RT-AC66U: increased stability after replacing power supply

[TerminX]

I don't think this explicitly relates to the Merlin firmware, but that's what I'm running so I figured I'd mention it here. My RT-AC66Us have been unstable for months. Random reboots, incredible slowdowns to 1-2mbit/s every day (or even multiple times an hour if I was unlucky enough) requiring either a reboot or running restart_wireless over ssh, complete instability when trying to use CTF, etc. I didn't think my power adapters were faulty as I have 4 of these routers (only 2 in use currently) and the symptoms still happened regardless of which of the 4 adapters I tried.

Recently, I noticed that the replacement power supplies sold by Pwr+ on Amazon for the RT-AC66U were 19v 2.1a versus the 19v 1.58a of the stock charger, likely so Pwr+ could save on manufacturing costs and just produce one model that works with all the ASUS devices requiring 19v and the same power jack (which includes a few netbooks I believe). It occurred to me that my longstanding issues are very similar to what happens if you try to run a PC with an inadequate PSU (or if you're overclocking and bump the CPU voltage up to levels that your motherboard and PSU can't sustain) so I bought a couple of the 2.1a power supplies from Amazon and my issues seem to have gone away.

It was a last ditch effort before scrapping the RT-AC66Us entirely and I would be looking at spending several hundred bucks replacing my setup if I hadn't noticed this and given it a try. A previous effort to solve the problem included redoing the thermal paste and giving them active cooling via fans that run off USB, which lowered my temperatures on both the 2.4GHz and 5GHz side by ~10C each, but had no effect on the symptoms. However, the higher output power supplies seem to have actually made a difference. Even the stupid "received packet with own address as source address" log messages stopped, which I strongly suspect were correlated with the slowdowns I was experiencing!

Your mileage may vary, but anyone else who has a misbehaving RT-AC66U they feel like taking outside and shooting is definitely advised to give this a try before spending hundreds to solve the problem.

 

[TonyH]

I don't think this explicitly relates to the Merlin firmware, but that's what I'm running so I figured I'd mention it here. My RT-AC66Us have been unstable for months. Random reboots, incredible slowdowns to 1-2mbit/s every day (or even multiple times an hour if I was unlucky enough) requiring either a reboot or running restart_wireless over ssh, complete instability when trying to use CTF, etc. I didn't think my power adapters were faulty as I have 4 of these routers (only 2 in use currently) and the symptoms still happened regardless of which of the 4 adapters I tried.

Recently, I noticed that the replacement power supplies sold by Pwr+ on Amazon for the RT-AC66U were 19v 2.1a versus the 19v 1.58a of the stock charger, likely so Pwr+ could save on manufacturing costs and just produce one model that works with all the ASUS devices requiring 19v and the same power jack (which includes a few netbooks I believe). It occurred to me that my longstanding issues are very similar to what happens if you try to run a PC with an inadequate PSU (or if you're overclocking and bump the CPU voltage up to levels that your motherboard and PSU can't sustain) so I bought a couple of the 2.1a power supplies from Amazon and my issues seem to have gone away.

It was a last ditch effort before scrapping the RT-AC66Us entirely and I would be looking at spending several hundred bucks replacing my setup if I hadn't noticed this and given it a try. A previous effort to solve the problem included redoing the thermal paste and giving them active cooling via fans that run off USB, which lowered my temperatures on both the 2.4GHz and 5GHz side by ~10C each, but had no effect on the symptoms. However, the higher output power supplies seem to have actually made a difference. Even the stupid "received packet with own address as source address" log messages stopped, which I strongly suspect were correlated with the slowdowns I was experiencing!

Your mileage may vary, but anyone else who has a misbehaving RT-AC66U they feel like taking outside and shooting is definitely advised to give this a try before spending hundreds to solve the problem.

IMO, in that case, it should've been first ditch effort, not last. That is why I always connect my home network router
to a lab. grade DC source with V/A meters.

 

[TerminX]

Well, since it happens with all 4 of my ASUS branded power supplies, all of which test fine via multimeter, it doesn't seem likely that they're all defective straight out of the box. The first place you turn to after going through 4 OEM power supplies isn't usually a lower priced third party alternative.

IMO, the 1.58a supply the router comes with is underspecced, though I can't really say if that's because the hardware somehow requires a bit more current to operate correctly as it ages, or if new features or a newer Broadcom driver added in later firmwares causes the peak current draw to be higher than it was with the stock firmware when the router was released.

 

[System Error Message]

1.58A may be enough for a wifi router with USB3. You have the main platform itself which for MIPS uses up to 20W (spec of 15W + Wifi + switch) and than you have usb ports. USB 2.0 specifies 500mA at 5 V (2.5W) and USB 3 specifies 1500mA at 5V ( 7.5W). Adding both together you need up to 30W in total

Its a good thing you did replace the PSU as running a lower spec PSU can damage the system. Perhaps the output from the ASUS supplied PSU just isnt stable but it is good to actually measure the stability of a PSU output to avoid fakes. There was one instance of a women in australia using a fake plug to charge her ipad and it ended up killing her. its also possible for some dishonest suppliers to give you non genuine plugs that look like the original.

A PSU that is being overloaded isnt just unstable but would also be hot. You can also open the plug and check the quality of the work that is put into it.

 

[Per Salmi]

Very interesting, I will have to test this with my AC66U. Just this week it has started to reboot at random times, sometimes as often av 3-4 times an hour. Tested different firmwares without any effect regarding the problem. Strange as the router worked perfectly well until about a week ago. Is it possible that the power supply degrades slowly over time?

 

[Engineer]

Very interesting, I will have to test this with my AC66U. Just this week it has started to reboot at random times, sometimes as often av 3-4 times an hour. Tested different firmwares without any effect regarding the problem. Strange as the router worked perfectly well until about a week ago. Is it possible that the power supply degrades slowly over time?

Yes, it is very possible. Also, it could be a capacitor on the main board, especially if the capacitor has started leaking. What you describe sounds like a capacitor has started leaking (but that's just a guess based on previous experience).

 

[Per Salmi]

Yes, it is very possible. Also, it could be a capacitor on the main board, especially if the capacitor has started leaking. What you describe sounds like a capacitor has started leaking (but that's just a guess based on previous experience).

Ok, this issue with random and frequent reboots of the AC66U seems to be quite common...

 

[TerminX]

Since this thread was bumped, I just wanted to report back in and say that I'm at about the one week mark with no issues. Was even able to enable CTF and bump the CPU clock speed up to the 662MHz setting on both units!

For anyone curious, the power supply Pwr+ sent me for this is nothing like what it comes with. Instead of a wall wart it's a whole laptop-style power brick with a detachable, grounded power cord with "Mickey Mouse" style connector.

 

[Per Salmi]

Since this thread was bumped, I just wanted to report back in and say that I'm at about the one week mark with no issues. Was even able to enable CTF and bump the CPU clock speed up to the 662MHz setting on both units!

For anyone curious, the power supply Pwr+ sent me for this is nothing like what it comes with. Instead of a wall wart it's a whole laptop-style power brick with a detachable, grounded power cord with "Mickey Mouse" style connector.

What is CTF?

 

[mattiL]

https://pcdiy.asus.com/2014/08/august-2014-the-best-802-11ac-router-rt-ac87u-rt-ac87r/

 

[Per Salmi]

Ok, resolved my stability issues by first having the RT-AC66U replaced as it was still under warranty, the new one has a slightly different power supply (it still says 1.58A) and a higher hardware revision on the router hardware so some fixes might have been done. The new hardware worked better than the old one but not completely stable until I disabled the NAT acceleration under LAN->Switch Control. Now it has been running without any interruptions for almost 6 days. Very strange that the NAT acceleration had to be turned off as that setting worked well on the old router for almost 1.5 years until it suddenly started the random restart behavior.

 

[TerminX]

Ok, resolved my stability issues by first having the RT-AC66U replaced as it was still under warranty, the new one has a slightly different power supply (it still says 1.58A) and a higher hardware revision on the router hardware so some fixes might have been done. The new hardware worked better than the old one but not completely stable until I disabled the NAT acceleration under LAN->Switch Control. Now it has been running without any interruptions for almost 6 days. Very strange that the NAT acceleration had to be turned off as that setting worked well on the old router for almost 1.5 years until it suddenly started the random restart behavior.

Try the 2.1a power supply and see if your NAT acceleration issue goes away. NAT acceleration is CTF.

Both of my RT-AC66Us remain solid after 10 days of running at 662MHz with CTF enabled.

 

[Marko Polo]

and USB 3 specifies 1500mA at 5V ( 7.5W). Adding both together you need up to 30W in total

Well, are you sure? As stated in USB specification:

USB 3.0 ports may also implement other USB specifications for increased power, including the USB Battery Charging Specification for up to 1.5 A or 7.5 W, or, in the case of USB 3.1, the USB Power Delivery specification for charging the host device up to 100 W

The word "may" is the keyword here. Are you sure AC66/68/88 line provides 1500 mA to it USB3.0 port?
I am using 2.5 HDD (requires 700 mA) connected to USB3.0 and cannot figure out if the issues I observe are related to insufficient supply or smth other.
Have anybody ever measured these 3.0 ports with tool like this?

 

[Jack Yaz]

Holy thread necro batman!

 

[ColinTaylor]

The word "may" is the keyword here. Are you sure AC66/68/88 line provides 1500 mA to it USB3.0 port?
I am using 2.5 HDD (requires 700 mA) connected to USB3.0 and cannot figure out if the issues I observe are related to insufficient supply or smth other.

This post clarifies the "available" power also referred to here.

So to summarize; USB2 can supply 500mA and USB3 can supply 900mA. But it is the device itself that decides how much power it will draw and whether or not it is compliant with any of the standards. In other words, if you plug a simple lamp into a USB3 socket it can draw up to 900mA @ 5v.

 

[System Error Message]

Ok, resolved my stability issues by first having the RT-AC66U replaced as it was still under warranty, the new one has a slightly different power supply (it still says 1.58A) and a higher hardware revision on the router hardware so some fixes might have been done. The new hardware worked better than the old one but not completely stable until I disabled the NAT acceleration under LAN->Switch Control. Now it has been running without any interruptions for almost 6 days. Very strange that the NAT acceleration had to be turned off as that setting worked well on the old router for almost 1.5 years until it suddenly started the random restart behavior.

firmware/hardware bug.