MoCA 2.0 Actiontech ECB6000

[pec]

I purchased a pair of Actiontec ECB6000's, which are MoCA 2.0 (100 MHz single channel - "Baseline Mode"), and installed them yesterday. Consistently getting >325 Mbps (max I have seen is 350 Mbps). The ECB6000's replaced a pair of Actiontec ECB2500C's that got 95 Mbps. This is on a dedicated run of coax from first floor to second floor. While Actiontec cites the 670 Mbps PHY rate, the MoCA Alliance site says you can expect 400+ Mbps MAC throughput from MoCA 2.0 Baseline Mode.

I have not seen the ECB6000's advertised, but you can purchase them from Actiontec's online store.

-pc

 

[kjacques1]

Fantastic news. I've been waiting for Moca 2.0 adapters to finally come to the consumer.

 

[bds3151]

I just got a pair of these and on a shared cable tv line with 2 splitters I'm getting about 50~70 mbps, the same as moca 1.1! I think they need a dedicated cable to perform at 2.0 rates. I don't have that so I'm going to try amplifiers next.

 

[pec]

Dedicated cable should not be required, but I have not tested this since I had a dedicated piece of coax going from the first to second floors. The MoCA field trial preliminary results from earlier this year were >350 Mbps in 95% of their test houses. I am typically seeing on the average ~330 Mbps.

You do need splitters with a 2 GHz frequency response (the 100 MHz channel is between 1.125 and 1.425 GHz).

-pc

 

[bds3151]

Dedicated cable should not be required, but I have not tested this since I had a dedicated piece of coax going from the first to second floors. The MoCA field trial preliminary results from earlier this year were >350 Mbps in 95% of their test houses. I am typically seeing on the average ~330 Mbps.

You do need splitters with a 2 GHz frequency response (the 100 MHz channel is between 1.125 and 1.425 GHz).

-pc

I do have the appropriate splitters. Thanks for the response.

 

[bds3151]

By the way, is the 325mbps each way or combined as in the spec. for 1.1 - 170mbps divided by two or 85mbps each way?

 

[pec]

My 325-350 Mbps result was from Speedtest.net. The 400+ Mbps cited in the MoCA Alliance is total throughput, so I assume you are correct that a full-duplex stream over MoCA 2.0 Basic would be 200 Mbps.

 

[stevech]

I purchased a pair of Actiontec ECB6000's, which are MoCA 2.0 (100 MHz single channel - "Baseline Mode"), and installed them yesterday. Consistently getting >325 Mbps (max I have seen is 350 Mbps).
-pc

Is that 325Mbps net yield at the IP layer, after all overhead? E.g., run speedtest.net across MoCa.

 

[pec]

Yes, 325-350 Mbs is the downlink speedtest.net results from my ISP (1 Gbps service from ISP). All versions of MoCA use a reservation system for scheduling transmissions, and for MoCA 2.0 Basic, the MoCA Alliance estimates 400+ Mbps of total MAC throughput. There are TCP Acks coming back from the 4-streams of HTTP traffic used by Speedtest.net and some small amount of traffic from the other devices connected that would go into the total throughput calculation.
-pc

 

[azazel1024]

I just got a pair of these and on a shared cable tv line with 2 splitters I'm getting about 50~70 mbps, the same as moca 1.1! I think they need a dedicated cable to perform at 2.0 rates. I don't have that so I'm going to try amplifiers next.

Do you have ANY other MoCA devices on that shared Coax? Like a DVR box or something? Because by MoCA standards, ALL MoCA adapters will default to the lowest currently implemented MoCA standard on the wire.

So if you have a MoCA 1.0 or 1.1 device on there, they'll bump down to that, even if they are capable of 2.0. Also if the condition of the Coax, splitters and what not are poor, you won't get faster performance. If you are only getting 50-70Mbps, I'd suspect a bad setup. Just like Wifi or Powerline, you'll only see higher performance with "faster" standards if the medium can support it. If it is noisy, poor signal attenuation, etc., you won't get things faster or only marginally.

 

[bds3151]

Do you have ANY other MoCA devices on that shared Coax? Like a DVR box or something? Because by MoCA standards, ALL MoCA adapters will default to the lowest currently implemented MoCA standard on the wire.

So if you have a MoCA 1.0 or 1.1 device on there, they'll bump down to that, even if they are capable of 2.0. Also if the condition of the Coax, splitters and what not are poor, you won't get faster performance. If you are only getting 50-70Mbps, I'd suspect a bad setup. Just like Wifi or Powerline, you'll only see higher performance with "faster" standards if the medium can support it. If it is noisy, poor signal attenuation, etc., you won't get things faster or only marginally.

Do you have ANY other MoCA devices on that shared Coax? Like a DVR box or something? Because by MoCA standards, ALL MoCA adapters will default to the lowest currently implemented MoCA standard on the wire.

So if you have a MoCA 1.0 or 1.1 device on there, they'll bump down to that, even if they are capable of 2.0. Also if the condition of the Coax, splitters and what not are poor, you won't get faster performance. If you are only getting 50-70Mbps, I'd suspect a bad setup. Just like Wifi or Powerline, you'll only see higher performance with "faster" standards if the medium can support it. If it is noisy, poor signal attenuation, etc., you won't get things faster or only marginally.

Yeah, I thought about that also. I removed the only 1.1 device and no change. FIOS tv works great over the same cable so I'm thinking its attenuation since I have about -14db because of the splitters (they are MOCA 2.0 compatible). I can't find a MOCA 2.0 amp in stock anywhere to order that would compensate for the signal loss. Too new I guess.

 

[bds3151]

I did some further testing back to back with a 3' coax against a Synology NAS that can do about 150mbps read on a direct ethernet cable and got about 120mbps. So it seems that my coax infrastructure is approximately halving the data rate. I don't know where to go from here as I need to share the coax via splitters. Seems like my situation represents the more typical environment so the actiontec devices will only work at 2.0 speeds if you dedicate coax or, when amplifiers are generally available (untested theory). So for my situation = fail.

 

[azazel1024]

Limit the splitters. 14dB means a LOT of splits.

What you can do is this, where you need MoCA connection, us ONLY a dual, 3.5dB splitter. Then connect that to the rest of the infrastructure.

That way on the leg that is MoCA to MoCA, you'd only have 3.5dB of attenuation from the splitters. Also remove any and all connections that aren't required. Do you really have 4 different devices connected (maybe you do)? If not, you can move to a 3 way splitter and always put the 4-way back in later if you need it. Shorten any and all Coax that you can. If you are using RG-58 instead of RG-6, replace as much of the cable as you can. Cap any and all unused coax jacks. This will all improve things.

Also...how old is that Synology if it can only do 150Mbps? That is pretty slow.

 

[bds3151]

Limit the splitters. 14dB means a LOT of splits.

What you can do is this, where you need MoCA connection, us ONLY a dual, 3.5dB splitter. Then connect that to the rest of the infrastructure.

That way on the leg that is MoCA to MoCA, you'd only have 3.5dB of attenuation from the splitters. Also remove any and all connections that aren't required. Do you really have 4 different devices connected (maybe you do)? If not, you can move to a 3 way splitter and always put the 4-way back in later if you need it. Shorten any and all Coax that you can. If you are using RG-58 instead of RG-6, replace as much of the cable as you can. Cap any and all unused coax jacks. This will all improve things.

Also...how old is that Synology if it can only do 150Mbps? That is pretty slow.

I have 2 -4 way's with 7db loss each. Anyway, I tested the loss theory by totally removing half of the splitters and there was no appreciable speed change. The only improvement I've seen is to use a very short run of rg6 (3') with the devices back to back. All of the endpoints are capped. Re: Synology - yeah it's probably 5 or more years old. I also totally isolated the cable from FIOS TV so see if that was conflicting - also no change in speed. Next step is to add a moca blocker/reflector at POE to see if that helps any. I don't think an amplifier would do anything based upon the results removing one 4 way splitter. Seems to be more of a noise issue maybe. I might try to re terminate some of the coax with better compression fittings.

 

[azazel1024]

Yeah, I would suspect you are either getting noise introduced somewhere or some of the coax run(runs) have bad cable.

 

[pec]

Are you using the so-called MoCA splitters (e.g., Holland Electronics), which claim to have reduced isolation between the output ports in the MoCA D-band? The MoCA reflector is also recommended in the installation guides to go at the POE, or in your test case you could just put it at the input to the splitter that has the MoCA segments.

It looks to me like the existing MoCA amps should work fine with MoCA 2.0 since the ECB6000 uses the same part of the D-band as the earlier devices. Since you reduced the path loss in your test, however, you are probably right that the problem is not attenuation in the forward path.

Since you got exactly the same throughput with MoCA 1 and 2 devices, I also wondered if there was a stray MoCA 1 device that you had not removed from the FIOS setup, but it looks like you have already done this.

-pc

 

[stevech]

likely not noise.
Weak signal
Splitter that isn't rated for 2GHz.

 

[azazel1024]

Yeah, I'd rule out the splitters and the wire by using a female to female union and connect up the two coax cables that you'd be running between the two bridges. None of this using a 3ft section of cable. Use the actual wiring you'll use later, but with no splitter in place. If there is still a performance issue, the issue is in the wiring or the termination at the box (open the wall boxes and check for a bad connection between the coax and the keystone jack terminals!!!)

 

[stevech]

female to female coupler a.k.a. barrel connector.
TV coax connectors in general are "F" connectors.

F connectors are famous for unreliable long term - corrosion, shield not contacting connector, center conductor too short, etc. CableCo techs use super high quality F connectors and proper tools for installing them on coax. RG6 coax, not RG59 (crappy).

 

[Thucydides]

pec, how is latency using the ECB6000s ? I have the Moca 1.1 ECB2500Cs at home, and they seemed to add 3-4 ms (from memory here) to ping times across the network.