HomePlug AV2 Head-To-Head

[kamaran]

Great Article Tim, the website gets better all the time and is one of my favorite resources.

It's a shame these HomePlug manufacturers aren't as flamboyant with their technical specifications as they are with their marketing!

Question for you: Were the results for these AV2 SISO adapters (around 20-50% improvement over AV 500) consistent with your expectations?

If so, what are your expectations for the AV2 MIMO adapters coming later this year?

Thank you Sir!

 

[Dr_ARCHer]

Since Linksys uses the Broadcom chipset and Zyxel uses the Qualcomm/Atheros chipset, is it possible to do some cross pairing testing to find out how compatible are the two AV2 implementation?

 

[thiggins]

Since Linksys uses the Broadcom chipset and Zyxel uses the Qualcomm/Atheros chipset, is it possible to do some cross pairing testing to find out how compatible are the two AV2 implementation?

As the article states, both products use the Qualcomm Atheros QCA7450 / AR1540.

 

[thiggins]

Question for you: Were the results for these AV2 SISO adapters (around 20-50% improvement over AV 500) consistent with your expectations? [/quote]My only expectation was that performance would degrade significantly with distance.

If so, what are your expectations for the AV2 MIMO adapters coming later this year?

Same as above.

 

[kamaran]

So at the risk of sounding naive, what is the benefit of waiting for the MIMO versions?

 

[thiggins]

So at the risk of sounding naive, what is the benefit of waiting for the MIMO versions?

MIMO should bring higher throughput to the table. Both "AV 500" and AV2-SISO-500 designs primarily (only?) depend on the wider frequency band to bring higher throughput. MIMO should bring more bandwidth, even if it doesn't access all of the increased frequencies. But proof is in the silicon.

 

[Dr_ARCHer]

As the article states, both products use the Qualcomm Atheros QCA7450 / AR1540.

Sorry. Don't know why I kept thinking that the Linksys is using the Broadcom BCM60500 chipset.

Anyway, does cross pairing changes things? Does Linksys implementation differs from Zyxel implementation allows both to operate at AV2 speeds?

 

[thiggins]

Anyway, does cross pairing changes things? Does Linksys implementation differs from Zyxel implementation allows both to operate at AV2 speeds?

It might. But I expect differences would be mainly for same-room / same outlet use, which isn't a practical application.

At any rate, no further review is planned for now.

 

[smagro]

Zyxel are not the only ones calling them 600Mbps. TP-LINK do also (TL-PA6010) and Allnet (ALL168600).

It seems they feel they need to call it 'something bigger' than 500Mbps so that customers know that it is an upgrade over existing 500Mbps models.

 

[thiggins]

Zyxel are not the only ones calling them 600Mbps. TP-LINK do also (TL-PA6010) and Allnet (ALL168600).

It seems they feel they need to call it 'something bigger' than 500Mbps so that customers know that it is an upgrade over existing 500Mbps models.

The Allnet at least is AV2. The TP-Link is not (according to their specs).

 

[thiggins]

Hi Tim:

Great review, but I noticed that on the bottom of page 2 showing the firmware of the two adapters, the images are identical.

Nope. Look carefully at the firmware versions. They are different. Use the larger images.

 

[Fraoch]

Nope. Look carefully at the firmware versions. They are different. Use the larger images.

Whoops, I was looking at the utility version. <blushes>

I'll delete that post.

 

[bsotak]

Do you think the same-outlet test would offer similar results to the same-room?

Unfortunately I have my desk on the opposite side of the bedroom as the ethernet jack in my room. It would take a significant amount of effort to run the wire under the carpet, or around the baseboards (doorways on each side of the desk), or up through the wall and down. This could be a great solution, but I didn't know how the same-outlet versus same-room compared to one another.

 

[thiggins]

Do you think the same-outlet test would offer similar results to the same-room?

Depends on what else is plugged into the room. Cellphone chargers and other wall-warts tend to knock throughput down.

Best way to know is to try. Return the product if it doesn't do what you want.

 

[kamaran]

Tim any further insight as to what products and when are in the pipeline for AV2 MIMO?

I'm holding off upgrading my home network until they are available.

 

[thiggins]

Tim any further insight as to what products and when are in the pipeline for AV2 MIMO?

I'm holding off upgrading my home network until they are available.

next should be AV2 MIMO with link rates around 1Gbps before the end of this year.

 

[sm00thpapa]

Great article wish I read this before I purchased the item below. Now I am aware of the AV and AV2 standard. Thank you Tim.

http://www.newegg.com/Product/Product.aspx?Item=N82E16833704165

 

[krellboy]

AV2 comparison article

Thanks for the review, I have a few questions/observations that may be relevant. Please note I haven't actually tried the Zyxel, but it appears identical to the Allnet which I have tested. I do have the Zyxel on order and will likely test it this week.

The ground prong isn't just there for safety—it's actually used for signal transmission.

Regrettably it isn't there for safety or signal transmission, it's just tied to neutral or line through a MOV, I can't remember which.

While the boards are not identical, I imagine they don't deviate very far from Qualcomm Atheros' reference design.

My examination suggests that the Zyxel/Allnet is an un-optimized copy of the reference design, while the Linksys is reworked. The Zyxel/Allnet is slightly "crummier" in choice of components as well.

The Downlink Throughput plot below shows both the AV2 products start out with almost a 2-to-1 throughput advantage over the AV 500 products.

I'm entirely lost by what you mean by "uplink" and "downlink" in this context, even after looking at your diagrams.

Comparing with the ZyXEL's firmware version shown below shows the Linksys uses a slightly later version.

I screwed around a bit moving firmware and tonemaps between the Allnet and US Linksys (because in theory the Allnet should have been unblocked but performed significantly worse than the Linksys). [I strongly warn anyone against doing this.] Really no joy, my conclusion was that the Linksys performed better because it had a better analog front-end. I'm probably going to revisit this.

I've been doing my testing using iperf a specialized testbed where I can control the attenuation and add broadband noise and other impairments, since the results aren't directly comparable to yours I don't think they'd be useful but I could collect up some number in a table next time I get around to testing again (if anyone is interested).

 

[thiggins]

I'm entirely lost by what you mean by "uplink" and "downlink" in this context, even after looking at your diagrams.

Described here, linked at the top of the Performance section.
http://www.smallnetbuilder.com/lanwan/lanwan-howto/31625-how-we-test-powerline-products

 

[krellboy]

Described here, linked at the top of the Performance section.
http://www.smallnetbuilder.com/lanwan/lanwan-howto/31625-how-we-test-powerline-products

Thanks, I had read that.

I apologize if I don't get the "level" right on this question, I'm a firmware engineer working on small home net devices (and in the past larger ones like routers and switches), I don't know your background so I'll just have to wing it.

Let me try to clarify a bit: Imagine I send you many special, gold-plated 200-ft pieces of "monster-type" cat-6 and asked you to review them using your powerline methodology (just by running the cable through the house to the same test sites). You collect up the data and of course it's faster.

Then you think, hmmm, well, I have all of these types of fancy cable, I'll make a new product report category. So you take pictures and create charts and graphs and as you've already collected the uplink and downlink speed data (you used the powerline methodology) you decide to publish that to your review also.

Someone reads your review. They say "huh? uplink and downlink? It's a piece of CABLE, how could those numbers be different, I don't understand" So you send him a link to your "test methodology".

My question is this: After he reads your methodology would you expect him to say "Ah, that makes sense" or still be confused as to why a piece of cable has different "uplink" and "downlink" speeds?

Thanks,