Ethernet from your TV Outlet: NETGEAR MoCA Coax-Ethernet Adapter Kit Reviewed

[picman1]

Thank you for the explanation but looking at Netgears hookup of their MCAB1001 extenders from their website the coax comes from the modem which comes from the street. The ethernet starts from the MCAB1001 and goes through the router. Isn't the ethernet speed at 100mbps from the extender and to the router and thus throughout the network or am I misunderstanding what you are saying? Thanks again.

 

[thiggins]

As I said, any traffic flowing through the MoCA bridges will be limited to 100 Mbps, which is the link rate of the 10/100 Etherent connection. But traffic between Gigabit Ethernet devices will not be limited by the speed of the MoCA devices.

 

[picman1]

Thank you I got it now. I just got a couple of Netgear extenders and they worked like most everyone else on the thread, right out of the box no config needed. Streaming 1080P from my PC server and the PS3 through 2 floors no problem.

One thing I noticed and I didn't see much on the subject is streaming a movie from the internet is not as good. It is watchable, no pixilation but some minor stutters. Are there any adjustments to improve internet streaming or these extenders limited for intenet streaming. Thank you again.

 

[thiggins]

The Internet streaming problem has nothing to do with the extenders. The problems can be due to any number of things related to the path from the Internet-based server to your router. But nothing to do with the MoCA connection.

 

[xdarque]

Configuring QoS or priority

Anyone know how to configure the QoS or priority of traffic on these adapters? Currently, whenever I copy a large file over my LAN using the MoCA connection, the adapter that listens for WAN traffic completely ignores any and all WAN packets until the large LAN file transfer is done.

For example, if one of my other machines starts a bandwidth test on speakeasy.net, the test will just pause midstream (as soon as the large LAN file transfer starts) and will only resume once the LAN file transfer is done.

I already changed the config from "All Pass" to "D Band" in order for my cable modem (which is connected to one of the MoCAs) to even work...

Thanks in advance,
-Dan

 

[eaadams]

Up and running now time to split

Ok so we got our Netgear MoCA up and running on one of our two Coax networks.

It works! However, it kills the image quality on our farthest TV (via a distribution box, a 13 year old version of THIS) and it also kills the IR repeaters over COAX (via the same distribution box) for the TV's to which the MOcA is attached.

Luckily we have a 2nd COAX network that we once used for Satellite TV. So I will adjust this for MoCA.


What sort of Coax splitter can I use for MoCA?

Can I use this: http://www.homedepot.com/webapp/wcs/stores/servlet/ProductDisplay?storeId=10051&langId=-1&catalogId=10&productId=100479584&N=10000003+90401+500106+4294926876

or this: http://www.homedepot.com/webapp/wcs/stores/servlet/ProductDisplay?storeId=10051&langId=-1&catalogId=10&productId=100579552&N=10000003+90401+500106+4294926876

or this: http://www.homedepot.com/webapp/wcs/stores/servlet/ProductDisplay?storeId=10051&langId=-1&catalogId=10&productId=100396625&N=10000003+90401+501636


[EDIT] I have to go:

3 MoCA adapters via 1 Coax Spliter

 

[eaadams]

I read this about moca spliters: "Do not use any splitter with a "power pass" output on any ports as this will cause pixilation of your broadcast video and unreliable MoCA connections"

What does that mean?

 

[eaadams]

I thought I would post my experience.

We have a netgear switch that goes about 70' to a Home Depot Coax 5-2.1ghz splitter. From the splitter it goes another 80'+ underground to a detached garage.

It works perfectly. No problems and so easy to setup. I am so pleased.

I'm eagerly awaiting the moca router that has wifi and I'm going to use moca to do out upstairs.

This works so much better than our old Airport with daisychained repeaters. Quite amazing.

 

[phdeane]

Need Help With Throughput Test

Hello - hoping someone can help me.

I'm trying to test the throughput of my Netgear MoCA adapters. I was seeing relatively slow file transfer speeds when I was using the adapters in my real-world cable system. I say relatively as it was way less than the advertised speed, which we know to be too high. I was also having freezing issues using Netflix streaming (PlayOn software to PS3) across the adapters.

So I thought I would do a rudimentary test. I removed the two adapters from my real-world environment and put them on a test bench. I connected them via the supplied 3' RG6 coax cable by connecting the Coax IN on both adapters. I connected the Ethernet port on one adapter to my laptop (XP Pro). I connected the Ethernet port on the other adapter to my D-Link DIR-655 wireless n router (gigabit), which is connected to my cable modem. I have a home computer (Vista Home Premium) connected to that same router. The MoCA adapters were left in the default setup configuration (although I did try all other combinations, too).

To be clear, there is no cable TV signal going through the system and no other computers involved in the test. I also have all firewalls on the computers turned off, as well as anti-virus programs.

To run my test, I simply copied over a folder containing numerous large video files from my laptop to a hard drive on my home computer. The total size of the files was 2,868,251,724 bytes and it took 7 minutes, 27 seconds to copy over the files. I ran the test numerous times and got the same results. I calculate that transfer rate to be 51.33 Mbps (megabits per second).

I got the exact same transfer rate when going in the other direction (home computer to laptop), as well as in the real-world setup where my adapters are going across about 50' of RG6 coax cable.

By the way, to get a baseline, I first connected the same laptop to the same home computer on the same router, and ran the exact same copy test. No MoCA adapters involved. In that case, the files transferred over in 62 seconds, which I calculate to be 370.10 Mbps. This should rule out any hard drive read/write access speed issues, router issues (unless the router conflicts with the MoCA adapters), etc.

I spent hours on the phone and email with Netgear. I even got an engineer to run the exact same test (at least he says he did, but he's all the way in India, so not sure about that), under the same test conditions, and he got just short of 175 Mbps (which, based on the review here, I don't even know how that is possible). In any event, they deemed the adapters to be defective and sent me two new ones. I got the same result. Netgear is now unable to help me.

My question is this. Is my test valid or flawed? Is a large file transfer a valid way to test throughput? Netgear said it was a valid test, but then why do others use software programs to test throughput?

I'm hoping someone more technical than me can point me in the right direction. If I didn't have the Netflix issues, I probably would not have run the tests in the first place, but now I either want to resolve this or get different MoCA adapters from another manufacturer.

Thanks in advance!

 

[corndog]

Hello there,

I would say that in your case it looks like a valid test. As tests go, it's one of many. You are only testing large files - you could also test many small files, or a random mixture, which would give you different, but still meaningful, results.

However, since your main purpose is to stream video, the large file copy would seem to make sense.

I would suggest, to make your testing more informative, run it again with a MUCH larger file. Maybe 30 or 40 gigs in size.

While the test is going on, check the graph of network throughput in task manager. Is it pretty stable? Or does it peak and then drop off in spurts?

I normally get 60 to 90Mbits/sec on my MoCA, but my graphs show that it is pretty stable around 60 to 70, with jumps up to 100. I have no drop-outs, however.

Even high definition video should be able to play without the slightest problem in a situation like this, as long as you're not dealing with uncompressed raw video (which would be crazy anyway).

Something to try anyhow...

 

[thiggins]

What speed are you expecting to see? The review showed a maximum of 70 Mbps using IxChariot (which has less overhead then Windows file copying) and TCP/IP. Higher speeds are obtained only when using multiple connections.

You could try simulating this by doing two file copies simultaneously.

 

[phdeane]

I would suggest, to make your testing more informative, run it again with a MUCH larger file. Maybe 30 or 40 gigs in size.

While the test is going on, check the graph of network throughput in task manager. Is it pretty stable? Or does it peak and then drop off in spurts?

Thank you for the quick reply. I will give the larger file size a try. I'm not real confident about it, though, as I so consistently get the 51Mbps throughput. I will also try more smaller files, but I didn't mention that the folder I'm copying does contain over 30 files in three folders. Can't hurt trying both, though.

What speed are you expecting to see? The review showed a maximum of 70 Mbps using IxChariot (which has less overhead then Windows file copying) and TCP/IP. Higher speeds are obtained only when using multiple connections.

You could try simulating this by doing two file copies simultaneously.

Thank you for your reply, too. I am expecting the 70Mbps, or at least close. 51Mbps just seems too far off. I realize that should be fine for streaming any even moderately compressed HD files, but I have seen intermittent problems that I mentioned previously.

As for the two files, I hadn't tried that, as I'm sure I will get an improved throughput. However, that doesn't represent the real-world streaming I do. It is also still worth a try to see what type of throughput I get in comparison to the review.

Thanks again!

 

[phdeane]

The review showed a maximum of 70 Mbps using IxChariot (which has less overhead then Windows file copying) and TCP/IP.

Ahh, I glanced over the above bolded statement. Do you think that would account for my slower throughput? Could copying files using Windows account for such a significant difference? I don't have a program to test, but if there is a freeware or free-trial program that would work like IxChariot, I'm open to suggestions. My only concern is also that the Netgear engineer said he duplicated my exact test (which, of course, is suspect, as how can he have an "exact" duplicate test environment?) and got much higher throughput.

 

[thiggins]

There is a lot of overhead in Windows file copying with XP. Vista SP1 improved file copy speed significantly (after screwing it up in the initial release). But you need two Vista SP1 (or higher) systems to achieve the full benefit.

You can try FTP instead of file copy for lower overhead. iperf/jperf is probably the closest you can come to IxChariot for low-overhead throughput testing. But the settings you use are important because the defaults don't necessarily provide the highest throughput. See Measuring Network Performance - Iperf and Measuring Network Performance - Jperf.

 

[phdeane]

There is a lot of overhead in Windows file copying with XP. Vista SP1 improved file copy speed significantly (after screwing it up in the initial release). But you need two Vista SP1 (or higher) systems to achieve the full benefit.

You can try FTP instead of file copy for lower overhead. iperf/jperf is probably the closest you can come to IxChariot for low-overhead throughput testing. But the settings you use are important because the defaults don't necessarily provide the highest throughput. See Measuring Network Performance - Iperf and Measuring Network Performance - Jperf.

Thank you, Tim. You'v been very helpful. I will give those suggestions a try.

 

[phdeane]

And thank you for the great review, by the way. It was your review that spurred me to purchase the Netgear MoCA adapters.

 

[phdeane]

Just to give an update. I ran jperf across the MoCA adapters and using TCP and a TCP Window Size of 128 KBytes, I consistently get a throughput of about 90 Mbps - good enough.

Thanks again to Tim and the others who replied. I certainly learned a bit through this process.

 

[thiggins]

Thanks for letting us know how it turned out.

 

[Unregistered]

Thanks, and a thought on file copy throughput

First, thanks Tim for a really useful review. I bought a pair of MCA1001's based on it, and I'm very happy with their performance. I'd rate them 100 out of 100 if I they allowed remote management. Right now I'm looking for a third one, but don't want to spend the bucks to buy another set of two. Maybe Netgear will eventually be willing to sell these singly.

Regarding Windows file copy throughput, I'll bet the the culprit is the latency added by MoCA -- up to 9 msec each way if I recall correctly. I believe Windows file copies have frequent two-way communication, so latency will reduce throughput. (I've seen this effect in other situations with latency.) FTP, on the other hand, is designed to work efficiently with long latencies, and so will give better throughput. Any streaming protocol designed to run on the internet also should be fairly immune to network latency.

 

[richardd]

Mac/TV question

Tim,
Thanks so much for all your helpful information.
I am all-MAC, no-PC here. I want to connect a TV and Blu-Ray to my Netgear MOCA COAX-ETHERNET adapter. Will I need to change any settings on the adapter? Will I need to find someone's PC to do so? Will it work out of the box? Thanks.