You could just do the speedtests between computers within your network using the likes of iperf.

You can also test within the local network to a machine on GbE using something like iperf, although that's easier on Linux.

Asteros:

 

Thanks Steve, I hear what you're saying and I understand from reading previous GZ posts that speed tests are tricky to draw conclusions from about wifi speeds.

 

I'm on Spark. Out of curiosity, I have ordered Fibre Max (up to 900 Mbps) to remove that bottle neck and will continue testing once it's been installed. I look forward to reporting back.

 

 

 

Such a minor difference isn't enough to draw any sort of conclusion about the performance - and without knowing the PHY rates of the client on both radios when the test was being performed you can't draw any conclusions from it.

Wireless doesn't cope well with packet loss and retries, so conducting a TCP speedtest which will saturate your UFB connection meaning traffic is either policed by your RSP or Chorus, and has the potential to deliver slightly different results across different brands of access points.

If you want to be testing performance you need to have a faster connection or as mentioned run iperf locally on your network and test to that.

Your biggest limitation is that your client radio is only 2x2 so you're only going to get pretty average speeds anyway - you're certainly not going to get anywhere near the 700Mbps or so that this AP is capable of with a 4x4 client.

Hi I've completed testing with Fibre Max (940 Mb/s) now installed. Both access points are performing well and are within the specifications described in the previous posts. Testing the access points has been an interesting process, and it's surprising how even 1 normal Gib board wall can affect Wifi speeds now I have a faster wired line speed. Thanks for the help!

I got *close* to 800Mbit on a 80MHz channel, Wireless AX on a 2x2 access point.

You're not going to get those sorts of speeds on a 2x2 AC device. And using 160MHz channels will make everyone around you very grumpy with you for taking a large chunk of the spectrum (also will likely not do you any favors in terms of performance either).

Read this: https://www.duckware.com/tech/wifi-in-the-us.html

Asteros: Hi, out of curiosity how do people obtain 700-800 Mbps on AC 2x2 devices? Do they use 160 MHz channels?

Yes you would need to be using 160MHz channels which will give you a PHY rate of 1733.3 

In the real world nobody uses 160MHz channels though for obvious reasons, and certainly in an average home it's totally pointless since you'd have very poor coverage unless you were planning on an AP in every room. That then poses re-use issues which will impact performance.

mAYH3M: I have found out that my speeds are better on a 5GHz 40MHz wide channel than 80Mhz wide.Also unlocking the DFS spectrum has worked well as most ISP supplied one stick to the 32-64 or 149-165 range.

For some reason SNR takes a huge hit when I try to bond all three.

Unless you're in the same room as the AP you're testing (and more importantly all clients are in the same room as well) that's totally normal. You get a lower power output using 80MHz so coverage is reduced.