Disappointing results with Moca

I just purchased a pair of Actiontec Moca 1.1 adapters (WCB3000N01) in order to provide a wired connection to an upstairs room. I have 150 Mbps Internet service, and when I do a test at speedtest.net I get around 170 Mbps download speed pretty consistently. After hooking up the adapters I did a test at the upstairs adapter and got download speeds in the low 30's. That's certainly good enough for most applications, but with Moca 1.1 advertised as providing up to 170 Mbps (I know, marketing-speak), I thought I would at least get up near 100 Mbps if not a little over.

My first thought was that it was due to the fact that my coax run to upstairs is pretty long, or maybe the cable is slightly flawed. So I brought the upstairs adapter downstairs and hooked it up directly to the other one using the coax cable that came with one of the adapters. I plugged into an Ethernet port of the adapter on the other side of the coax connection, ran a few tests at speedtest.net, and got download speeds in the low - mid 30's, pretty much the same as with the long coax run. I know that technically I should measure the speed directly between the two devices instead of from the Internet, but given that my Internet connection obviously isn't the bottleneck, I think it was a valid test.

I've seen other people post on various forums that they're getting better results, so I'm puzzled as to why I'm not. Any ideas?

Edit: I think at least one of the MoCA 1.1 adapters (WCB3000N01) was defective. I got better results using my old MoCA 1.0 adapters. I ended up buying a pair of new Actiontec MoCA 2.0 adapters, which work very well. See my follow-up post further down.

Are your low download speeds with the WCB3000N01 wired or wireless?

If wired, is the connection 10/100 or is it 1000/Mbps as indicated on your device's network status?

If wireless, what band are you using to test with and what channel width are you using? Have you tried different channels on each band to find the best one for your environment?

I connected to a gigabit Ethernet port on the WCB3000N01, and the status of the Ethernet adapter on my computer showed a connection speed of 1.0 Gbps.

Does the computer's adaptor or the WCB3000N01 have an option to enable or disable Jumbo frames?

If they do, are the set to the same setting (on or off) and if on, are they both set to the same size?

L&LD said:

Does the computer's adaptor or the WCB3000N01 have an option to enable or disable Jumbo frames?

If they do, are the set to the same setting (on or off) and if on, are they both set to the same size?

Click to expand...

Jumbo frames were disabled on the computer adapter. There doesn't seem to be a way to set it on the WCB3000N01, so I tried both sizes (4088 and 9014) on the computer adapter. The download speed seemed to improve marginally (maybe 4 or 5 Mbps) with it set to 9014. But of course there's no way (that I know of) to make that change to the PS3 that I use for streaming anyway.

I seem to have run out of ideas for you to test further with.

Hopefully someone that knows Moca more intimately can come by and offer additional ideas.

One last suggestion; turn off all network equipment including modem, router(s) and switches and leaving the units unplugged from the AC for up to an hour, see if that happens to reset things to normal.

Thanks for your suggestions, L&LD. I had previously tried powering everything off (though I only left them off for a few minutes, not an hour), but to no avail.

Earlier today during lunch I went to Best Buy and got a Netgear Powerline 1200 kit. I just set it up and ran a few tests, and am getting consistent 100 Mbps download speeds in the upstairs room. I'll make sure it operates consistently over the next few days and assuming it does, the Moca adapters are going back. However, I'm still puzzled that they didn't work better, so if anyone has any other suggestions in the next couple of days I'll try them.

My MoCA (old boxes) delivers a net yield at the IP layer of about 70Mbps (bits/sec). That's all I can expect with gen 1.

Like WiFi, marketeers like to slap the raw burst bit rate on the box. That burst rate is far higher than the net IP layter yield because of
Duplexing (WiFi, MoCA are half-duplex, not full-duplex like (proper) cat5, cellular data (excluding WiMax).

You're welcome. Sorry they didn't help in the end though.

Glad that the powerline kit seems to be working for now, but I would not depend on them working at a consistent level over the (extended) long term.

There is no way to run an Ethernet Cat5e cable there (look at monoprice for the flat variety at the length you need)? That would be the solution I would be pursing, if possible.

http://www.monoprice.com/Product?c_id=102&cp_id=10208&cs_id=1020814&p_id=9557&seq=1&format=2

Yeah, my results with the old Netgear MoCA boxes are about 85Mbps using a large file (~6GB) file transfer, to a wired laptop. Don't know what's going on there, but the powerline networking results sound good. Just hooked up the MoCA boxes at least a couple of years ago, and they worked perfectly, and have continued to do so. I have to restart them sometimes if we have a power failure, and very occasionally when I change routers, neither of which is unexpected by me *smile*. I change routers a bunch to play with various new firmware releases.

I'm interested to hear how consistent the powerline networking throughput is turning out to be for you. I was consider trying the Netgear PL1200's to see how they compared to the MoCA, and how consistent the transfer rate is.

Thanks.l

Powerline is very much hit and miss for various reasons. And as signal-attenuating devices are added in the home, or as interference-generators come and go (like some light dimmers), it's a battle.

That's why I use MoCA. Things rarely change with coax.

Pinotphile said:

I just purchased a pair of Actiontec Moca 1.1 adapters (WCB3000N01) in order to provide a wired connection to an upstairs room. I have 150 Mbps Internet service, and when I do a test at speedtest.net I get around 170 Mbps download speed pretty consistently. After hooking up the adapters I did a test at the upstairs adapter and got download speeds in the low 30's. That's certainly good enough for most applications, but with Moca 1.1 advertised as providing up to 170 Mbps (I know, marketing-speak), I thought I would at least get up near 100 Mbps if not a little over.

My first thought was that it was due to the fact that my coax run to upstairs is pretty long, or maybe the cable is slightly flawed. So I brought the upstairs adapter downstairs and hooked it up directly to the other one using the coax cable that came with one of the adapters. I plugged into an Ethernet port of the adapter on the other side of the coax connection, ran a few tests at speedtest.net, and got download speeds in the low - mid 30's, pretty much the same as with the long coax run. I know that technically I should measure the speed directly between the two devices instead of from the Internet, but given that my Internet connection obviously isn't the bottleneck, I think it was a valid test.

I've seen other people post on various forums that they're getting better results, so I'm puzzled as to why I'm not. Any ideas?

Click to expand...

Just wanted to follow up on this. The Powerline adapters have continued to work fine. However, last week I saw the thread about the Actiontec MoCA 2.0 adapters becoming available, so of course I had to order a pair of the ECB6200 (bonded). By the way, I think at least one of the WCB3000N01 (MoCA 1.1) adapters that I tried previously must have been defective. I got more than double the download speeds with my old Motorola NIM100 (MoCA 1.0) compared to what I got with the WCB3000N01's.

As for the ECB6200's, they work great. First I ran a couple of tests at speedtest.net with my computer hooked up directly to my router, and got download speeds around 230-235 Mbps. Then I connected the two MoCA adapters together with a short coax cable, connected one of them to the router, and connected my computer to the other one. I got the same results with this test as when I was connected directly to the router. This was obviously an unrealistic test, but given that the previous Actiontecs performed poorly on the same test, it was good to see the new ones perform as expected.

Next I moved one of the adapters upstairs (where I'll actually use it). I again got the same results as being connected directly to the router, despite the long coax run. The room with the router is downstairs on one end of the house, while the upstairs room is on the opposite end of the house above the garage. Both cable runs go to the basement, where I joined them directly with a barrel connector (they used to be on a splitter with a couple of other cable runs that I no longer use). So needless to say, I'm quite impressed with the performance of the ECB6200's.

Pinotphile said:

Just wanted to follow up on this. The Powerline adapters have continued to work fine. However, last week I saw the thread about the Actiontec MoCA 2.0 adapters becoming available, so of course I had to order a pair of the ECB6200 (bonded). By the way, I think at least one of the WCB3000N01 (MoCA 1.1) adapters that I tried previously must have been defective. I got more than double the download speeds with my old Motorola NIM100 (MoCA 1.0) compared to what I got with the WCB3000N01's.

As for the ECB6200's, they work great. First I ran a couple of tests at speedtest.net with my computer hooked up directly to my router, and got download speeds around 230-235 Mbps. Then I connected the two MoCA adapters together with a short coax cable, connected one of them to the router, and connected my computer to the other one. I got the same results with this test as when I was connected directly to the router. This was obviously an unrealistic test, but given that the previous Actiontecs performed poorly on the same test, it was good to see the new ones perform as expected.

Next I moved one of the adapters upstairs (where I'll actually use it). I again got the same results as being connected directly to the router, despite the long coax run. The room with the router is downstairs on one end of the house, while the upstairs room is on the opposite end of the house above the garage. Both cable runs go to the basement, where I joined them directly with a barrel connector (they used to be on a splitter with a couple of other cable runs that I no longer use). So needless to say, I'm quite impressed with the performance of the ECB6200's.

Click to expand...

If you test against a NAS, You should be getting 600~800 mbps. I'm getting 300mbps plus on the non bonded ECB6000's.

bds3151 said:

If you test against a NAS, You should be getting 600~800 mbps. I'm getting 300mbps plus on the non bonded ECB6000's.

Click to expand...

I do so wish that when people post speeds, they state briefly is this at the MAC/layer2 point or at the IP layer?

I wish people would say what useful THROUGHPUT they see, where throughput is TCP speed with a data source that cannot color the results. And mind their B's and b's.

Ok Steve. Just to clarify, in my case the 300mbps is useful throughput at the TCP layer 3.