Powerline AV2 1800 adapter

[moron+genius]

in case some of you may be interested... just launched not yet available retail

hope SNB can conduct a proper speed test soon

http://www.sineoji.com/products/products-details/pl1800e

http://www.youtube.com/watch?feature=player_embedded&v=N4ZLjC-6T-g

 

[thiggins]

MIMO AV2 powerline adapters are not going to be available in the U.S. until September or October.

 

[themishk]

this is crazy
1800m

the test only shows the powerplug can go up to 300m real DL rate
and it seems u can buy this product now @ singapore?

 

[Dr_ARCHer]

According to the company website, this product will be on sale tomorrow (Singapore time) at the local (Singapore) computer exhibition. I will be dropping by to see if I can pick up some units. The AC87U will also be launched there (units came in today).

 

[stevech]

this is crazy
1800m

the test only shows the powerplug can go up to 300m real DL rate
and it seems u can buy this product now @ singapore?

A bit better grammar here would help us comprehend. Sorry.

 

[themishk]

A bit better grammar here would help us comprehend. Sorry.

my apology.
i am just saying the results are pretty impressive (getting 300mbs download speed), but since I am not living in Singapore, I would not be able to test it out myself

 

[themishk]

According to the company website, this product will be on sale tomorrow (Singapore time) at the local (Singapore) computer exhibition. I will be dropping by to see if I can pick up some units. The AC87U will also be launched there (units came in today).

any update on yr trip? or any idea how ppl outside Singapore can buy it?

 

[Dr_ARCHer]

any update on yr trip? or any idea how ppl outside Singapore can buy it?

Unfortunately, it seems that there were some production delays, and they could only accept pre-orders. Funny thing is that all local ads and write-ups didn't mention that it is for pre-orders only. Trip down was wasted, but fortunately, I picked up two units of the AC87U

 

[moron+genius]

also Aztech 1200 av2 with passthru avail for preorder today in sg

 

[moron+genius]

the speedtest in the video is limited by the tester's broadband speed, wonder what its max performance is. hope SNB can conduct a proper test... but I doubt it..

my apology.
i am just saying the results are pretty impressive (getting 300mbs download speed), but since I am not living in Singapore, I would not be able to test it out myself

 

[themishk]

i think the time for homeplug has came!

 

[BBM]

One more product
http://lea-networks.com/americas/wp...ine_communications/NetSocket1800US_070914.pdf

This one uses NA plug.

 

[azazel1024]

the speedtest in the video is limited by the tester's broadband speed, wonder what its max performance is. hope SNB can conduct a proper test... but I doubt it..

In my limited experience with homeplug stuff, you are generally looking at 25% or so of "rated speed" for maximum usable speed and that is on the same circuit without much feeding back in to the electrical wiring. IE ideal conditions.

PHY rate typically seems to be between 40-60% of max PHY rate listed for the product under ideal conditions and you are generally going to get around 50% yield on PHY rate once ECC, rebroadcasts, etc are factored in.

So a 1800Mbps product under ideal conditions and if the firmware, adapter, etc are good, you are probably looking at around 450Mbps at best. Which is rather good, but this is under ideal conditions, which isn't often what you'll be under, since you generally aren't using a homeplug on the same circuit, because that is often just across the room or maybe one room over.

So depending on the installation and everything else, figure more in the 150-300Mbps range...which is still very good.

Its nothing like 1000Mbps full duplex of gigabit ethernet though. Or for that matter, not nearly as fast as a good 1300Mbps bridge 802.11ac connection. I've seen plenty of guys at decent distances bridge AC1750 routers (in 5GHz 3:3 80MHz mode of course) get usable numbers of over 300Mbps. Heck, Line of sight and 20ft away I can get over 440Mbps between my Intel 7260ac and my AC1750 router, and that is only 2:2, not 3:3. Stick a wall in the way and 30ft and I can still get over 300Mbps. That might not be the kind of situation you'd have for powerline, the wireless situation is often worse, which is why someone is opting for powerline.

However, powerline still really remains the networking choice of last resort. It is getting better and by leaps and bounds the last couple of years, but it is still the 3rd rate networking choice (at least with the newest PL adapters, probably better than MoCA, at least 1.1 and maybe even 2.0 (which pretty much can't be found) under some select circumstances with this upcoming crop of 1800Mbps PL adapters).

 

[thoma]

In my limited experience with homeplug stuff, you are generally looking at 25% or so of "rated speed" for maximum usable speed and that is on the same circuit without much feeding back in to the electrical wiring. IE ideal conditions.

Mark Hazen at Intellon estimates a 3:1 ratio between the raw PHY rate and the maximum practicable TCP transmission rate.

• 200 Mbps raw PHY
• 150 Mbps coded PHY
• 92 Mbps UDP
• 65 Mbps TCP
• 35 Mbps Windows file transfer

Source: http://www.electronicproducts.com/A.../Data_rates_of_HomePlug_and_802_11_Wi-Fi.aspx

 

[azazel1024]

That is interesting, but not accurate. Unless they are pulling SMB numbers from somewhere odd (I don't know what SMB1.0 overhead is, I know it is chatty compared to SMB2 and later), overhead is tiny.

You are talking only 2-3% of the network traffic with a 1500MTU is going to be server message block traffic. The rest is going to be TCP overhead (something like 6% with a 1500MTU) and payload (which, as far as TCP is concerned, is part of the payload). That is based on "over ethernet".

Wireless has its own overheads to add, as does home plug, on top of all of that.

However, if the raw PHY is something like 150Mbps, with 11n wireless and without custom extensions to it (like low density parity checking, which is more of an 11ac thing, but some 11n drivers/radios support that) you get about 114Mbps out. 76% net yield is roughly the maximum possible rate, as ECC, beacon frames and other bits take up the rest. That is supposing zero retransmits due to dropped/garbled packets or other possible legacy things.

11g the max net yield under theory is closer to 55% (Maybe a little more?).

I can that if there is a high retransmit rate, that UDP would be a lot higher than TCP traffic (though in practice, on a "lossless" (as in lost packet-less) connection, UDP is typically only about 6% faster than TCP, because again, TCP only takes up around 5-6% of a packet for 1500MTU packet size, with jumbo frames the difference is less, though the retransmit penalty is higher).

I can't think of anything that would explain the difference between 65Mbps for TCP and 35Mbps for windows file transfer, using a 200Mbps PHY raw rate. If anything the Windows file transfer should be more like 60Mbps based on the other numbers.

 

[smagro]

I think Homeplug alliance have got a real problem with their naming schemes. Every time I read about Homeplug on deal sites, 90% of the comments are people complaining about the real world speed vs the label on the box.

It's only going to get more ridiculous in the future. How many times faster in real world usage will 1800Mbps models be compared to 200Mbps? Joe Average would expect a 9x difference based on the naming.

They would have been better off naming them Gen1, Gen2 etc or some other scheme where customers know they are getting the newer, 'better' one but without any numeric connection rates on the box or in the model name.

 

[thiggins]

I think Homeplug alliance have got a real problem with their naming schemes. Every time I read about Homeplug on deal sites, 90% of the comments are people complaining about the real world speed vs the label on the box.

Unfortunately, HomePlug is following Wi-Fi's lead. When did you ever get the speed of the "label on the box" for any of those products?

 

[azazel1024]

Unfortunately, HomePlug is following Wi-Fi's lead. When did you ever get the speed of the "label on the box" for any of those products?

Can't argue. Personally I just don't know that there is a particularly good way to name them or rate them. The signaling rate isn't necessarily the worst way, and does make it easy for someone "in the know" to understand the product. Average consumer, not at all. However, I don't think it is necessarily something that can be easily broken down the consumer.

Its a fair bit more complex than "horsepower" of a car...and horsepower measurements don't tell you much of anything without knowning the actual torque curve, gearing and vehicle weight. Most consumers aren't going to think anything other than "more is better", without understanding anything else. Though of course, most times they get to test drive a car before they buy it (not so with a wifi router).

Wifi and powerline also have to deal with noisy environments which make it that much more complex. A particularly noisy electrical system might see very modest gains going from AV200 to AV1800 because the system is just too loud to do much more, no matter how complex or how many spatial streams you want to try to run. Same thing with wifi, if a band is noisy enough, moving from 65Mbps 2.4GHz to 600Mbps 2.4GHz might lead to only small gains in performance, not the nearly 10X increase it sounds like on the box.

That said, in relatively ideal conditions, at least for wifi, you CAN come resonably close to following that trend. With an ideal setup (close to the router, zero interfence, very good client and basestation), a 600Mbps 2.4GHz connection is likely to get you in the area of 8-9x faster performance than that 65Mbps 2.4GHz link.

I'd imagine the same is true for Powerline. Under very good/near ideal circumstances, AV1800 might well be roughly 9x faster than AV200...its the difficulty in finding those circumstances.

 

[ChristineBCW]

I wonder what marketing options are available for 'speed claims'. I think WiFi, MoCA and HomePlug suffer from so many hidden, interfering and distance issues that marketeers simply can't come up with a better option.

In the above posts, we see a good stab at some equation: "About a quarter of the claimed speed". That's my experience as well.

During the last years of the CRT market, marketeers started using "Viewable screen" as a better claim because, for once, they listened to the grief and comlaints of consumers - "My 15" inch monitor isn't even 14.1 viewable!"

I don't believe that honesty made the CRT industry suffer one bit. I saw relief on consumer faces, actually - and certainly felt that, myself. "Finally - honest marketing claims!"

I suspect WiFi, MoCA and HomePlugs have discovered some truthful claim but, once again, they are reluctant to be the first ones on the block to be forthcoming. So, we're left with marketing claims and consumers tsk-tsk, throw a wet-blanket over purchasing and take a sit-and-wait attitude.

Wireless is giving us even more reasons, of late, with the wave upon wave of "new tech" or at least "new tech marketing claims".

As K once said, "I guess I'll have to buy The White Album again."

 

[thiggins]

I actually like the current Wi-Fi naming scheme (N900, AC1750, etc.) that has become the de facto standard. By using max PHY rate, they are following the precedent set by Ethernet. Once you understand the rules, it's easy to understand what you are buying.

Now if marketeers would stop the overoptimistically rounding up (I'm looking at YOU, ASUS, and all you HomePlug AV 2 marketeers) and everyone use the same numbers, it's at least a step toward unconfusing consumers.

Even Ethernet has derating factors (most people would be surprised at how far from 1,000 Mbps their Gigabit Ethernet connections really run). It's just that the performance variation in alternate-to-Ethernet technologies is so high and people notice more.