Asus AC68U vs Asus AC88U range, temperature and stability

[Wutikorn]

I'm going to buy a router in a month, and I am choosing between Asus AC68U and Asus AC88U. I currently have Buffalo WZR-HP-AG300H, and I found out that its range is just there, but it is not stable. I just want stable connection at 20Mbps while I have wireless N devices. My internet speed is 20Mbps/7Mbps. I will have WD DL2100/ 1-2 torrenting devices/ a 720P Wireless IP Camera and 5-30 web-surfing devices. I will use Ai-Protection, adaptive QOS and OpenVPN(Max speed is 7Mbps due to my Internet speed). I am concern about heat, the temperature can goes up to 35-40C in the room. The router will be put in a place that nothings are around. I am also concern about stability. I have to restart my Buffalo router weekly in order for it to work properly, but I don't want to do that in the new router. And I wonder if I should use Asus Merlin firmware or not. So which router should I buy?

 

[pete y testing]

I currently have Buffalo WZR-HP-AG300H, and I found out that its range is just there

thats cause its a relic from the past and any wireless AC router will perform better coverage wise as long as the environmental conditions and construction materials of the location allow it

I am concern about heat, the temperature can goes up to 35-40C in the room.

thats getting close to max heat any device will work in , do you not have air conditioning where you are ?

I am also concern about stability. I have to restart my Buffalo router weekly in order for it to work properly,

thats prob cause you killing it with torrent connections and again a decent wireless AC router will do far better

And I wonder if I should use Asus Merlin firmware or not.

ether stock or merlins wont matter as far as that goes

So which router should I buy?

so if your main concern is coverage then there is little between them

see my post in the link below

http://www.snbforums.com/threads/wireless-ac-5-gig-coverage-per-class-comparison.32362/

but the issue of coverage is a 2 way street , you also need to consider what wireless adapter you are using in the client device as well

but between the 68u and the 88u coverage wise at 1/2 the price the asus rt-ac68u should do you fine

 

[Wutikorn]

thats cause its a relic from the past and any wireless AC router will perform better coverage wise as long as the environmental conditions and construction materials of the location allow it



thats getting close to max heat any device will work in , do you not have air conditioning where you are ?



thats prob cause you killing it with torrent connections and again a decent wireless AC router will do far better



ether stock or merlins wont matter as far as that goes



so if your main concern is coverage then there is little between them

see my post in the link below

http://www.snbforums.com/threads/wireless-ac-5-gig-coverage-per-class-comparison.32362/

but the issue of coverage is a 2 way street , you also need to consider what wireless adapter you are using in the client device as well

but between the 68u and the 88u coverage wise at 1/2 the price the asus rt-ac68u should do you fine

Thanks for all you information and link, I wish I had all those routers so that I could test range at around 40 meters. I probably buy Asus AC68U as it's not that different. Thanks again

 

[Wutikorn]

I just thought of one more difference. Will 256MB RAM matter to me at all?

 

[pete y testing]

I probably buy Asus AC68U as it's not that different.

except it cost twice as much

so that I could test range at around 40 meters.

reality 40 meters indoors is just not possible unless its open plan , if your house or dwelling is that big and has lots of walls i suggest you look at having ethernet run from the router to close to the 40 meter distance and run a second wireless access point there

reason being even if you could get any router to work 40 meters away through a few walls etc the connection would be so low and slow it would be painful at best , far better to run wifi close to where its needed to get better connection and better speed and better throughput

 

[pete y testing]

I just thought of one more difference. Will 256MB RAM matter to me at all?

the more ram the harder the router can work and handle the load but you would be pushing it hard indeed to bother 256M ramk and 128M is just fine for most power users

 

[Wutikorn]

except it cost twice as much



reality 40 meters indoors is just not possible unless its open plan , if your house or dwelling is that big and has lots of walls i suggest you look at having ethernet run from the router to close to the 40 meter distance and run a second wireless access point there

reason being even if you could get any router to work 40 meters away through a few walls etc the connection would be so low and slow it would be painful at best , far better to run wifi close to where its needed to get better connection and better speed and better throughput

Didn't see that in the test, there were a few walls. So 25 meters is about what I expected.

the more ram the harder the router can work and handle the load but you would be pushing it hard indeed to bother 256M ramk and 128M is just fine for most power users

ok, thanks

 

[pete y testing]

Didn't see that in the test, there were a few walls. So 25 meters is about what I expected.

consider this at 10 feet in the same room im getting close tp 110 MB/s read speeds , at 25 meters through a few walls im getting 36 MB/s , so a loss of around 66% , now take that out to 40 meters and it would be more like 10 MB/s and would certainly be having issues with noise and other wifi interference

rssi wise anything over -70db and you get into the area where the client devices will start looking for other connection because the signal is considered average , at 40 meters you would be well in-80 to -90db rssi

 

[Wutikorn]

I already ordered Asus AC68U. Will come here in two days. I didn't see part of your comment at first. I have do not have air conditioning in that room, but I will keep it where it should be between 30-35C instead of 35-40C.

 

[Deleted member 22229]

Even at 30-35 c the router is going to run very hot. You most likely will need some kind of active cooling.

 

[Wutikorn]

Even at 30-35 c the router is going to run very hot. You most likely will need some kind of active cooling.

I still need a fan? What is really maximum temperature I can put my router at without putting a fan?

 

[pete y testing]

I still need a fan? What is really maximum temperature I can put my router at without putting a fan?

anything below 30 deg C should be ok , anything above and i would certainly consider some form of active cooling

even if the router worked ok in those heat conditions you would certainly reduce its life span of the router considerably

 

[Deleted member 22229]

I still need a fan? What is really maximum temperature I can put my router at without putting a fan?

I believe Merlin mentioned at one time if the router is 90c or above that's getting to hot. That's 194F. 100c, 212F and water boils.

 

[pete y testing]

I believe Merlin mentioned at one time if the router is 90c or above that's getting to hot.

thats the cpu heat sink interface he was talking about , we are talking ambient temperature where the router lives

 

[Deleted member 22229]

I have owned a few 68's and if his room temp is 30-35 c I would bet the router will be at or very close to 90 c.

 

[sfx2000]

I've been watching different threads in passing regarding Asus CPU/Core Temps - I've always seen them unexpectedly/abnormally high, too high for the type of Cores/Process nodes and the workloads...

I'm wondering if either Broadcom (SDK) or Asus is calculating them correctly... anybody care to dig into the code and see?

 

[Deleted member 22229]

Well I can say this. Netgear R7000 basically the same hardware as the 68 Asus routers and they run much cooler on the CPU and the radio temps. This was measured using the vortex/Merlin fork. So same software different router.

There have been people who have taken the 68 apart and see the heat sinks are not fit properly and don't make good contact with the parts they are designed to cool.

 

[RMerlin]

I'm wondering if either Broadcom (SDK) or Asus is calculating them correctly... anybody care to dig into the code and see?

Without access to the SDK, it's impossible to tell if the formula is accurate or not. Wireless value is derived from a formula used by Asus, and CPU temperature is read through DMU. Based on how the temperature evolves with newer HW revision, they look plausible to me.

 

[sfx2000]

CPU temp just doesn't seem right... 90C is pretty darn hot, and over an extended period of time, the housing on a passively cooled device would be more than just "warm to the touch", it would be very uncomfortably hot...

As a point of reference - the Intel C2358 dual core CPU in my pfSense router rarely gets above 42C, and it's a tiny passive heatsink on it. Yes, I know, smaller geometry node, but it's clocked at 1.7GHz, and hella traffic through it, so I'm not sure if Asus is working the numbers correctly...

 

[RMerlin]

Replace that with a cheap heatsink, and put a thermal pad instead of thermal compound. Put it slightly askew, and watch your CPU temperature sky rocket...

The BCM4709 isn't "just" a CPU, it also contains other blocks such as the BCM5301x. the USB controler, etc... It's a complete SoC, you can't compare that with a plain CPU.

Based on the temperature you can feel by touching the case, I'd say the temperature looks right to me. And people's results when they just replace the thermal pad with proper compound showed a temperature drop totally in line with what I would expect.