Wi-Fi Signal Strength RT-AC86U v RT-AX88U

[brummygit]

I've arrived here researching poor 2.4Ghz signal on my new RT-AX88U bought to replace my RT-AC68U. I needed more ethernet ports, memory and CPU power and wanted to future proof a little so went with this model even though I don't have any AX clients, however I'm having problems due to the range of the 2.4Ghz signal from the AX88U.

I normally run RMerlin's software but am currently on stock firmware version 3.0.0.4.384_9107 as I wanted to try the new QoS categories. I believe I am seeing the same issues being reported by many in this thread, and already commented on another thread before finding this one. I haven't yet had the chance to install an old firmware prior to the Broadcom driver changes but do plan to do this at some point when I can persuade the family to let me.

In contradiction to some of the focus of contributors in this thread, I am not interested in the throughput I am getting, my problem is 2.4Ghz signal strength and therefore range. I have lost reliable connectivity to some of my 2.4Ghz only devices. The signal is actually lower than my 5Ghz signal from the AX88U. In the following screen grab from InSSIDer taken on my laptop in my living room, the MAC addresses starting AC are my AC68U covering the opposite end of my home- you can see that I get a marginally stronger 2.4Ghz signal from it as would be expected, but in the case of my AX88U (MAC starting A8) the 5Ghz network is 15dBm stronger than 2.4Ghz.

This causes me real issues with my more remote 2.4Ghz devices. I am not trying to hold out in 1997 as has been suggested, but I have devices such as a new Google Nest Cam IQ Outdoor that is 2.4Ghz only and is now struggling to remain a reliable connection having previously been reliable with my old router in the same position. The signal is just too weak for this and a number of other more remote devices. By contrast my Ring Pro doorbell uses 5Ghz and has a rock solid connection in a similar position.

Thanks to many on this thread who have been trying to diagnose this issue - I've raised a support request with Asus in the hope that more reports prompt some action.

 

[brummygit]

Both Tim's lab tests and L&LD's experience make it very unlikely that the AX88 is deficient. The remaining conclusion is that your example is defective.

Those tests were measuring throughput rather than signal strength so not particularly relevant. They also pre-date the firmware which is is suspected to have introduced the issue.

 

[brummygit]

@sturmstar if you haven't used a new SSID when testing new routers, that may one reason why the old one seems so much better.

I see this stated quite often and the implication is that the client device stores and uses properties of the target network to reconnect. I understand this for the security elements such as WPA type and pre-shared key, but I believe that the connection characteristics are negotiated at the time of connection.

Can anyone point me to sources which confirm the original point as this would be quite fundamental when building large corporate multi-radio networks (I've been involved in building over 2,000 APs in a single campus using common SSIDs), and would impact ISPs such as BT UK who run hotspots from multiple generations of ISP router, and organisations that provide WiFi in their premises across many sites (eg Starbucks)?

 

[brummygit]

I've been able to perform some more tests this evening and the results are interesting.

I was using my AX88U in it's normal position with the AiMesh node removed. For each test I fully reset back to factory defaults, reconfigured manually and then set it to manual channel allocations with 2.4Ghz at 20Mhz on channel 6, 5Ghz at 20/40/80Mhz on channel 100. I tested using an AC equipped laptop using a mix of ethernet, 2.4Ghz and 5Ghz connections.

Using Asus firmware 384_6210 and 384_6436 I could see that when I was on Ethernet the 2.4Ghz signal was better as per the attached . When I'm connected using 2.4Ghz I still see similar signal strengths but connecting with 5Ghz this is effectively reversed

Using 384_7756 and above the results have changed. When using Ethernet the 2.4Ghz signal is noticeably weaker but if I connect using 2.4Ghz the signal improves to previous levels and when I'm using 5Ghz it's the 5Ghz that improves. The same is seen with the latest 384_9107.

It seems that the new wireless drivers introduced with 384_7756 have indeed changed the resting 2.4Ghz signal strength, however once I established a connection with either the 2.4Ghz or 5Ghz radio the signal on that band is optimised to give acceptable levels, but the other is reduced. In fact the 5Ghz signal in more recent firmware versions is very good.

I also compared with a BT WiFi disk in the same spot as the router and the Asus 2.4Ghz is equal at it's best, but falls back in it's weaker states, but the Asus 5Ghz is better.

 

[Kees17760]

It seems that the new wireless drivers introduced with 384_7756 have indeed changed the resting 2.4Ghz signal strength, however once I established a connection with either the 2.4Ghz or 5Ghz radio the signal on that band is optimised to give acceptable levels, but the other is reduced. In fact the 5Ghz signal in more recent firmware versions is very good.

I also compared with a BT WiFi disk in the same spot as the router and the Asus 2.4Ghz is equal at it's best, but falls back in it's weaker states, but the Asus 5Ghz is better.

See if i get this right.
So with no clients connected at a given band, be it 2,4 or 5GHz, the Asus somehow dims it's radio as in some sort of energy saving mode.
That is not what i encountered, mainly because i always have a few 2,4 GHz clients up. That would also mean that when none of my 2,4 GHZ clients would be up, the Asus's signal (being already low) would definitely sink into nothing.

In a way this makes sence, but on the other hand a potential culprit is introduced, because:

- When in dimmed mode, how will clients know the Asus is still there?
- And when they do, will the Asus "awake" and raise it's level fast enough?
- Is it just an energy save mode, or a protective cooling measure?
- Is there a way to circumvent or bypass this mode?

 

[GarrHaal]

I did actually disable 5ghz band on ax88u and it made no difference in signal strength on my phone or laptop. I went out to test it too. And the tx power adjustment setting didn't actually do anything either.

 

[brummygit]

See if i get this right.
So with no clients connected at a given band, be it 2,4 or 5GHz, the Asus somehow dims it's radio as in some sort of energy saving mode.
That is not what i encountered, mainly because i always have a few 2,4 GHz clients up. That would also mean that when none of my 2,4 GHZ clients would be up, the Asus's signal (being already low) would definitely sink into nothing.

In a way this makes sence, but on the other hand a potential culprit is introduced, because:

- When in dimmed mode, how will clients know the Asus is still there?
- And when they do, will the Asus "awake" and raise it's level fast enough?
- Is it just an energy save mode, or a protective cooling measure?
- Is there a way to circumvent or bypass this mode?

There may have been some other clients attached during my tests as I used my normal SSID but this is what a typical client would experience on my network. There’s never a time when nothing is connected.

I think it’s a huge jump from seeing different signal levels when connected or not to concluding it’s due to power management. My own guess is more on antenna management, beam forming or similar but I don’t have an explanation.

What I’m seeing is that on older firmware the 2.4Ghz is strong when I’m looking at broadcasting networks I’m not connected to, but later firmware broadcasts a lower signal level which then increases when I’m connected. Maybe the single antenna discussed at length on this thread is used for 2.4Ghz broadcast purposes and then multiple antennas manage the connection once established. I really don’t know.

I do think that the WiFi driver update in 384_7756 improved 5Ghz performance at the expense of 2.4Ghz somehow.

 

[ollimoe]

@brummygit - I can confirm all your observations, this discussion has been going on for some time unfortunately ending up in philosophical debates. I also remember some forum members being in touch with ASUS on this problem but I have not seen any progress. As I can confirm the same observation since Merlin 384.14 it clearly indicates a problem on ASUS or Broadcom - going back to 384.13 my 2.4 is at an acceptable level compared to my previous 86U, 5 Ghz similar quality. Going to 384.14 and beyond I clearly see a significant drop on 2.4 while I see an increas on signal strength on 5 Ghz. And yes, fixed channels, no beamforming, no move of the HW etc ...

 

[Ninko]

ASUS got back to me, saying:

"Thank you for your email.

This has not been duplicated, though we do not have the exact devices you mention to test

But due to the poor connection quality, we suspect it would be an issue with the 2,4GHz band/antenna

To arrange a return/RMA of the component, please contact your seller to arrange the return for repair, replacement or refund. Please note that the return can be arranged via the distribution chain only. As ASUS does not sell any products directly to end users we cannot be seen as the seller.

Our apologies for the inconveniences caused to you.


Kind Regards,
Peter S."

I'm not sure what's best to do now, get it replaced or not? I'm just not sure it's a hardware problem.

 

[Kees17760]

ASUS got back to me, saying:

"Thank you for your email.

This has not been duplicated, though we do not have the exact devices you mention to test

But due to the poor connection quality, we suspect it would be an issue with the 2,4GHz band/antenna

To arrange a return/RMA of the component, please contact your seller to arrange the return for repair, replacement or refund. Please note that the return can be arranged via the distribution chain only. As ASUS does not sell any products directly to end users we cannot be seen as the seller.

Our apologies for the inconveniences caused to you.


Kind Regards,
Peter S."

I'm not sure what's best to do now, get it replaced or not? I'm just not sure it's a hardware problem.

Why not send it in. It doesn't hurt to try. Also read previous posts from user Garrhaal

 

[GarrHaal]

ASUS got back to me, saying:

"Thank you for your email.

This has not been duplicated, though we do not have the exact devices you mention to test

But due to the poor connection quality, we suspect it would be an issue with the 2,4GHz band/antenna

To arrange a return/RMA of the component, please contact your seller to arrange the return for repair, replacement or refund. Please note that the return can be arranged via the distribution chain only. As ASUS does not sell any products directly to end users we cannot be seen as the seller.

Our apologies for the inconveniences caused to you.


Kind Regards,
Peter S."

I'm not sure what's best to do now, get it replaced or not? I'm just not sure it's a hardware problem.

Thats almost the same answer I got. I sent it in for repair.

 

[Ninko]

@GarrHaal Was it any better after sending it back for repair?

 

[GarrHaal]

@GarrHaal Was it any better after sending it back for repair?

Sent it last week back, 2-3 weeks estimated repair.

 

[ollimoe]

I'd be surprised if the problem is solved through HW repair as there were many folks here in the forum reporting similar issues after FW upgrade. In any way, please report back to see if/what may have changed! Good luck!

 

[KrispiL]

I recently upgraded my main AC68U with an AX88U and noticed a drop in 2.4Ghz signal strength with the new router. I've read through this thread which has been really helpful in at least confirming I'm not the only one seeing this.

However, even though the signal strength is lower, I've done some transfer rate tests using iperf v2 and I still get better throughput with the AX88U compared to the AC68U using the default setup on each and latest stock firmware.

For example, testing using a Surface Pro (2017 model) from my office which is the room furthest from the router.

OFFICE: AX88U 2.4Ghz: Signal Strength: -66 dBm Transfer Rate: 41.50 Mbits/sec
OFFICE: AC68U 2.4Ghz: Signal Strength: -56 dBm Transfer Rate: 28.70 Mbits/sec

I tried a different device, a Mac Book Pro (2013 model):

OFFICE: AX88U 2.4Ghz: Signal Strength: -75 dBm Transfer Rate: 41.40 Mbits/sec
OFFICE: AC68U 2.4Ghz: Signal Strength: -70 dBm Transfer Rate: 38.60 Mbits/sec

I ran the tests in several different rooms and got similar results, 2.4Ghz signal strength between 5/10 dBm worse on AX88U but throughput the same or better.

For comparison, 5Ghz gives similar signal strengths but slightly better throughput on the AX88U in the same room:

OFFICE: AX88U 5Ghz: Signal Strength: -74 dBm Transfer Rate: 47.90 Mbits/sec
OFFICE: AC68U 5Ghz: Signal Strength: -75 dBm Transfer Rate: 41.10 Mbits/sec

I'm not sure what's going on with the signal strength and I admit my tests were not done in lab conditions, but overall the AX88U performs better than the AC68U it replaced.
I can post the rest of the results from different rooms if folks are interested. I'm now using the AC68U as a repeater for the guest network.
(Note: first post but long time lurker in these forums since I got a N66U back in 2013)

 

[GarrHaal]

Tiny update on ax88u, they are sending a new unit. Should arrive on tuesday.

 

[Ninko]

Unless they've changed any of the hardware, i.e. hardware version, I can't see there be any difference sadly. That's why I never bothered to get mine replaced, I'm still deciding whether or not to get a refund on it and buying something else.

 

[GarrHaal]

So got my replacement AX88U rev A1.1 made in china with firmware 3.0.0.4.384
Updated to newest firmware and can say it is waaaay better now, no problem reaching out to the garage with 2,4ghz now. Seems like it was a faulty unit.

 

[Ninko]

Was the one you had before the same hardware version, i.e. A1.1?

 

[GarrHaal]

Was the one you had before the same hardware version, i.e. A1.1?

Yes it was