Mixing Powerline Adapters

[Stompa]

If I had a pair of TP-Link AV200 powerline adapters connected to each other, and added a pair of AV1000 powerline adapters also connected only to each other, at what speed would each pair connect? There would be no connecting of an AV200 to an AV1000.

I've tried searching online, but opinion seems to vary a lot!

Has anybody actually tried something similar?

Thanks

[eepee]

Surely you have or can get the specifications for the devices?

So I think what you are asking is what speed you can expect?

It is virtually impossible to answer that question except by measuring at the end devices, presumably computers, themselves.

The reason is that computers do not generally make mistakes. There is so much transfer integrity coding that, if something does not arrive at the other end, the receiving machine asks for a re-send, generally called re-tries.

Let us say you have a dodgy hard drive. The transfers within the machine are working at full speed. However because the receiving CPU finds integrity faults in the data, it asks the drive to re-send it. So in effect the genuine transfer rate has been halved, at least for that particular bit if data.

So, if I have understood your question correctly, why could the speed vary from the specifications?

Are the re-tries caused by poor signal quality from the devices?
Or from poor wiring which has a lot of 'kinks' causing echos?
Or from dirty mains because of other equipment?
Or indeed by signal clash between your two systems.

As you can see, nearly impossible to measure.
Not sure if any of this helps!

Regards,
ep

[Allitnil]

I would expect somewhat less from each pairing than you would get if you only had the one pair. There's bound to be some cross interference but it's pretty much impossible to forecast how much that would reduce bandwidth compared to the inherent reductions in your wiring system. At the end of the day, unless your AV200 adapters are only just delivering enough bandwidth for your needs the answer is probably moot.

[Stompa]

Surely you have or can get the specifications for the devices?

So I think what you are asking is what speed you can expect?

It is virtually impossible to answer that question except by measuring at the end devices, presumably computers, themselves.

The reason is that computers do not generally make mistakes. There is so much transfer integrity coding that, if something does not arrive at the other end, the receiving machine asks for a re-send, generally called re-tries.

Let us say you have a dodgy hard drive. The transfers within the machine are working at full speed. However because the receiving CPU finds integrity faults in the data, it asks the drive to re-send it. So in effect the genuine transfer rate has been halved, at least for that particular bit if data.

So, if I have understood your question correctly, why could the speed vary from the specifications?

Are the re-tries caused by poor signal quality from the devices?
Or from poor wiring which has a lot of 'kinks' causing echos?
Or from dirty mains because of other equipment?
Or indeed by signal clash between your two systems.

As you can see, nearly impossible to measure.
Not sure if any of this helps!

Regards,
ep

Thanks for the detailed response.

My own guess is that I would expect 200 from the AV200's, and 1000 from the AV1000's. I appreciate that in reality there will be a degradation in performance for a variety of reasons, so the speed will end up being somewhat less (perhaps by quite a lot).

However, I have read reports which suggest that the maximum speed of the entire system will be that of the slowest connected adapter. So effectively the presence of the AV200's on the same circuit will drag down the top speed of the AV1000's to 200, despite the fact that they won't be paired with each other.

I guess I was hoping that somebody may have actually tried something similar and could confirm which is correct.

[Stompa]

I would expect somewhat less from each pairing than you would get if you only had the one pair. There's bound to be some cross interference but it's pretty much impossible to forecast how much that would reduce bandwidth compared to the inherent reductions in your wiring system.

Thanks, I'd agree with that.

[servodude]

If I had a pair of TP-Link AV200 powerline adapters connected to each other, and added a pair of AV1000 powerline adapters also connected only to each other, at what speed would each pair connect? There would be no connecting of an AV200 to an AV1000.

I believe each combination would try and establish a unique pair of links (and that this is pretty dynamic it's not what you've got it's what you're doing with it )
- at least that's how it is shown in tpPLC (as a directed baud rate depending on which adapter you choose to inspect)

FWIW I've got a mix of AV500 and AV600 - max displayed link rate is ~180mbps going down to ~95
- implying that it's the wiring that makes the biggest difference
- but as our ISP rate is only 55mbps it's moot for most of what we're doing

- sd

[xeny]

I suspect that the speed you get out of each pair will vary depending on how much traffic there is going across the other pair. If one link is idle, the other will run at best speed (which from memory is about 60Mbit/sec for AV200?). As the other link sees more traffic the first link will slow down.

You may get better consistency if you can pair all four together, as then you'll get AV200 between those units and AV1000 between the faster ones - see https://superuser.com/questions/935902/ ... line-plugs , and they'll at least know about each other rather than talking over each other.

[Stompa]

OK, thanks all, I'll have to try it and see.

[Midsmartin]

I think the speed may also vary depending on the brand/model of devices you use plus all the usual variation due to your particular wiring, and other devices plugged in that may interfere. You will just have to try it and see.

I stopped using powerline due to the extreme difficulty of tracing faults when it doesn't work. In my case, the power supply in my desktop pc, while apparently fine, completely stopped powerline working. It was ok when I replaced it.
I have seen other domestic appliances cause the same problem. I'd lean towards getting some mesh WiFi units, if that might work in your house.

[Infrasonic]

I think the speed may also vary depending on the brand/model of devices you use plus all the usual variation due to your particular wiring, and other devices plugged in that may interfere. You will just have to try it and see.

I stopped using powerline due to the extreme difficulty of tracing faults when it doesn't work. In my case, the power supply in my desktop pc, while apparently fine, completely stopped powerline working. It was ok when I replaced it.
I have seen other domestic appliances cause the same problem. I'd lean towards getting some mesh WiFi units, if that might work in your house.

Running data over mains wiring is a very 'noisy' solution - and power supply design is going to be cheap rather than competent with commodity electrical products. I think some of the better (£££) PL adapters have decent filtering for mains harmonics et al and running desktop PC's with a good UPS will cut down on PS noise too.
A commercial grade wireless AP (e.g Ubiquiti/Ruckus) should cover all but the biggest houses with decent bandwidth - if you can stomach the cost.
(They do come up S/H or refurbed on eBay...)