New GT-AX6000 setup questions

[Megahurtz]

I just upgraded to a GT-AX6000 and am pretty pleased with it's performance, but I had a few questions I want to throw out and get some feedback on.

1. What speed should I be expecting from the 2.4gig band? I never really checked it on my previous router, but on the new setup I'm getting around 120-150Mbit. That strikes me as quite slow!

2. I've always read that turning on any of the Ai Protect functions has a dramatic impact on throughput. I did turn it on and the router's built in speed test was still maxing out my connection at 1.4Gbit. However actual speed tests from AX devices connected to it would get more like 600-700 Mbit. I thought there was something wrong with my wifi settings, but disabling Ai Protect get my wireless client back to testing at 1.4gig. So it clearly does seem that Ai Protect has a severe performance penalty, but why doesn't it affect the built in speed test too?

3. My new router was a refurb from Amazon that came in a plain brown cardboard box sealed with Asus tape. SSH'ing into it and running "wl country" tells me that the country code is set for Canada. Is that typical for North American Asus routers? Any limitations on me using it in the US?

4. What is the preferred transmit power setting for the wifi radios? I think they both defaulted to max.

 

[Tech9]

1. Excellent for 2.4GHz band.

2. GUI speed test is CPU limited.

3. Yes, up to 200mW on channels 36-48 vs up to 1000mW for the US model.

4. Depends on the desired coverage area. Performance setting is the default.

5. Your router has to work as DNS proxy for this. Post screenshots of your WAN and LAN settings.

 

[glens]

1. What speed should I be expecting from the 2.4gig band? I never really checked it on my previous router, but on the new setup I'm getting around 120-150Mbit. That strikes me as quite slow!

I've got a GT-AX6000 in the basement on the other side of a wall from my laptop. Occasionally I find the laptop associated with the 2.4GHz radio which is set for 20/40 AX and can handily get mid 300s both directions.

 

[Crimliar]

1. I've a few AX devices that occasionally decide they are going to connect to 2.4GHz and they tend to come in at around 300Mbps. AC dual streams manage 144Mbps, while the plethora of single stream "N" devices usually just hit those 65-72Mbps targets!
2. Ai-Protect - no it doesn't <nuff said>

 

[Megahurtz]

I found a solution to my DNS problem by turning on Tools>Other Settings>Wan: Use local caching DNS server as system resolver (default: No). I guess it was sending all my local queries upstream, hence they were failing. It says having that disabled should speed up resolution and improve reliability, but I don't understand why the default settings would prevent local name resolution from working. That seems like something you'd probably want to have working!

Back to my other questions!

1. I thought I had better throughput on my old router's 2.4g band. I guess I could hook it back up and compare the wireless settings. Nothing I have on the 2.4g band really needs huge speed, but 120Mbit seems abnormally slow.

3. Is there a way to permanently change the country setting? I'm able to modify it through an SSH shell, but it reverts back to Canada after a reboot. My 5g band is using channel 36 and I had noticed that the signal seemed a bit weaker than with my previous router.

 

[Ripshod]

3. Is there a way to permanently change the country setting? I'm able to modify it through an SSH shell, but it reverts back to Canada

I doubt anyone here would voluntarily help with what is technically an illegal operating mode.

 

[Tech9]

illegal operating mode

US settings used in US are okay, but I've seen US routers sold by NewEgg in Canada. This is a violation of local Canadian regulations.

Is there a way to permanently change the country setting?

Not an easy one for this specific model router and not something we can discuss on this forum. You have to look for clues elsewhere.

 

[glens]

My 5g band is using channel 36 and I had noticed that the signal seemed a bit weaker than with my previous router.

So those lightwaves coming from the router aren't shining as brightly as they could legally be made to do. Does it actually affect the communication performance with your clients? Are the clients even transmitting /that much/ power themselves back to the router?

Asked another way, if you could make it so a client could discern the light from 600 feet as opposed to only 200, would that enable the client to ever reach back from 200+ feet itself? Isn't the client's transmit power usually the limiting factor in a fringe scenario?

 

[L&LD]

That's asking the wrong question ('asked another way').

If the router can project its signal farther, it may (should!) be able to detect the reply too.

It is not up to only the client. The routers receiving RF design is important here (and it can and should be more discerning than what any client devices are capable of). And this has proven to be true many times as new models are introduced (even when they are in the same class as the earliest models, with inferior RF designs and implementations).

 

[Tech9]

Isn't the client's transmit power usually the limiting factor in a fringe scenario?

Yes and this translates into asymmetric link speeds to/from the client at distance. Higher AP power in home environment may have benefits though since most of the time upload is a small % of download. Your 20mW client may have 6Mbps uplink only, but enough to maintain the connection with 200Mbps downlink from 200mW AP near by. The "distance" can actually be small, but with 2-3 walls between the AP and the client. The fact is channels 149-161 (up to 1000mW) in Canada work better than 36-48 (up to 200mW) for dense condo environments, for example. The same in Europe for channels 100-112.

 

[glens]

If the router can project its signal farther, it may (should!) be able to detect the reply too.

Two completely different conditions. You could make a 1/10 wavelength antenna visibly glow by driving it with hundreds of Watts to get great transmit distance but it mightn't work for beans receiving. Then there's the receiving radio qualities to consider as well, and I agree that a typical router will have better signal gathering and processing abilities than a typical client. But all that breaches the bounds of the immediate topic.

this translates into asymmetric link speeds to/from the client at distance.

Which I consider less desirable. I'm a fan of symmetry as much as possible, hence opting for a fiber feed (same up/down) when DSL and/or cable may also be available, but asymmetrical (sometimes /drastically/ so). I realize the bulk of my Internet traffic is consumption, but...

 

[L&LD]

No, a router's job is to transmit and receive. Nothing different except your perspective.

As stated, we have more than a few examples of what I state above.

 

[glens]

In this case I believe I'd return the router and obtain one low-level-configured for my locale as opposed to doing hex-edits, if it didn't meet expectations.

 

[Tech9]

I'm a fan of symmetry as much as possible

In this case you have to stay in the same room as your AIO router. The further you go the client's uplink will slow down for the reasons above.

hence opting for a fiber feed

Your ISP speed up/down is unrelated to what we discuss.

 

[glens]

No, a router's job is to transmit and receive. Nothing different except your perspective.

As stated, we have more than a few examples of what I state above.

I'm not arguing with you, I agree, but not in the way I think you're thinking.

 

[L&LD]

What do you think I'm thinking?

What part are you agreeing with?

 

[glens]

In this case you have to stay in the same room as your AIO router. The further you go the client's uplink will slow down for the reasons above.



Your ISP speed up/down is unrelated to what we discuss.

Except for the symmetry factor which I hadn't introduced but was ever a consideration nonetheless. Limiting the topic to AP / client communication only, asymmetry is still noticeable and undesirable IMO.

 

[glens]

What do you think I'm thinking?

That the pertinent electronic circuit qualities of radio receivers and transmitters are the same, or something?

 

[L&LD]

No. That is not what I stated.

 

[glens]

This getting a little un-fun. I didn't state you'd stated it. I stated that I thought you were thinking it.