ac66u signal strength and speed weaken over time

[jathor]

Long story short, ever since i bought the asus ac66u (about one year ago and had it replaced a month or so ago) I've had the same issue with every firmware version including merlins. After a couple days of excellent reception and speed it would:

1) signal strength would drop at least 10db (using inssider) and whatever device i have in the same place would either lose reception or the signal bar would drop a couple bars

2) speed would slow down quite a bit (YouTube videos would buffer many times, games would lag, ping was high etc etc.

Reboot solves the issue but comes back after a few days

I noticed this a while after i bought it (so i couldn't return it) and happened with every single update and 2 of merlins versions also. I sent it in to rma to asus and they sent me another but still the same issue.

I don't want to spend another couple hundred on another router as this one is pretty much new but I'm sick and tired of having to worry about rebooting every couple days.

Does this only happen to me?

 

[jms493]

U are not the only person. This router has been terrible so far. My apple router was much better. I wish i never got rid of it.

sent from my flip phone.

 

[jathor]

Anyone know of a solution or reason as to why this is happening?

 

[jerry6]

Anyone know of a solution or reason as to why this is happening?

Sure , send it back . I just got my replacement RMA AC 66 and connection rates are 60-130 mbps on one client 80- 120 on an AC client , with my N 66 it's 220 - 240 mbps and AC client 270 - 300 mbps . The AC stands for a b s o l u t e c r a p . Another thing is transfer rates have dropped by more than 3-7 MBPS used to get between 7 and 10 MBPS now lucky to get 768 KBPS on the same clients . Can no longer watch video with the AC 66 as is is always buffering , unreal what a pos they sent to me .
That and the package they sent it back in was below the standards they ask for when you send in the router .

Giving Asus repair depot a call Monday morning

 

[jathor]

I already did send it back and it's the same story with the "new" one they sent me...

 

[jerry6]

I'll keep sending mine back untill I get an AC 66 that works the way it should , at least have the same or better signal /transfer rate as an N 66 . MY RMA returned router is going back , no way an I accepting a router that cannot match an old early N router in performance .

 

[L&LD]

I'll keep sending mine back untill I get an AC 66 that works the way it should , at least have the same or better signal /transfer rate as an N 66 . MY RMA returned router is going back , no way an I accepting a router that cannot match an old early N router in performance .

I would not advise your strategy: a specific model has specific design parameters and hardware specifications that will give the same performance (within a fairly small range).

If the new device doesn't live up to your expectations (and it seems it doesn't), then either return/exchange it if you can or sell it outright and buy a model that works closer to your expectations.


The customers I have that use an RT-AC66U don't have any issues. But at the same time; they also upgraded from something much, much worse than the RT-N66U too.

 

[jerry6]

Right , only thing is I expect the AC 66 to perform AS WELL or better than an N 66 not 1/4 to 1/2 the connection rate and 1/4 the transfer rate . I was going to trade it in at Amazon , but they offer 50 cents for the AC 66 ,$52.00 for the N66 , guess I'll have to call Asus and return the unit they sent me . The AC 66 I sent back because it bricked was eaual or bettered my N66 most of the time

 

[Buck115]

I don’t have an AC66U, I have an RT-AC68U; so this might not be applicable. I’ve never had the issue you’re having on stock ASUS firmware, but I’ve experienced a similar problem on Merlin’s firmware (version .41 and .42).

I discovered when I perform a Site Survey, within 30 seconds the 5GHz band’s signal strength diminishes -10dBm from -49dBm at my test computer (ASUS PCE-AC68 wireless card) to -59dBm measured by inSSIDer. My first visual indication is the Windows Network Connection Status Icon in the Notification Area goes from a constant full 5 bars to a constant 4 bars. When I look at the RT-AC68U Wireless Log, RSSI is the same -58dBm, but RX/TX has gone from 1053/1053 to 1053/702. A Reboot from the GUI gets the 5GHz signal strength back to normal. 2.4GHz doesn’t seem to be affected. For me this is 100% repeatable.

 

[jerry6]

I don’t have an AC66U, I have an RT-AC68U; so this might not be applicable. I’ve never had the issue you’re having on stock ASUS firmware, but I’ve experienced a similar problem on Merlin’s firmware (version .41 and .42).

I discovered when I perform a Site Survey, within 30 seconds the 5GHz band’s signal strength diminishes -10dBm from -49dBm at my test computer (ASUS PCE-AC68 wireless card) to -59dBm measured by inSSIDer. My first visual indication is the Windows Network Connection Status Icon in the Notification Area goes from a constant full 5 bars to a constant 4 bars. When I look at the RT-AC68U Wireless Log, RSSI is the same -58dBm, but RX/TX has gone from 1053/1053 to 1053/702. A Reboot from the GUI gets the 5GHz signal strength back to normal. 2.4GHz doesn’t seem to be affected. For me this is 100% repeatable.

Should not make a difference using Merlin or Stock FW , they both use the same drivers , my problems persist using stock and or Merlin FW

 

[Buck115]

Should not make a difference using Merlin or Stock FW , they both use the same drivers , my problems persist using stock and or Merlin FW

No, it shouldn't. However, the Site Survey process isn't a feature in the stock ASUS firmware, only Merlin's as far as I know. That's were I find the signal strength issue when a Site Survey is performed. The OP's problem is so similar I thought it was worth mentioning.

 

[RMerlin]

No, it shouldn't. However, the Site Survey process isn't a feature in the stock ASUS firmware, only Merlin's as far as I know. That's were I find the signal strength issue when a Site Survey is performed. The OP's problem is so similar I thought it was worth mentioning.

It's possible that the site survey might have an impact on the radio performance.

The site survey code comes from stock firmware, however Asus only uses it for configuring the router as a repeater or media bridge (this is the code they use to scan for available APs to pair with).

Out of curiosity, can you reproduce the same issue with both active and passive scans, or just one of the two?

 

[Buck115]

It's possible that the site survey might have an impact on the radio performance.

The site survey code comes from stock firmware, however Asus only uses it for configuring the router as a repeater or media bridge (this is the code they use to scan for available APs to pair with).

Out of curiosity, can you reproduce the same issue with both active and passive scans, or just one of the two?

I didn't try passive. I've reinstalled Asus 5656, but tomorrow I'll reinstall .42 and try passive scans and report the results here in this thread. It has been 100% repeatable for me with active scans (default).

 

[Buck115]

Out of curiosity, can you reproduce the same issue with both active and passive scans, or just one of the two?

Merlin, I flashed your .42 firmware last night and performed the test you requested. Unfortunately, it appears there’s no way for me to test to see if a passive Site Survey scan causes the same 10dBm reduction in 5GHz signal strength (-49dBm to -59dBm) as an active scan. The reasons are:

When you enter the Site Survey page a scan runs automatically, so the first time you enter that page it runs an active (default) scan, which in my case causes the 10dBm reduction in 5GHz signal strength. At that point you can select passive and perform a rescan, but the reduction has already taken place and no matter how many active or passive scans you run, it neither restores the 10dBm loss nor does it make it worse. The only way to restore the 5GHz signal strength is to reboot the router, which resets the Site Survey page to active (default) again even though you’d set it to passive. So, once again, when you enter the Site Survey page the first time after a reboot, an active scan runs and the 5GHz signal strength is reduced by 10dBm once again; a vicious circle.

I tried to trick it by setting the Site Survey to passive, saving that configuration, rebooting to restore 5GHz signal strength, and reloading the saved configuration before entering the Site Survey page after the reboot. However, it appears that the passive setting isn’t saved, so when I entered the Site Survey page after restoring the configuration, it was set to active (default) and an active scan ran, which again resulted in a 10dBm reduction in 5GHz signal strength.

I noticed that there is no “Apply” button at the bottom of the Site Survey page to apply the change from active to passive scan. However, I found that if I change from active to passive and perform a rescan it seems to establish the setting as passive and from that point forward you can enter the Site Survey screen and passive scan is the retained setting. But, again, unless I reboot, the 10dBm loss of 5GHz signal strength isn’t restored.

When I originally observed this anomaly on your .41 firmware, I waited around 24 hours to see if it would correct itself, but it didn’t. I don’t know if waiting any longer would help.

I still have .42 flashed, so if you have any other suggestions I’d be glad to try them.

 

[RMerlin]

I noticed that there is no “Apply” button at the bottom of the Site Survey page to apply the change from active to passive scan. However, I found that if I change from active to passive and perform a rescan it seems to establish the setting as passive and from that point forward you can enter the Site Survey screen and passive scan is the retained setting. But, again, unless I reboot, the 10dBm loss of 5GHz signal strength isn’t restored.

That should still be the case after a reboot. If I recall correctly, I do save the last used scan mode to nvram.

 

[Buck115]

That should still be the case after a reboot. If I recall correctly, I do save the last used scan mode to nvram.

Unfortunately, after a reboot, even if passive scan has been established, it reverts to active.

 

[RMerlin]

Unfortunately, after a reboot, even if passive scan has been established, it reverts to active.

Odd. Try manually setting the variable through ssh:

nvram set wlc_scan_mode=1
nvram commit

You could try setting it right after a reboot. Once it's set, then access the webui and try a site survey.

 

[Buck115]

Merlin, I've never performed the ssh procedures, but I'll read back through some of the threads that discuss it and give it a try; thanks!

 

[Buck115]

Merlin, I found your tutorial on how to use ssh, downloaded Xshell4, installed it, and ran the commands you provided. It did indeed set the Site Survey to passive, so when I entered the Site Survey page a passive scan was executed. Unfortunately, with the same results; within 30 seconds I observed a 10dBm decrease in 5GHz signal strength at my test computer.

For those that are experiencing a decrease in 5GHz signal strength with your firmware, I wonder if this is the culprit?

 

[RMerlin]

Merlin, I found your tutorial on how to use ssh, downloaded Xshell4, installed it, and ran the commands you provided. It did indeed set the Site Survey to passive, so when I entered the Site Survey page a passive scan was executed. Unfortunately, with the same results; within 30 seconds I observed a 10dBm decrease in 5GHz signal strength at my test computer.

For those that are experiencing a decrease in 5GHz signal strength with your firmware, I wonder if this is the culprit?

Random throughput/range decreases over time have been reported a couple of times over the months, and they seemed more frequent with the RT-AC66U than the other models, for some odd reason. The issue predates by addition of the Site Survey feature, and was also reported with the stock firmware.

It's still odd than conducting a scan would cause the output to become affected until the radio gets restarted. Only Broadcom would know why.