RT-AC68U 5GHz RF stopped working. Weird CPU temp.

[Hale_JP]

RT-AC68U 5GHz started disappearing after a few hours after every reboot.
Stops both broadcasting SSID and receiving.

I have this router for almost two years. It was working fine until yesterday, when it started.
2.4GHz works fine as before.

There is nothing special running on the router. No torrent, no USB disks now. Just QoS, which I disabled without success.

The weird thing is temperature:
Legend: 2.4 GHz - 5 GHz - CPU
Current Temperatures: 53 °C - 55 °C - 73 °C
While CPU load is close to zero:
Core 1 = 2%
Core 2 = 2%
RAM = 27%
(Merlin was always installed on this router, but I did not update or touch it for w while. The current ver. is 384.13)

Looks like some hardware problem.
Did anyone had the same issue, or do know how to fix it?

 

[Hale_JP]

I guess, I know what triggered the problem.
Two months ago I moved to new larger apartments, so the working laptop moved behind two cement walls.
The router should have started tuning itself to higher power.

I disassembled it and found that amplifiers there have no thermal pads, soldered just inside isolated boxes.
It is possible, the 5G amp got burned out. Maybe.

=============

That's it.

U just put a cube of high-end thermal plaster on every amplifier(?), and 5GHz has stabilized a bit. (I guess these small chips between antenna splitter, and radio are either amps, or switches, or combined)

Still, the CPU temp is too high, don't know why, I replaced that plaster as well.

When 68U's radio, 2.4GHz, or 5Ghz starts disappearing, that means the amp has boiled itself out.
Very bad power and thermal design.

 

[ColinTaylor]

The weird thing is temperature:
Legend: 2.4 GHz - 5 GHz - CPU
Current Temperatures: 53 °C - 55 °C - 73 °C

Those temperatures for that model are perfectly normal. There are many, many posts comparing temperatures in these forums. If anything your CPU is running cooler than most. My typical temperatures are as follows and have been running like that or hotter for years.

52°C 54°C 86°C

 

[OzarkEdge]

Did anyone had the same issue, or do know how to fix it?

Flash it with stock Asuswrt, reset with Initialize from the webUI, configure it minimally from scratch, and if the WiFi is not present and stable, replace it.

OE

 

[Hale_JP]

Please, don't post Microsoft's canned prescripts HERE. It is not a S&M forum.

For reasonable people, I repeat. The problem is in overheating RF frontend. The temp, monitored by Merlin is not related to the problem. CPU and both RF MACs have basic 2-2.5mm soft heat-taking pads. Other RF chips don't. I didn't measure which is overheating, just placed a 3mm high-performance thermal pad on each, and the router has stabilized a bit.

Probably not completely, I just run it for 3 days. But it got much better.

I recommend preparing 1, 2, and 3.5 mm non-conducting thermal pads, and replace/add everything there. And apply wide 1 mm pads to heatsinks; original pads are to small.

//Why this forum got so sluggish? I need 2 minutes just to list the thread. Connecting from Japan using NTT's fiber.

 

[OzarkEdge]

Hey Mr., you are a provocateur,
I guess, I ruined your monopoly on solutions of the "last hope" to RF failing in RT-AC8X routers. I must say, it is a pretty common thing , placing heatpads and radiators on overheating and failing RF components.

I didn't say that every AC router is affected.
But upon my investigation, I found that it is the most common problem of RT-AC8X boards. And on my impression, suggestion of yours NEVER worked, on no "helped" reports were ever posted.

Because it is a hardware problem, don't you see? When either 2.4GHz, or 5GHz start disappearing without any manipulations done; while 2.5GHz and 5GHz MAC/Phy's are so different, from different makers, don't you find it suspicious?
The frontend Is the same! There's a Wilkinson, and some square chips, which I didn(t try to identify. But these can be some last stage amps, switches, matching IC's or less probable, delay lines (the first chip in a pair). These components can go bad and overheat. Especially when there is NO air convection under the shield.

Now, TADA!, the photo I found later, in a "disassembly" article.
https://www.smallnetbuilder.com/ima...u/asus_rtac68u_board_component_no_shields.jpg
See that little blue brick, the small one?
It is a thermal pad placed probably by ASUS QC, or RMA department.

No, speaking about a "solution". On my experience, some chip gone bad, and it will not help for long. Frequently overheating RF components can go bad irrevertibly.



//Why this happen?
A wild guess. this is a hardware timebomb placed by sneaky designers. They have statistics on component life under pour cooling conditions. They use it to phase consumer goods out of the market. Philips does the same, but not this dirty. They place grease bombs in remote controls, so after 5-6 years PCB starts emitting grease telling that the TV got very old, and it is a time to buy a new one,

I'm not debating the nature of the hardware failure or your attribution of nefarious intent by Asus. I only suggest that given a hardware failure, get new hardware. Your proposed solution to modify the hardware is unproven and doubtful and not useful to anyone in general and will most likely not fix your hardware failure that is rarely experienced by other AC68U owners. So, your post is not useful and made less so by your intent to assign blame for a hardware failure that is most likely blameless in the scheme of consumer electronics. You got almost two years of service for your money... average that into your overall consumer electronics cost and you are doing just fine in the year 2020. If you disagree, then blame yourself for not choosing how to spend your money more wisely in the first place. Once you buy the router, it's your responsibility after the two year warranty expires... you did try to get an warranty RMA before you opened it up and played with it, yes?

OE