TP-LINK TL-PA8030P Reviewed

[sfx2000]

http://www.smallnetbuilder.com/lanw...it-passthrough-powerline-starter-kit-reviewed

Generally works - the end-points see 1 GigE... very easy to set up, and with the TP-Link util, easy to secure it.

Bridge performance - depends on a lot of things, but my going across a circuit breaker and then a long run across the house... that bridge was/currently an Apple Airport extended link, and generally work - that extended airport network was a solid 216Mbps on 2.4GHz WiFi, but WiFi being half-duplex, it's not quite the same as HPNA...

Even though this kit suggests AV1200 - I'm seeing about 257Mbps on that link at the moment, more than fast enough, but perhaps less than expectation.. it is faster than a WiFi 802.11n 3-stream connection on narrow channels...

the bigger deal for me is the plug-blockage... see below - cribbed from the SNB review of this kit, but I'll confirm that a 3-prong plug will run into mechanical interference... there's enough free space inside the housing to clear things for the big plug for US markets - understand that this is a global product and some plugs are big, and some are small...

 

[Nullity]

lol @ plug-blocking picture.

 

[abailey]

My understanding is that both HPNA and MOCA are also half duplex, just like Wifi. Is that not right?

 

[sfx2000]

My understanding is that both HPNA and MOCA are also half duplex, just like Wifi. Is that not right?

It's half-duplex across from one unit to the other - and shared if more than two units...

With this device, note that the Gigabit ports are full-duplex, and across the ports, it's gigabit, which can be very handy for tying together units within the home entertainment center, as connectivity on the device across the ports is full speed..

I'm pretty happy so far with the kit - I've got ethernet in most areas, except for a span over to the TV/Entertainment complex, where previously I had a wireless bridge (Buffalo WLI-TX4-AG300N) which is 802.11n, but doesn't support the UNII-3 channels (149-161) due to it being a Japanese device - and the ports on the back, while switched/full duplex, were limited to 100BaseT.

The AV1200 TP-Link kit is better than what I had, and so far, so good... I'm getting a consistent 100Mbps link across the connection, sometimes higher, but I haven't seen anything lower than 100Mbit/sec...

 

[sfx2000]

My understanding is that both HPNA and MOCA are also half duplex, just like Wifi. Is that not right?

HomePlug is half-duplex, but it's a scheduled half duplex - similar to Token-Ring, where each station transmits in a given time slot, and while it's transmitting, the others cannot, and must listen until the time slot is done...

The actual terms are Beaconing and Time Division Multiplexing, and the scheduling is determined by a master node (which is typically the first one turned on). Very different than WiFi or ethernet as most folks use...

Took some digging into the home plug specifications to understand what is going on there...

 

[sfx2000]

Here's a neat white paper that describes how HomePlug AV works - good read...

http://www.homeplug.org/media/filer...aa82-179e6e95cab5/hpav-white-paper_050818.pdf

 

[sfx2000]

After a week - pretty happy with it - just works, and I haven't seen any major issues... but that's expected as it's pair from a single vendor.

Not doing general internet across the homeplug span, it's basically doing video work over to the entertainment center... so using the TPLink util, that's the basic QoS profile in use...

Funny thing digging around in wireshark - these are basically an ethernet cable - no visibility into the data stream at all...

 

[sfx2000]

FYI - don't put these on a power strip - TP-Link and others are not kidding...

For my mock-up - one end was on a power strip, the other was direct to socket -- Moving the one on the power strip to a socket more than doubled the bandwidth across the span...

AV1200 is pretty sensitive...

 

[sfx2000]

Interesting - doing some quantifying of my network across different spans - gigabit ethernet vs. homeplug below...

host is on a switch that is connected to both... the latency, and variability is quite a bit higher over AV1200...

 

[L&LD]

sfx2000, it would be great if you could explain the graphs and what they're showing in this and the other threads you're sharing this type of info.

It is almost impossible to know if the upper or lower graph is the 'better' one. And why.

Graphs without titles are also less than informative. What is actually being measured here (yes, I know; the latency, but I mean the hardware too) and how does this affect the actual network use?

Sorry, not trying to be critical here (I thank you for all the info you share with us), but I look at these graphs and say '??????'. They may mean something to you with a glance. I would love to learn to look at them and glean something from them too.

Thank you.

 

[sfx2000]

the script sends 20 ICMP pings every 5 minutes to various hosts, and calculates the round trip time for the returns...

the scale on the right maps that RTT over time.. and also indicates if there is packet loss or not - the green line is the median value, and the grey is min-max...

As one can see - HPAV is roughly 4x latency in this setup compared to gigabit ethernet...

 

[sfx2000]

And note - both client target specs are the same - they are my WAP's, hardware/firmware are the same... specifically configured for this test...

 

[sfx2000]

After a week - pretty happy with it - just works, and I haven't seen any major issues... but that's expected as it's pair from a single vendor.

After a month or so in use - starting to see the random dropout issue that others have seen/reported elsewhere on the forums - lot of complaints on TP-Link's support forum, but little action from them.

I'm putting this into the "do not recommend" bucket until TP-Link provides a fix...

 

[sfx2000]

Quick update - since flashing the AV1200's with the Mar 2016 firmware, the drop-out issues I've been having seem to have gone away.

Can grab the update here - http://www.tp-link.us/download/TL-PA8030P-KIT.html#Firmware

BTW - the admin utils they have on the Web Site are Windows only, but if you contact TP-Link support, they do have a beta version for Macintosh... be persistent, not all reps on Tier 1 know about it, but Tier 2 does...

 

[sfx2000]

The version of Firmware you want is TL-PA8030P(US)_V1_160328 - the other link is the previous version - remember to flash both items in the archive (there's a readme) and flash both adapters..

 

[Dan-H]

Thanks for the review.

I just picked up a pair of these on sale at costco. It looks like it shipped with an older firmware version ..._150209 I'll likely flash them both tonight or tomorrow.

They worked fine out of the box, and are performing better than my 5GHZ router in bridge mode.

One comment is the documentation is slim around the QOS settings.

 

[sfx2000]

One comment is the documentation is slim around the QOS settings.

And it's not that obvious either - HomePlug AV has different treatment of QoS compared to ethernet and WiFi - I posted a chart some time back that mapped out the QoS values across different access technologies, and for the life of me, can't find it now... (with 7800+ posts, it's hard to find a single one in all those, lol)

 

[Dan-H]

excerpt from: http://www.tp-link.us/res/down/doc/TL-PA8030P_KIT(US_V1_UG.pdf

How important is it to wire this to match the picture on the left vs the sad face picture on the right?

The "local" side adapter is plugged into a wall outlet with only the ethernet cable to the router connected. The router is on the other side of the room connected to a power strip plugged into a different outlet.

The "remote" side is plugged into a wall outlet and the desktop computer is plugged into a power strip on a different outlet than where the adapter is plugged in.

This is US home wiring and the local outlet are and remote outlet are on different circuit breakers ( if this info matters).

I would simply test this but in both cases getting to either power strip involves moving furniture so it is not that easy.

 

[thiggins]

My Location E for the powerline tests has many wall-wart adapters plugged in. I unplug them ALL while I test, even adapters in the next room.

That said, I've found an outlet strip, even cheapies with no surge protection or filtering, can provide some high-frequency noise filtering. So you have that going for you in your setup.

 

[System Error Message]

The frequency used by powerline isnt really that high, its low enough that i can measure with my oscilloscope. As long as 200Mhz and lower isnt filtered that should be fine.