Question on TP-Link AV1200 vs AV2000 & Powerline wi-fi

robdrums2097

Hi guys,

I have an ageing TP-Link Powerline AV600 system consisting of 3x ethernet only modules and 1x wi-fi / ethernet extender across my house. Which works nice and stably, but I've come to realise that as it's not gigabit or 5ghz wi-fi, it's strangling my Virgin Media 250mb connection in the locations I have the plugs. My router setup is a VM Superhub 3 in modem mode, an Apple Time Capsule (Extreme) router and an Airport Express upstairs connected to ethernet by Powerline. Speedtest shows me I am getting >200 download speed in my lounge next to the router by ethernet or 5ghz (which is where I do the most heavy duty streaming), but elsewhere in the house more like 30-40. Which is passable and stable but not amazing. So, I am considering upgrading the four modules to one of the latest models.

I've spoken to TPL, who naturally advised me that the AV2000 would give me the best combined powerline   ethernet performance, so I had decided to buy a 9610 KIT and a 9020P KIT to give me 4x AV2000 spec units, one with wi-fi for my dining room. My question was (and still is): is the AV1200 8730 KIT with AC1750 wi-fi going to be better for my dining room where I extend the wireless? I had thought maybe of buying a 9020P KIT AV2000 kit for the router end and where I plug my computers in wired, and an 8730 KIT for my wi-fi area as its wi-fi performance is (strangely, given that it's an older model) rated better than the 9610. This would be cheaper than the 9610   9020P. TPL also said that the speed of the wi-fi is governed more by the incoming speed down the cable than by what theoretical maximum the range extender could do, meaning having the AV2000 kit throughout would give me better overall wired and wireless performance than the 9020   8730. Still with me?

The reason I ask is because I then read AVForums' review of the 9020P kit (TP-LINK TL-PA9020P Powerline Starter Kit Review) and was surprised by their (actually quite unique) claim that the performance of the AV2000 kit yielded (them) poorer results than the old AV1200 kits. I'm sure this review is a thorn in the side of TP Link's otherwise universal praise for the AV2000 stuff, and every other review I have read says every time the AV2000 is superior, but it was interesting nonetheless as I was close to pulling the trigger on £200 worth of stuff.

SO... Does anyone have any real world experience of this predicament, and am I likely in the real world to:

a) see no difference at all between 1200 vs 2000,
b) definitely be better off with 2000 all round, even with its 'inferior' wi-fi,
c) better off with AV1200 all round,
or d) best off with my proposed mix of 2000 and 1200?

Also, if I did mix it (and even if I kept some of my AV600 stuff on the same network), would the overall speed be dragged down to the weakest link in the chain, or should I see superior performance where a 2000 is used and the same performance I currently get at the 600 (as long as the master unit plugged into the router end was 2000 spec)? TPL told me having the lower speed device(s) mixed with newer ones wouldn't affect the speed of the 2000 spec units.

Also, I wonder if a newer version might be on its way this year to offer AV2000 (or better?), more ports and better wi-fi like the older kit...

Hope this makes sense, and thank you in advance for any thoughts and input!

Rob

MarkyPancake

As I understand it and from past experience, your Powerline network will run to the lowest spec device, so one AV1200 adapter in an AV2000 set up, will reduce the whole network to the AV1200's capabilities.

robdrums2097

That’s what I thought, though TP-LINK claim not... I’d be interested to know if anyone else has experienced what AVF have & found that the new 2000s are inferior to the older ones, and the wi-fi performance.

Rob

ChuckMountain

Can you not do wired ethernet to access points as required.

Powerlines are always going to be the limiting factor in fast networks and as Internet speed has increased then this is going to cause issues and slowdowns.  A lot of tests show the speeds of most devices struggling to get past the 200Mbps but it all depends on your wiring etc.

That is also point to point if you introduce more into the environment then it's one thing at once communicating so slows down the network potentially.

robdrums2097

Sure, I know there’s no substitute, and perhaps if I ever rewire the house I might. But for now, the PLs offer a good, stable alternative. As the stuff that needs the real fast connections are next to my router, I’m not too worried about getting gigabit speeds upstairs - it’d just be nice to maximise what I do have for as little cost& hassle as possible.

mickevh

Unless it has changed recently, it is not the case that HomePlug devices are all required to operate at some "lowest common denominator" link rate.

Like Wi-Fi, each HomePlug negotiates the best link rate is can with any other plugs it wants to talk to (and possibly - again like Wi-Fi - not necessarily the same rate in both directions, but I'd need to research that.)

This makes sense if one contemplates: If one had (let's say) three HP's, two relatively close together that can sustain a high link rate and one further afield that can "only" sustain a low link rate, why constrain the two faster ones...?

It's some additional management overhead to maintain such a regime, but the management overhead of the HomePlug paradigm is pretty high anyway (compared to, say, ethernet) so it's not really a big deal.

The sort of data networking technologies we have at home are a heterogeneous mix and every communicating pair negotiate their link rates (ever erroneously called "speed") independently of all the others. As such the "speed" of one link does not influence the "speed" of another, just like on the road network the "speed" on one road has no bearing on the "speed" of another.

However, when one assesses the end-to-end throughput, (note "throughput" and "link rate" are not at all the same thing,) of any pair of end stations, (including with things like "speedtest") one sees the cumulative effects of all the "hops" along the pathway between the end stations. Generally, the slowest hop in the pathway has the greatest effect, though traffic levels can have a great effect too.

Hence, when trying to improve the end-to-end performance of any network pathway, one generally starts by finding out which hop is the slowest and doing something there. But of course, all that does is move the ignominy of being "the slowest hop in the pathway" somewhere else, so rinse and repeat until one achieves acceptable performance, runs out of money, has done the best one can, gets bored or any combination of the such.

MarkyPancake

All I know is, when I got AV1200 adapters and still had an AV200 adapter connected, the devices connected to the AV1200 adapter ran at the AV200's performance until I removed the AV200 adapter.