Question Glad but disappointed with my new Wireless Router ?

[Boris_yo]

I've bought TP-Link Archer C6U (EU version) recently to replace my old 802.11n router which was actually transmitting at low power being environmentally-friendly. Not sure how 5V vs 12V makes a difference on a planet with irresponsible consumers filled with plastic.

Been using the new one for a few weeks and mostly I am satisfied with it. This one has automatic restart schedule and a physical restart button on its rear panel. My old one didn't have either and sometimes it would deny access to its web management interface which I would fix by restarting it. The new router has USB port for USB thumbdrive. I don't know if thumbdrives are meant to stay in the port long-term or to be used when necessary. I left one inside for an hour and I could tell it was very warm on touch.

The download speed has improved definitely in comparison to speeds of old router and are now reaching my full internet package speed of 200/5. Being 1 meter from the router gives me over 192Mbps (1 on screenshot) and in my bedroom 22.5 feet (6.8 meters) and 2 wooden doors away from the router gives me between 166Mbps (2 on screenshot) up to 186Mbps:

Locked channels is my disappointment though. In this router's settings I can only choose channels 36, 40, 44 and 48 but I remember seeing people choosing other channels that are not available to me. This prompted me to contact technical support of distributor. I was told to update firmware but I tend to be careful with these things because I don't like to fix what isn't broken. By the tone of technical support person I sensed that I am dealing with FAQ equivalent so I decided to not update.

I visited TP-Link's forums and found other people, including people from EU complaining about only UNII-1 band being available. I checked with US FCC counterpart in my country and EU and found out that other bands are allowed in both so why are UNII-2, UNII-2 Extended and UNII-3 are locked in this device? I know UNII-2 and UNII-2 Extended spectrum is used for weather radar and military installations but I think that unless they are not in a proximity where your router would cause them interference it shouldn't be an issue. If there is something close isn't DFS supposed to solve these kind of issues. One person reported being in a clear zone and DFS not being triggered in months.

In my case the speeds I have reported were measured on 80Mhz channel width (AP: Not Available 5). There is only one 80Hz channel width available that spans 5170MHz - 5250MHz spectrum of UNII-1 band and where I am located there are 2 other users sharing 80MHz channel width:

I understand that given the speed test above, the interference is small to be worried about DFS channels. But one person on the forum who said he is living in apartment building was concerned due to UNII-1 band being congested. I haven't seen any reasonable reply from TP-Link forum personnel so I inquired about the issue. And rightly so because after back and forth between technical support and their engineers the answer I received was reasonable. However it was not until I was told that there is no solution:

"Thank you for your valued reply.
Archer C6U support Band 1/2/3. However, band 2 and 3 is in professional used and will not be opened for home routers used. Only Band 1 could be used for Home routers.
It is not only limited by TP-Link routers, many other routers with different brands are the same.
We are so sorry that we have no solution to solve this issue.
Highly appreciated for your understanding."

If it all comes down to entry level, high level and enterprise level routers why not mention the limitations in fine print on product's page or at least on device's packaging? If I want a mid-range router is the price range the only guide I should follow or is there something else? Are there SOHO or a company labels? When I was buying 2.5GHz 802.11n router I expected 13 channels and I got 13 channels. Would the same thing be with high-range router of same band and protocol? TP-Link mentioned that other router brands conduct same practices. I don't have experience with other routers but if I was in a situation where UNII-1 band was congested I would most likely benefit from channels in DFS and low power spectrum. Just having UNII-1 and UNII-2 available gives me second 80MHz channel width to switch to in case there is a congestion. By the way the AP Zaguri earlier was on 80Mhz channel width covering 52, 56, 60, 64 channels. I don't know what router they use. But just look at how much more there is to the rest of spectrum:

 

[maikutech]

I see two problems with that, if you are using the old ethernet cables for your new router, that could be a problem of why your mbps download and upload speed is not at full peek.
The other problem your mbps internet speed needs to be adjusted by your isp so you are no longer being choked while using the internet for devices greater than 8 in a home/building.

Go with CAT 8 ethernet cords for your router and modem:
https://www.amazon.com/Ethernet-Int...cphy=9011163&hvtargid=pla-1352124083652&psc=1

 

[tennis2]

Aside from feature/function additions/improvements, there are oftentimes security vulnerabilities that are patched in router updates. Checking for router updates is a good idea.

The C6U is a pretty basic router. Cost in the US varies, but somewhere around 50 USD? Not sure what you thought you were getting. You said speeds were drastically improved and matching your service speed. Coverage can be highly variable based on home construction and whatnot. Maybe overthinking this?

Use wired if/where possible.

 

[USAFRet]

Go with CAT 8 ethernet cords for your router and modem:

"Cat 8" is irrelevant.

Anything up to gigabit performance, out to 100 meters....good quality Cat5e is just fine.

 

[tennis2]

Go with CAT 8 ethernet cords for your router and modem:

CAT 5e will carry Gigabit. CAT 6 will carry Gigabit more reliably over a significant distance. Past that point, you're in diminishing returns of connections/junctions/etc. If the price difference isn't large, sure, get CAT 8, but CAT8 is more of a datacenter/enterprise use cable.

 

[maikutech]

So you would recommend cat 5e for his internet connection ?
I had good reason not to recommend a cat 5e ethernet cable.

 

[tennis2]

So you would recommend cat 5e for his internet connection ?
I had good reason not to recommend a cat 5e ethernet cable.

CAT6 will do 10Gbps up to ~150'. How much more do you need?

Generally CAT6 is price-equivalent to CAT5e. For shorter patch cables (<10m) I see no issues with CAT5e for gigabit. If it's longer distances, more potential for interference, more "permanent" installations, I go CAT6.

CAT8 is 3x-4x more expensive than CAT6, and what benefit are you getting in home use with a $50 gigabit router?

 

[USAFRet]

So you would recommend cat 5e for his internet connection ?
I had good reason not to recommend a cat 5e ethernet cable.

Sure.

What reasons do you have for not recommending it?

 

[bill001g]

I see you have found the chart on wifi channels. This one is a bit more detailed but is basically the same

https://en.wikipedia.org/wiki/List_of_WLAN_channels

.

So part of the DFS rules are you can't manually set those channels you have to let the router in auto mode. In theory it will detect things like weather radar. Problem is it does not always pick what you think is best.
Many routers just do not want to deal with this mess.

This is exactly the same reason most wifi6 stuff, especially end devices, is only 80mhz rather than 160mhz.

The solution is going to be wifi6e.....we hope but wifi vendors tell so many lies.

 

[Boris_yo]

The C6U is a pretty basic router. Cost in the US varies, but somewhere around 50 USD? Not sure what you thought you were getting. You said speeds were drastically improved and matching your service speed. Coverage can be highly variable based on home construction and whatnot. Maybe overthinking this?

I am okay but imagine for a moment hypothetical situation. I am that guy from apartment who upgraded from 2.5GHz device to 5GHz hoping for better speeds who buys the router without knowing that the router he is buying offers only 4 channels. Later after experiencing interference and speed fluctuations he finds out that the area is heavily congested with neighbors using and switching between the same channels he is using. This could be different if manufacturer mentioned in a fine print that the router is restricted only to UNII-1 band.

Not sure why manufacturers combine 2.5MHz and 5GHz bandwidths unless it's for sugar-coating. Surely I won't get 1900Mbps if I buy 2.5GHz 600Mbps + 5GHz 1300Mbps router at the same moment. But not informing customer about channel limitation is not right. Most home users won't bother with it but there will be outliers.

 

[bill001g]

You know being a engineer we hate the marketing guys....but then again they have to deal with the customers who demand stupid stuff.

Most end users roll their eyes when you talk about channels and even more so stuff like UNI bands. They go "big number good". You want to get even more mad go read about the mesh stuff. Especially the ones that do not have dedicated radios for the backhaul. BUT that makes the issue of lack of channels even worse because you now need 2 80mhz blocks to make it work well.

I guess I have been around long enough to just know wifi numbers are massive lies. In reality you might get 300mbps maximum at a reasonable distance from the router on 5g and much less using 2.4g. Wifi6 assuming you can use 160mhz with little interference people can get 600mbps but the vast majority drop back to 80mhz and it is no faster than 802.11ac.

If you want to be the test case get a wifi6e router and end devices. In the EU there are 3 160mhz blocks and there are about 7 in the USA. Since wifi6e has only recently be easier to find there likely are not a lot of users. This lack of real consumers using it also does not tell us how well this really is going to work. Mostly you see marketing bull and people who like to brag how large their "number" is.

 

[tennis2]

I am okay but imagine for a moment hypothetical situation. I am that guy from apartment who upgraded from 2.5GHz device to 5GHz hoping for better speeds who buys the router without knowing that the router he is buying offers only 4 channels. Later after experiencing interference and speed fluctuations he finds out that the area is heavily congested with neighbors using and switching between the same channels he is using. This could be different if manufacturer mentioned in a fine print that the router is restricted only to UNII-1 band

I suppose that depends on "that guy's" knowledge level of wireless networking. As I said, the C6U looks to be an entry level ac router at an entry level price, you get entry level features. If living in an apartment, it's expected that wireless traffic is highly congested based on the proximity of all the residents. Surely it's difficult to know exactly what you're getting, or what's worth paying more for, but TP-Link's product page for the C6U (here) is fairly detailed, even if it doesn't advertise how many channels there are (which would be nice). I'm not entirely sure that you've established that the 4 channel "limitation" is an issue anyway though(?).

Honestly I'd say that your level of knowledge/research has exceeded the vast majority of users in general, especially for buyers of this tier of wifi router. I'll admit that this exceeds my own level of wifi knowledge (or interest in learning about). I've done some wifi channel scanning, but ultimately end up leaving that on auto since most other peoples routers in range of interference are switching channels. Better/pricier routers may/should have more channels/bands to select from to keep you "above" the average/congested ranges.

 

[Boris_yo]

Wifi6 assuming you can use 160mhz with little interference people can get 600mbps but the vast majority drop back to 80mhz and it is no faster than 802.11ac.

I thought Wifi6 is not much different than Wifi5 however has coloring feature, that is marking nearby broadcasts that interfere and avoiding them. tennis2 user mentioned that most routers are set to auto-channel switching. I think coloring improves that even further for better response times and less collisions?

This is the answer (#2) by TP-Link tech support to someone who had EU version of router. The support claims that channels are restricted to comply with CE regulations. Yet ETSI (European Standards Organization) allows for extended channels on a broad 5GHz spectrum. I wonder what CE standard his device does not conform to:

https://community.tp-link.com/en/home/forum/topic/232992

If living in an apartment, it's expected that wireless traffic is highly congested based on the proximity of all the residents. Surely it's difficult to know exactly what you're getting, or what's worth paying more for, but TP-Link's product page for the C6U (here) is fairly detailed, even if it doesn't advertise how many channels there are (which would be nice). I'm not entirely sure that you've established that the 4 channel "limitation" is an issue anyway though(?).

Better/pricier routers may/should have more channels/bands to select from to keep you "above" the average/congested ranges.

How do you know the router you are getting is not entry level one without guessing? There are AC entry level routers but there are AX as well. Are AX and higher pricing good indicators of extended bands being available in that router? A small fine print would not hurt but I am no marketing specialist. In my use case I do not experience issues but just wanted to try extended channels in my testing so no big deal for me.

This is the post by end user that I mentioned:
https://community.tp-link.com/en/home/forum/topic/219674

This user claims that ASUS has more channels:
https://community.tp-link.com/en/home/forum/topic/104049

This user has VR600 router and says that the solution in his case would be availability of extended channels. Would be nice to mention that they are not available in his router:
https://community.tp-link.com/en/home/forum/topic/233958

 

[Boris_yo]

@bill001g I was surprised to receive almost 190Mbps on my laptop that uses 802.11n network adapter. I thought these kind of speeds are possible only on 802.11ac but it turns out not. This is probably due to 40Mhz channel width that can now function in its full and without much interference that it encountered before?

If my download speed package was 500Mbps then I would probably get even closer to theoretical 300Mbps possible with 802.11n standard?

 

[bill001g]

This is my favor chart on the speeds verse the types.
https://docs.google.com/spreadsheet...nKQ1O7abij/pubhtml?gid=1367372895&single=true
The orange is 802.11n which is a fully supported subset of 802.11ac which is the green. The rest of the chart are the fancy wifi6 protocols.

You generally will not get much over 1/2 because wifi is half duplex and it is combine transmit and receive speeds together. This is like calling a 1gbit ethernrt 2gbit but with ethernet it can actually send 1gbit and receive 1gbit at the same time because it is full duplex.

Also the more complex the encoding and the more mimo channels you use the less you gain. So going from 1x1 mimo to 2x2 give a lot more but not really double going from 2x2 to 3x3 gives less and less. They would use 100x100 mimo if this was magic.
The more fancy the encoding the more susceptible to interference it is. Things like Qam1024 pretty much only work in the same room for most people.

Now your nic and router might also support some of the non standard encodings. qam256 is not really in the standard for use on 2.4g but some devices use it.b You also find some 802.11ac devices using qam1024 which really is only defined for wifi6.

I would be happy and don't jinx yourself getting 184 on 802.11n is very good.

 

[Boris_yo]

You generally will not get much over 1/2 because wifi is half duplex and it is combine transmit and receive speeds together. This is like calling a 1gbit ethernrt 2gbit but with ethernet it can actually send 1gbit and receive 1gbit at the same time because it is full duplex.

If it was full-duplex that would mean I would get 300Mbps received + 300Mbps sent?

I would be happy and don't jinx yourself getting 184 on 802.11n is very good.

If 25Mbps is required to stream UHD movie to 1 device, I would be able to stream up to 7 devices in total at home?

 

[bill001g]

It can't really be full duplex because it is radio. If it used different radio channels for send and receive you could get 300mbps up and down in theory at least. Many modern forms of LTE on cell towers use different radio frequencies for download and upload.

Wifi if you run very unrealistic test cases can technically get 300mbps which is why the manufacture have that number. You would need some kind of data protocol that say only transmitted. With normal internet type traffic the router sends a packet and the pc sends a acknowledgment back that it got it. This mean it must constantly switch back and forth from transmit to receive. It is actually suprising it works as well as it does espceially if you factor in multiple devices sending and receiving data with the router even using different data encodings. All this switching stuff around takes time and causes you do not be able to get the magic theory numbers.

 

[Boris_yo]

It is actually suprising it works as well as it does espceially if you factor in multiple devices sending and receiving data with the router even using different data encodings. All this switching stuff around takes time and causes you do not be able to get the magic theory numbers

The router I bought recently has dual-core 880MHz MediaTek CPU and USB 2.0 slot. I wonder if having dual-core CPU has positive effect on speed in terms of reducing delay with all that switching you mentioned or additional core is only for USB 2.0 stuff.

 

[bill001g]

Nope all the work is done completely independent of the router firmware or cpu. It is all done inside the wifi chipset. The firmware that runs inside the wifi chips is generally provided by the chipset manufactures and the router manufacture does not change it to avoid having to get things recertified by the fcc. This was the very first thing I discovered when looking at third party firmware for routers. I wanted a simple how many retransmission where happening and found it is all locked up in a BIN file that there is no source for and your only option is to load it into the wifi chipset.

 

[Boris_yo]

Nope all the work is done completely independent of the router firmware or cpu. It is all done inside the wifi chipset.

Why they used dual-core processor then if a single-core processor can handle operations?

Well that's the reply from TP-Link I expected:

Later found this reply from their forum member saying it is lazy programming. Even higher range routers seem to have band restrictions:

 

[bill001g]

Mostly having multiple cores is a marketing thing. People always think more is better. Just like regular CPU it greatly depends on your application. Some applications can use multiple cores and other can not.

A router is similar. I mean some routers can act as a NAS and that takes CPU power and you would prefer if it did not impact your traffic. Multiple cores can help in that case. Things that require encryption or traffic filtering tend to be single thread application and what is most important it the clock rate of the cpu more than how many cpu.

To a point it doesn't matter. Pretty much the only thing a router is doing is keeping track and translating the internal IP to the external IP unless you configure other stuff. This is the NAT function. It is very cpu intensive because it must recalculate the error checksums in every packet header. They have moved this function off the cpu to a hardware accelerator so the cpu is no longer involved. This is how a extremely cheap router can pass 1gbit of traffic wan/lan.

Now all this is beside the point of response from tplink.

You need to think of a router as 3 CPUs. You have the main cpu chip and you have 2 wifi radio cpu chips. All the wifi function is done by these other CPU.

What tplink doesn't want to admit is they do not actually make chips they just repackage them and put their name on the box. The wifi chips come with software images from the chipset vendor. Just as they say it affects many routers. For whatever reason the wifi chipset vendor does not support certain radio bands. The maker of the router can not do much about this other than maybe use other radio chips. Although technically they could contract to have custom software written for the wifi chipset/cpu they don't want to deal with the FCC certification stuff.

Most people have no idea it works this way but once you play with third party firmware you will find there is not a lot you can change in a router even when you can write the firmware yourself. All the really interesting wifi things are locked away and you just get a file you load into the chip and tell the wifi chip to boot.