter's to the 7.5db connections. This has cut my download speeds to 90-100 mbps. ... Would it be worth having Comcast come out to determine if we need an amplifier or anything like that?

My first guess would be that you may have Xfinity X1 whole home DVRs which use MoCA, and may be knocking the ECB6200 MoCA down to 1.1 rates.

One way to test the theory would be to keep everything connected as-is, power everything down for a few minutes, then power-up the gateway, then MoCA adapters ... but leave all Comcast DVRs or client boxes powered-down. Then test the speeds.

 

[krkaufman]

My second question is that I understand I need a MoCA 2.0 poe filter. I took a picture of our cable connection coming from the street because I'm not sure if it already has a poe filter on it.

Your picture does not show a POE filter, just a grounded connector. Here's an example of a MoCA POE filter. This thing should be attached to the input of your MoCA rated splitter.

What @Internet Man said. Typically, the “PoE” MoCA filter would optimally be installed on the input of the first split encountered by the incoming cable signal. You could trace the cable in either direction from that grounding bracket to see if the filter is installed elsewhere.

As a Comcast customer, you should be able to call them to have a tech come install one for free.

Would it be worth having Comcast come out to determine if we need an amplifier or anything like that?

A review of the coax plant is worthwhile, documenting how the rooms/devices connect, and through what components, noting brand/model# info. (diagrammed even better; example attached)

As for amplifying MoCA signals, MoCA has its own amplification built-in; regardless, cable/OTA amps don’t actually amplify signals at the MoCA frequencies (1125-1675 MHz) — more often than not amps are an impediment to MoCA communication.. At best, if an amp is required for cable/OTA signals (5-1002 MHz), you’d want to use an amplifier “designed for MoCA,” with a built-in MoCA filter, ensuring MoCA signals can pass between output ports.

.

 

[krkaufman]

My first guess would be that you may have Xfinity X1 whole home DVRs which use MoCA, and may be knocking the ECB6200 MoCA down to 1.1 rates.

Here’s one similar example, though the solution doesn’t seem optimal:

• forums.xfinity.com: Can MoCA work with Xfinity X1 Setup?

One thing to research/try would be taking advantage of MoCA 2.0’s ability to support two MoCA networks within MoCA’s Extended band D, by using MoCA privacy and a custom channel with the ECB6200 adapters. This appears to have succeeded for at least one Comcast customer, per >this post<. (followup w/ connection details, thread link)

—-
See also:

• http://www.dslreports.com/forum/r32173674-Equip-Xfinity-X1-moca

 

[krkaufman]

OP aside, has anyone else been able to get a MoCA 2.0 network (bonded or not) working in parallel to an Xfinity X1 whole home setup ... without physically isolating the coax?

 

[krkaufman]

#Buehler

 

[bmn1]

200-300 sounds about right considering you were using a splitter. You will NEVER see higher than ~375 Mbps DOWNLOAD with these. Only way you will see anything higher than that is to "cheat" by using "parallel" connections when using synthetic benchmarks like iPerf. Honestly, I think this is some kind of firmware bug that hasn't been addressed yet because I had no problem getting ~950 Mbps Upload, but could never get more than ~365 Mbps Download.

https://www.snbforums.com/threads/a...ts-download-speed-is-lower-than-upload.54687/

 

[Samir]

I wouldn't call iperf's parallel connections cheating--it's just simply how things work these days.

Even when you do a speedtest on your internet it's using parallel streams. The dslreports.com/speedtest will even tell you how many.

 

[bmn1]

I wouldn't call iperf's parallel connections cheating--it's just simply how things work these days.

Even when you do a speedtest on your internet it's using parallel streams. The dslreports.com/speedtest will even tell you how many.

It's "cheating" because it generates inflated numbers that the majority of users will never see in real world use! Most servers out there that serve up large files, think large video files, ISO files, etc. typically allow only ONE SINGLE CONNECTION to the server at a time! So unless you have client software like a Download Manager than can force multiple connections at a time (and that's assuming that the server will even allow it), then it doesn't matter what kind of test results you get from using iPerf with Parallel Streams because you will NEVER get that in real world use! I have setup and configured MANY Apache and nginx servers, I'm well aware of how they work. Why do you think download managers claim to "BOOST" your download speed? Because they do so by using multiple connections a.k.a. "Parallel Streams".

So once again, people should NOT be trying to disregard the fact that these MoCA adapters CANNOT achieve full DOWNLOAD throughput WITHOUT using parallel streams! Parallel Streams is just something that people are doing to get inflated numbers out of these adapters jsut to make themselves feel better. You could go do an iPerf test with Parallel Streams and get ~980mbps then turn around and send a 4GB file to another computer over your Windows Network and you'll be lucky to get HALF THAT! Why? Because it's only using a SINGLE TCP CONNECTION!

The real question is WHY am I able to get ~950mbps UPLOAD throughput WITHOUT using parallel streams, yet I can only get ~365mbps DOWNLOAD? Because there is an obvious harmware/firmware limitation with these specific adapters!

 

[Samir]

So if you know all this, then what you are testing is 'single stream bandwidth', not total bandwidth, which almost anywhere is now multi-streamed.

Believe me I didn't like it one bit when things first started rolling in this direction, but now everything else is multi-streamed too so it doesn't matter as much. Even network transfers are multi-streamed, so yes the full bandwidth will be used.

Test it out for yourself and run some on dslreports.com/speedtest if you have a gigabit Internet connection. If you're not hitting 900Mbs+ like the upload then you have a problem.

 

[bmn1]

Explain then WHY I can get ~950 mbps Upload with iPerf, WITHOUT USING PARALLEL STREAMS, but I only get ~365 mbps going the other direction? (see MY THREAD for test results) The two boxes are communicating to each other over the same damn wire for EACH TEST! So how in the hell does that make sense other than this is an obvious FW issue? Are you really that simple to not be able to wrap your head around that simple FACT? I hate to be rude but DAMN! SMH