• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Asus router making squeaking/high freq buzzing when stressed

Joined
Dec 13, 2008
Messages
354 (0.06/day)
System Name The Browser
Processor AMD FX-8350
Motherboard Asus 90 Aura Gaming
Cooling Noctua D14
Memory Crucial Balistix Slim 16gb
Video Card(s) Sapphire AMD 7950 3GB
Storage Crucial MX 240gb + 2X3TB WD Reds
Display(s) Dell 2515H 25"
Case Corsair Carbide 540 (piece of shit case)
Power Supply Seasonic G-650W Gold
Mouse Corsair M65 RGB
Keyboard Corsair K70
Software Win 10
Hi there !
So my nice asus AC1900 (AC68U) router started making buzzing noises from next to the rear plugs... Nothing fixed, just sounds like a nuclear radiation meter from Chenobyl videos. Pchhh Phchhhhhh Pshhhhhhttt.
Only happens when the Wifi is used.

Is it on its way out ? Because if it's a matter of time, I'd get a new one shipped over here so I don't get stuck in some weekend when it breaks.
 

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
Coil whine, it happens, sadly.
The reason it happens when Wi-Fi is used, is because the router draws more power when Wi-Fi is in use, it's the most power hungry part of a router.
As the coils are part of the power conversion circuitry, they get loaded and as they're imperfect components, you get whine due to internal vibrations.
Not much you can do about it.
It's highly unlikely that it'll affect your router in any way.
 
Joined
Dec 13, 2008
Messages
354 (0.06/day)
System Name The Browser
Processor AMD FX-8350
Motherboard Asus 90 Aura Gaming
Cooling Noctua D14
Memory Crucial Balistix Slim 16gb
Video Card(s) Sapphire AMD 7950 3GB
Storage Crucial MX 240gb + 2X3TB WD Reds
Display(s) Dell 2515H 25"
Case Corsair Carbide 540 (piece of shit case)
Power Supply Seasonic G-650W Gold
Mouse Corsair M65 RGB
Keyboard Corsair K70
Software Win 10
I see, so that router can hold many more years in theory ?
The thing is, it did not happen when it was new..... That's why I'm pissed about it :)
 

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
I see, so that router can hold many more years in theory ?
The thing is, it did not happen when it was new..... That's why I'm pissed about it :)
It can happen to pretty much anything that has coils in it, like power supplies, graphics cards, motherboards and obviously, routers...
I can't say how long your router will last, but it might be worth updating the firmware if you haven't done so, as if that would for some reason change something with the Wi-Fi, it might change the power draw and the whine might change, but it's a long-shot.

A PSU in my PC started to whine many moons ago and this was in a case that had a top mounted PSU. One day I got so pissed off with the noise, as it was really high-pitched, that I thumped the top of the computer so hard that the fan fell out the bottom of the PSU...
Don't do that...
 
Last edited:
Joined
Jul 25, 2006
Messages
13,411 (1.99/day)
Location
Nebraska, USA
System Name Brightworks Systems BWS-6 E-IV
Processor Intel Core i5-6600 @ 3.9GHz
Motherboard Gigabyte GA-Z170-HD3 Rev 1.0
Cooling Quality case, 2 x Fractal Design 140mm fans, stock CPU HSF
Memory 32GB (4 x 8GB) DDR4 3000 Corsair Vengeance
Video Card(s) EVGA GEForce GTX 1050Ti 4Gb GDDR5
Storage Samsung 850 Pro 256GB SSD, Samsung 860 Evo 500GB SSD
Display(s) Samsung S24E650BW LED x 2
Case Fractal Design Define R4
Power Supply EVGA Supernova 550W G2 Gold
Mouse Logitech M190
Keyboard Microsoft Wireless Comfort 5050
Software W10 Pro 64-bit
I agree with TheLostSwede in that this sounds like it could be coil whine. But personally, I have never heard coil whine from a low voltage/low wattage device like a home router that was so loud, I could hear it without putting my ear right over the device.

Geiger counters click. Coil whine typically is pretty high pitched. Did you check to make sure the cables are securely connected?

If it really is coil (or transformer) whine, it is impossible to tell if it will fail completely or last forever. If me, I would open the case and see if I could pinpoint the source - perhaps using a paper towel tube to my ear. This whine is usually caused by the coiled wires (or transformer plates) vibrating. If the offending device can be determined, a small dab of epoxy resin can often stop the vibration. Just note the cure time of the epoxy and give it time to thoroughly cure before using the device.
 
Joined
Jun 18, 2010
Messages
2,342 (0.44/day)
Processor Intel i7 970 // Intel i7 2600K
Motherboard Asus Rampage III Formula // Asus P8P67 Deluxe
Cooling Zalman CNPS9900MaxB // Zalman CNPS11X
Memory GSkill 2133 12GB // Corsair V 2400 32GB
Video Card(s) ASUS GTX1080 // MSI GTX1070
Storage Samsung 870EVO // Samsung 840P
Display(s) HP w2207h
Case CoolerMaster Stacker 830se // Lian Li PC-9F
Audio Device(s) onboard
Power Supply Seasonic X 850w Gold // EVGA 850w G2
Mouse Logitech G502SE HERO, G9
Keyboard Dell
Software W10 Pro 22H2
Joined
Nov 18, 2010
Messages
7,607 (1.47/day)
Location
Rīga, Latvia
System Name HELLSTAR
Processor AMD RYZEN 9 5950X
Motherboard ASUS Strix X570-E
Cooling 2x 360 + 280 rads. 3x Gentle Typhoons, 3x Phanteks T30, 2x TT T140 . EK-Quantum Momentum Monoblock.
Memory 4x8GB G.SKILL Trident Z RGB F4-4133C19D-16GTZR 14-16-12-30-44
Video Card(s) Sapphire Pulse RX 7900XTX. Water block. Crossflashed.
Storage Optane 900P[Fedora] + WD BLACK SN850X 4TB + 750 EVO 500GB + 1TB 980PRO+SN560 1TB(W11)
Display(s) Philips PHL BDM3270 + Acer XV242Y
Case Lian Li O11 Dynamic EVO
Audio Device(s) SMSL RAW-MDA1 DAC
Power Supply Fractal Design Newton R3 1000W
Mouse Razer Basilisk
Keyboard Razer BlackWidow V3 - Yellow Switch
Software FEDORA 41
Not only poor quality coils do whine, but also MLCC caps... and that's my first culprit there. If the PCB design engineer has no experience, it will end up like that. You have to poke around putting those.


I can hear various switching devices even low power ones, Well I guess I ain't deaf yet... worst came out from monitor LED/CCFL drivers. Why? Because they are cheapskates and do cheap designs, the parts are way pushed, inductors are saturated on load and it cause harmonics go wild across all ranges.. Also some phone chargers emit huge audible noise, well I can forgive them partly as there's no space.

Also if there is a SMPS then whole lot of rubbish is going on. The discontinuous mode SMPS generate noise depending on the load in kHz range and is greatly annoying, there are hybrid designs that during constant current draw change the charge rate of the main cap to evade generation of noise, but it costs money, and seldom who notices it, lets put a cheap shit buck converter also somehere for minor rails. Some even may use a feedback cap to the mains side and you still have dominant 100/120Hz hum. Then only the switching noise... and then the unorthodox ground pollution noise that creeps up from nowhere because how imperfect our life is.

Pretty much you cannot do anything about. Asus being shit maker recently.

You need to open the case and look.
See this: https://console5.com/wiki/Identifying_Bad_Capacitors
They are not bad tho... it is by design as their qualities overpass this slight weakness.
 
Joined
Jul 25, 2006
Messages
13,411 (1.99/day)
Location
Nebraska, USA
System Name Brightworks Systems BWS-6 E-IV
Processor Intel Core i5-6600 @ 3.9GHz
Motherboard Gigabyte GA-Z170-HD3 Rev 1.0
Cooling Quality case, 2 x Fractal Design 140mm fans, stock CPU HSF
Memory 32GB (4 x 8GB) DDR4 3000 Corsair Vengeance
Video Card(s) EVGA GEForce GTX 1050Ti 4Gb GDDR5
Storage Samsung 850 Pro 256GB SSD, Samsung 860 Evo 500GB SSD
Display(s) Samsung S24E650BW LED x 2
Case Fractal Design Define R4
Power Supply EVGA Supernova 550W G2 Gold
Mouse Logitech M190
Keyboard Microsoft Wireless Comfort 5050
Software W10 Pro 64-bit

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
It sounds like it could be a bad capacitor as well.

You need to open the case and look.

See this: https://console5.com/wiki/Identifying_Bad_Capacitors


Very unlikely, as the sound is only when he's using the Wi-Fi. This is exactly what happened when I worked at Securifi and we were developing the Almond+. They had to change the power design and use a different kind of coils/inductors to get rid of the whine. Apparently I was the only one that noticed...

Not only poor quality coils do whine, but also MLCC caps... and that's my first culprit there. If the PCB design engineer has no experience, it will end up like that. You have to poke around putting those.


I can hear various switching devices even low power ones, Well I guess I ain't deaf yet... worst came out from monitor LED/CCFL drivers. Why? Because they are cheapskates and do cheap designs, the parts are way pushed, inductors are saturated on load and it cause harmonics go wild across all ranges.. Also some phone chargers emit huge audible noise, well I can forgive them partly as there's no space.

Also if there is a SMPS then whole lot of rubbish is going on. The discontinuous mode SMPS generate noise depending on the load in kHz range and is greatly annoying, there are hybrid designs that during constant current draw change the charge rate of the main cap to evade generation of noise, but it costs money, and seldom who notices it, lets put a cheap shit buck converter also somehere for minor rails. Some even may use a feedback cap to the mains side and you still have dominant 100/120Hz hum. Then only the switching noise... and then the unorthodox ground pollution noise that creeps up from nowhere because how imperfect our life is.

Pretty much you cannot do anything about. Asus being shit maker recently.


They are not bad tho... it is by design as their qualities overpass this slight weakness.
That might be the case, but if you look at the PCB, this router the OP owns, has few regular coils/inductors as part of the power design and I believe that's what's causing the sound. This is due to the PA's on the Wi-Fi end drawing a lot of power when they're in use, which in turn means more power goes through the coils. Can't say I see any ceramic caps in that design.



Also note that this is a very old router design by now, which has been very popular and successful for Asus, so no need going around calling them names. In fact, I'm connected to a noise free one right now.
 
Last edited:
Joined
Nov 18, 2010
Messages
7,607 (1.47/day)
Location
Rīga, Latvia
System Name HELLSTAR
Processor AMD RYZEN 9 5950X
Motherboard ASUS Strix X570-E
Cooling 2x 360 + 280 rads. 3x Gentle Typhoons, 3x Phanteks T30, 2x TT T140 . EK-Quantum Momentum Monoblock.
Memory 4x8GB G.SKILL Trident Z RGB F4-4133C19D-16GTZR 14-16-12-30-44
Video Card(s) Sapphire Pulse RX 7900XTX. Water block. Crossflashed.
Storage Optane 900P[Fedora] + WD BLACK SN850X 4TB + 750 EVO 500GB + 1TB 980PRO+SN560 1TB(W11)
Display(s) Philips PHL BDM3270 + Acer XV242Y
Case Lian Li O11 Dynamic EVO
Audio Device(s) SMSL RAW-MDA1 DAC
Power Supply Fractal Design Newton R3 1000W
Mouse Razer Basilisk
Keyboard Razer BlackWidow V3 - Yellow Switch
Software FEDORA 41
Yeah, they can whine too - though typically not a ticking sound.

Measure wireless radio part peak currents and see the like packet behavior. It may sound like morse code doing radio traffic. There could be a choir of them and thus have baritone 1uF ones while the contra-tenor nanofarads do the high parts...

Also note that this is a very old router design by now, which has been very popular and successful for Asus, so no need going around calling them names. In fact, I'm connected to a noise free one right now.

They are not flowing intel's USB3 errata on 2.4GHz data, that's first, ASUS being shit again. My router has isolated traces with a shield can coming way down from the CPU. I cannot see what kind of step down converter they are using there but it should operate at ~2MHz range... but prolly the large sized ceramics are at fault there not the inductor, there are not such great powers to saturate it, to have it larger. Also there is kinda an unwanted style of sticking up two caps besides faking a CLC filter with the short trace in between... it is a known practice for motherboard and GPU VRM's, to kill HF radio noise as the CPU is a HF sensitive device, but as a fact it will cause only such unwanted effects, as we all know VRM's do whine and doesn't help with audible spectrum our ears are sensitive with. The caps are not Muratas for sure also and because of the budget they used some china OEM's and here we are. There is also funny that they are having a single chemi-con solid state, other is not placed, the other black ones are hybrids. The blue one for one of the step down converters, they had something fishy there already, they fought some anomalies and later gimped the design.
 
Last edited:
Joined
Jul 25, 2006
Messages
13,411 (1.99/day)
Location
Nebraska, USA
System Name Brightworks Systems BWS-6 E-IV
Processor Intel Core i5-6600 @ 3.9GHz
Motherboard Gigabyte GA-Z170-HD3 Rev 1.0
Cooling Quality case, 2 x Fractal Design 140mm fans, stock CPU HSF
Memory 32GB (4 x 8GB) DDR4 3000 Corsair Vengeance
Video Card(s) EVGA GEForce GTX 1050Ti 4Gb GDDR5
Storage Samsung 850 Pro 256GB SSD, Samsung 860 Evo 500GB SSD
Display(s) Samsung S24E650BW LED x 2
Case Fractal Design Define R4
Power Supply EVGA Supernova 550W G2 Gold
Mouse Logitech M190
Keyboard Microsoft Wireless Comfort 5050
Software W10 Pro 64-bit
Very unlikely, as the sound is only when he's using the Wi-Fi
That would not matter. The WAP (wireless access point) is a totally discrete network device, just as the router and the 4-port Ethernet switch are also totally discrete network devices. They just happen to share a common circuit board, power supply and associated power circuits and a shared case.

It is the same as an audio "receiver". The receiver is really a tuner, preamp and amplifier - 3 discrete device in one box sharing a common power supply.

The point is, when using wifi, those components unique to the WAP are then used. When not using wifi, they are idle. If the offending device is part of the common power circuits, additional load would be imposed when the wifi circuits are in use. And that additional load may cause the device to start singing.
 

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
They are not flowing intel's USB3 errata on 2.4GHz data, that's first, ASUS being shit again. My router has isolated traces with a shield can coming way down from the CPU. I cannot see what kind of step down converter they are using there but it should operate at ~2MHz range... but prolly the large sized ceramics are at fault there not the inductor, there are not such great powers to saturate it, to have it larger. Also there is kinda an unwanted style of sticking up two caps besides faking a CLC filter with the short trace in between... it is a known practice for motherboard and GPU VRM's, to kill HF radio noise as the CPU is a HF sensitive device, but as a fact it will cause only such unwanted effects, as we all know VRM's do whine and doesn't help with audible spectrum our ears are sensitive with. The caps are not Muratas for sure also and because of the budget they used some china OEM's and here we are. There is also funny that they are having a single chemi-con solid state, other is not placed, the other black ones are hybrids. The blue one for one of the step down converters, they had something fishy there already, they fought some anomalies and later gimped the design.
A lot of companies don't follow that... Neither did we at first when I was at Securifi, one of many issues I found as I was testing the early hardware. That part is actually quite dependent on the USB 3.x cables as well, but I found that even some expensive cables cause interference, not just cheap ones. One of the worst cables, was the one that at the time shipped with an external drive for WD. It would reduce the 2.4GHz Wi-Fi signal to a trickle and it killed the ZigBee signal. Takes a little bit of time to design out things like that, but it's actually not that hard.

Shielding isn't needed if the design is done right, but it's the easy route to take. We tried all kinds of shielding with the early design, it was simply a no go, as the USB 3.0 board traces were too long. In the end, the USB 3.0 host controller was moved to solve the problem.

Have you actually looked at how much power is being drawn when the PA's in a high-end Wi-Fi solution kicks in? The reason why most routers have massive heatsinks today, is not because of the SoC or the Wi-Fi chips, but because of the PAs. They're tiny little chips that draws an insane amount of power for their size and output a ton of heat. It's the single most power hungry and heat producing component in a modern Wi-Fi router.

I would almost bet money on it being the coils in this instance, since it's only happening when the Wi-Fi is being used. As I said, I have first hand experience of this exact same issue, but obviously in a different product.

You are aware that Murata actually just brand a lot of their products, right? Not everything from them is the best out there. Also, you don't need top grade components for everything electronic, although I agree that a lot of stuff out of China might not be an ideal choice. However, there's a lot of middle ground, where you can still make a good product without it being all Japanese brand components.

We're also way off topic here now.
 
Joined
Jul 25, 2006
Messages
13,411 (1.99/day)
Location
Nebraska, USA
System Name Brightworks Systems BWS-6 E-IV
Processor Intel Core i5-6600 @ 3.9GHz
Motherboard Gigabyte GA-Z170-HD3 Rev 1.0
Cooling Quality case, 2 x Fractal Design 140mm fans, stock CPU HSF
Memory 32GB (4 x 8GB) DDR4 3000 Corsair Vengeance
Video Card(s) EVGA GEForce GTX 1050Ti 4Gb GDDR5
Storage Samsung 850 Pro 256GB SSD, Samsung 860 Evo 500GB SSD
Display(s) Samsung S24E650BW LED x 2
Case Fractal Design Define R4
Power Supply EVGA Supernova 550W G2 Gold
Mouse Logitech M190
Keyboard Microsoft Wireless Comfort 5050
Software W10 Pro 64-bit
Shielding isn't needed if the design is done right
Ummm proper shielding is part of a done right design. This is especially true with RF circuits and where high speed switching/digital devices are used as those devices generate and are susceptible to RFI. And you are absolutely right about "quality" cables. Sadly, expensive does not always equal quality. In many cases, you are just paying for the brand. Some of the poorest made cables I have seen were expensive "Monster" brand cables. Quality cables tend to be "shielded" and, most importantly, the wires properly terminated in the connectors - quality connectors at that.
 
Joined
Dec 13, 2008
Messages
354 (0.06/day)
System Name The Browser
Processor AMD FX-8350
Motherboard Asus 90 Aura Gaming
Cooling Noctua D14
Memory Crucial Balistix Slim 16gb
Video Card(s) Sapphire AMD 7950 3GB
Storage Crucial MX 240gb + 2X3TB WD Reds
Display(s) Dell 2515H 25"
Case Corsair Carbide 540 (piece of shit case)
Power Supply Seasonic G-650W Gold
Mouse Corsair M65 RGB
Keyboard Corsair K70
Software Win 10
Very cool picture of the insides of my router there ! That's why I love coming to this place, a lot of actual professionals come here !


I can confirm that the noise is indeed coming from the area above the power connector, and indeed I see capacitors there ?

Another description of the sound would be like how a bad Lithium Polymer battery sounds. Irregular, high pitched, didn't-happen-when-new.....

I am about to order a set of two new Asusi because the noise just pisses me off at the this point. it's very loud and distracting .

Should I stick to these mid range Asus or is there something better made for the price ?
This particular Asus did serve for 5 years 24/7 now so I can't be too angry with it... I think T-Mobile also contracted with them to sell this model with new subscribers ? It came in a hundred different revisions all looking the same.
 
Last edited:
Joined
Jun 1, 2011
Messages
4,699 (0.95/day)
Location
in a van down by the river
Processor faster at instructions than yours
Motherboard more nurturing than yours
Cooling frostier than yours
Memory superior scheduling & haphazardly entry than yours
Video Card(s) better rasterization than yours
Storage more ample than yours
Display(s) increased pixels than yours
Case fancier than yours
Audio Device(s) further audible than yours
Power Supply additional amps x volts than yours
Mouse without as much gnawing as yours
Keyboard less clicky than yours
VR HMD not as odd looking as yours
Software extra mushier than yours
Benchmark Scores up yours
Hi there !
So my nice asus AC1900 (AC68U) router started making buzzing noises from next to the rear plugs... Nothing fixed, just sounds like a nuclear radiation meter from Chenobyl videos. Pchhh Phchhhhhh Pshhhhhhttt.
Only happens when the Wifi is used.

Is it on its way out ? Because if it's a matter of time, I'd get a new one shipped over here so I don't get stuck in some weekend when it breaks.
is the power cable completely in the outlet/power strip and device? No damage to it?
 

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
Very cool picture of the insides of my router there ! That's why I love coming to this place, a lot of actual professionals come here !


I can confirm that the noise is indeed coming from the area above the power connector, and indeed I see capacitors there ?

Another description of the sound would be like how a bad Lithium Polymer battery sounds. Irregular, high pitched, didn't-happen-when-new.....

I am about to order a set of two new Asusi because the noise just pisses me off at the this point. it's very loud and distracting .

Should I stick to these mid range Asus or is there something better made for the price ?
This particular Asus did serve for 5 years 24/7 now so I can't be too angry with it... I think T-Mobile also contracted with them to sell this model with new subscribers ? It came in a hundred different revisions all looking the same.
If you're looking at a new router, I would consider the Netgear R7800/X4S with Voxel's firmware. It's really great hardware in it and Voxel & Co are developing some really cool features for, if that's your kind of thing. Admittedly, anything supported by Merlin would be nearly as good, although, personally I do prefer QCA to BCM when it comes to the hardware side of things.

The R7800 is admittedly quite an old router by now, so you might want to consider something else simply because of that. Stay away from 802.11ax for now, as it's not ready for mainstream usage as yet, imho.

As an affordable router, the RT-AC68x range isn't bad, but it's not great either. The one I'm using at the moment (as I'm stuck abroad due to the Wuhan virus) has pretty poor 5GHz range.

Not sure what else you're looking for in a router in terms of features.
 
Last edited:
Joined
Nov 18, 2010
Messages
7,607 (1.47/day)
Location
Rīga, Latvia
System Name HELLSTAR
Processor AMD RYZEN 9 5950X
Motherboard ASUS Strix X570-E
Cooling 2x 360 + 280 rads. 3x Gentle Typhoons, 3x Phanteks T30, 2x TT T140 . EK-Quantum Momentum Monoblock.
Memory 4x8GB G.SKILL Trident Z RGB F4-4133C19D-16GTZR 14-16-12-30-44
Video Card(s) Sapphire Pulse RX 7900XTX. Water block. Crossflashed.
Storage Optane 900P[Fedora] + WD BLACK SN850X 4TB + 750 EVO 500GB + 1TB 980PRO+SN560 1TB(W11)
Display(s) Philips PHL BDM3270 + Acer XV242Y
Case Lian Li O11 Dynamic EVO
Audio Device(s) SMSL RAW-MDA1 DAC
Power Supply Fractal Design Newton R3 1000W
Mouse Razer Basilisk
Keyboard Razer BlackWidow V3 - Yellow Switch
Software FEDORA 41
Have you actually looked at how much power is being drawn when the PA's in a high-end Wi-Fi solution kicks in? The reason why most routers have massive heatsinks today, is not because of the SoC or the Wi-Fi chips, but because of the PAs. They're tiny little chips that draws an insane amount of power for their size and output a ton of heat. It's the single most power hungry and heat producing component in a modern Wi-Fi router.
I would almost bet money on it being the coils in this instance, since it's only happening when the Wi-Fi is being used. As I said, I have first hand experience of this exact same issue, but obviously in a different product.

Ye, I did look up the usual designs for RF power supply(albeit now modern, but low cost so it should fit), they are limited regards to the power limits, at max... already allmost illegal 200mW radio power, so lets tripple, like Mikrotik - 3 channels = 600mW usually at 3.3V put on some heat/efficiency loss, the reg must output around a 1A, okay let it be 1.5A, those coils there are around 3A, why so big?, to have the resistance parameter that meets up the specs, the driver datasheet dictates it, there is no need for large inductivity Henries.

Well... mobiles are much better in this state, they usually have integrated power management unit - PMU, that supplies CPU, RAM, etc a tailored perfect device, thus it works like a tank and are very sophisticated. Have we endured problems with iphones or qualcomm made CPU's? They are blazing fast and efficient and usually they are first. Yet a simple router often does funky things? Some may say because of the AC power source, but phones while being charged still do work fine... because the PMU costs an arm and leg and is sold as a set with the CPU, not few cent(almost wrote JayzTwoPesos, oh I did it) designs we look at now.

Regarding the Muratas... You got it right... Makers usually like Ti or Linear Tech, Maxim etc do examples, what kind of parts to use. They urge to use Class 1 MLCC, usually Muratas...as their datasheets do provide great information, oh well, I got respect for people and parts from Japan, there is still a reason for it.

I suspect... these are gimped ones - Class 2 devices there in caps. So due to poor materials, they exhibit the classic problem most don't know... when DC biased, those devices lose capacity and in magnitudes, in this situation tho it could be that the inductivity could start to act funny as the whole driver becomes unstable and may not even be driven at the desired resonant frequency, but that's not because of the the coil, the caps act weird, and that's only characteristic to class 2 and 3 MLCC materials. Also makers rarely do provide the DC biased capacity curve, also the AC one too. You take it out and measure, and nothing, seems good, like I ordered it. Very hideous things... The more the MLCC capacity, the more prone they are. I still support African slavery and genocide indirectly using Tantalum caps in my designs if I lack space, they have other bad things, but at least I know the beast I must fight on. It depends on the experience... and target BOM cost.

Some may argue why to gimp? Well... if you make few dozen of the devices no, if the device cost differs like 10$ if you do cheaper parts on mass production order, if you make a batch like for a 10 000 units... (depends on the small part roll size, if you do tings, minimum order is a roll, use it or not, that's reason many parts emerge on aliexpress, those are leftovers and sometimes legit) turns out to be 100 000$ savings. Nothing to sneeze at.
 
Joined
Dec 13, 2008
Messages
354 (0.06/day)
System Name The Browser
Processor AMD FX-8350
Motherboard Asus 90 Aura Gaming
Cooling Noctua D14
Memory Crucial Balistix Slim 16gb
Video Card(s) Sapphire AMD 7950 3GB
Storage Crucial MX 240gb + 2X3TB WD Reds
Display(s) Dell 2515H 25"
Case Corsair Carbide 540 (piece of shit case)
Power Supply Seasonic G-650W Gold
Mouse Corsair M65 RGB
Keyboard Corsair K70
Software Win 10
Where I am in Israel, the price of two Asus 68's is 220$.
That is also the price of a SINGLE netgear...
Knowing that, would you still go for the nighthawk ?

They also have D-Link and all the other Links but they look meh... No firmwares available.


And also an unrelated question while I have your attention ! What a nice DSL Modem I can use with these ? Or is the ISP modem usually enough ?


*very interesing Ferrum Master !

Is it true that QC chips are all backdoored to 3 letter agencies ? I think it was Snowden who said that ?
 
Joined
Jun 18, 2010
Messages
2,342 (0.44/day)
Processor Intel i7 970 // Intel i7 2600K
Motherboard Asus Rampage III Formula // Asus P8P67 Deluxe
Cooling Zalman CNPS9900MaxB // Zalman CNPS11X
Memory GSkill 2133 12GB // Corsair V 2400 32GB
Video Card(s) ASUS GTX1080 // MSI GTX1070
Storage Samsung 870EVO // Samsung 840P
Display(s) HP w2207h
Case CoolerMaster Stacker 830se // Lian Li PC-9F
Audio Device(s) onboard
Power Supply Seasonic X 850w Gold // EVGA 850w G2
Mouse Logitech G502SE HERO, G9
Keyboard Dell
Software W10 Pro 22H2
@Fif23

Take a good hard look at the ASUS AX58U also known as the AX3000.

Supported by Merlin Firmware.
 

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
Ye, I did look up the usual designs for RF power supply(albeit now modern, but low cost so it should fit), they are limited regards to the power limits, at max... already allmost illegal 200mW radio power, so lets tripple, like Mikrotik - 3 channels = 600mW usually at 3.3V put on some heat/efficiency loss, the reg must output around a 1A, okay let it be 1.5A, those coils there are around 3A, why so big?, to have the resistance parameter that meets up the specs, the driver datasheet dictates it, there is no need for large inductivity Henries.

Well... mobiles are much better in this state, they usually have integrated power management unit - PMU, that supplies CPU, RAM, etc a tailored perfect device, thus it works like a tank and are very sophisticated. Have we endured problems with iphones or qualcomm made CPU's? They are blazing fast and efficient and usually they are first. Yet a simple router often does funky things? Some may say because of the AC power source, but phones while being charged still do work fine... because the PMU costs an arm and leg and is sold as a set with the CPU, not few cent(almost wrote JayzTwoPesos, oh I did it) designs we look at now.

Regarding the Muratas... You got it right... Makers usually like Ti or Linear Tech, Maxim etc do examples, what kind of parts to use. They urge to use Class 1 MLCC, usually Muratas...as their datasheets do provide great information, oh well, I got respect for people and parts from Japan, there is still a reason for it.

I suspect... these are gimped ones - Class 2 devices there in caps. So due to poor materials, they exhibit the classic problem most don't know... when DC biased, those devices lose capacity and in magnitudes, in this situation tho it could be that the inductivity could start to act funny as the whole driver becomes unstable and may not even be driven at the desired resonant frequency, but that's not because of the the coil, the caps act weird, and that's only characteristic to class 2 and 3 MLCC materials. Also makers rarely do provide the DC biased capacity curve, also the AC one too. You take it out and measure, and nothing, seems good, like I ordered it. Very hideous things... The more the MLCC capacity, the more prone they are. I still support African slavery and genocide indirectly using Tantalum caps in my designs if I lack space, they have other bad things, but at least I know the beast I must fight on. It depends on the experience... and target BOM cost.

Some may argue why to gimp? Well... if you make few dozen of the devices no, if the device cost differs like 10$ if you do cheaper parts on mass production order, if you make a batch like for a 10 000 units... (depends on the small part roll size, if you do tings, minimum order is a roll, use it or not, that's reason many parts emerge on aliexpress, those are leftovers and sometimes legit) turns out to be 100 000$ savings. Nothing to sneeze at.
That's transmit power, which by the way is no longer 200mW, you need to catch up on about 5 years of changes in Wi-Fi products. 1W is allowed in the US for the 5GHz band, with other countries having followed suite.
As such, you have 3W of transmit power and most PA/LNA's are 5V or 3.7V, not 3.3V. Below is an example of a fairly typical 5GHz PA/LNA.
Trust me, this isn't as simple as you're making it out to be, I've been involved in the development of a couple of routers and I really know what I'm talking about here when it comes to power draw of what appears to be really small components.
Apologies for the blurry picture, but it's the only one I got laying around, since most of the pictures I took working there, is company property.

IMAG0047.jpg

Due to the six PA/LNAs at the bottom of that board, we had implement a heatpipe cooler, or the router would simply stop working after a couple of hours due to overheating.
IMAG0048.jpg

The back of the PCB is pretty much acting as a heatsink for the PA/LNAs.

Some pictures of the final PCB here.

Wi-Fi routers might look simple, but they take a lot more thought to do the PCB design for than your average ARM based board.

There's no reason you couldn't implement a PMU in a router and in fact, QCA has done that for some of their router SoCs.
A PMU isn't solving all the problems you're talking about though and phones are far from perfect. Plenty of phones overheat, explode or do other weird things. I think you're overestimating the quality of electronics in general. PMUs are not expensive at all, but they're in general tailed to one, or a few SoCs, which means if it's not a mass produced and mass sold solution, then yes, the PMU would be costly. There afaik, no universal PMUs. Sure, they cost more than a very basic power conversion circuitry, that's true, but they're not more than a buck or two.

Not saying Murata is bad, just a bit overrated at times. I mean, they sell a lot of Bluetooth modules, that use BCM or now Cypress parts inside them. Yes, Murata makes some special packaging etc. but it's still the same basic hardware inside.

You also forget a few things, you might've learned how to do this one way in school, with a more western mindset where cost isn't an issue. Anything made in Asia, is made with a very different mindset. First of all, a lot of the engineers only do what they've learnt in school and what their boss tells them. Secondly, it's all about cost, so they try to find the most cost efficient solution. Then there's usually someone else (sometimes without the original company's knowledge) that further cuts the cost down, to save a cent here or there.

This allows you and me to buy low cost electronics, but it also means that after a few years, it'll fail and end up in a landfill. Then again, I guess that's true for most electronics these days, as everything old has to be replaced with something new as fast as possible, even if we're seeing diminishing returns from the new product. I'm by no means against new stuff, but it feels that so many things are rushed these days, just so there's something new every year at a minimum. I feel we need to do more interoperability testing if nothing else of a lot of technology before it launches, especially things like network hardware. I also feel that all the device makers should be forced to issue regular software updates for at least two years for products that are connected to the internet, as far too many products receive zero updates and are in fact a security liability for the owner. This would if nothing else, reduce the amount of crap products in the market.

Personally I can't say I think Asus makes bad routers, at least not compared to the likes of D-Link, TP-Link, etc.
I'm surprised that my current Netgear router has been so good, as they make a lot of meh products too. I guess sometimes you're just lucky.
 
Last edited:
Joined
Nov 18, 2010
Messages
7,607 (1.47/day)
Location
Rīga, Latvia
System Name HELLSTAR
Processor AMD RYZEN 9 5950X
Motherboard ASUS Strix X570-E
Cooling 2x 360 + 280 rads. 3x Gentle Typhoons, 3x Phanteks T30, 2x TT T140 . EK-Quantum Momentum Monoblock.
Memory 4x8GB G.SKILL Trident Z RGB F4-4133C19D-16GTZR 14-16-12-30-44
Video Card(s) Sapphire Pulse RX 7900XTX. Water block. Crossflashed.
Storage Optane 900P[Fedora] + WD BLACK SN850X 4TB + 750 EVO 500GB + 1TB 980PRO+SN560 1TB(W11)
Display(s) Philips PHL BDM3270 + Acer XV242Y
Case Lian Li O11 Dynamic EVO
Audio Device(s) SMSL RAW-MDA1 DAC
Power Supply Fractal Design Newton R3 1000W
Mouse Razer Basilisk
Keyboard Razer BlackWidow V3 - Yellow Switch
Software FEDORA 41
Is it true that QC chips are all backdoored to 3 letter agencies ? I think it was Snowden who said that ?
I have no information about QC chips. You know why they are not used in routers that often? Because you have to have client base and net worth in dozen millions in order they speak with you and make an offer. The competition becomes crippled because of that actually... and here Mr. China comes in... they allow it to anyone, then who's fault it is? If you would have a backdoor, wouldn't you allow it to anyone?

Qualcomm has a subsidiary company named Atheros for their cheap designs. But their development documentation comes all and only in mandarin... think about it, is not only black and white... or 3 letter agencies... it is often about the highest bidder and best offer and that's it. So in the end.. who's side it is?

Especially you being in Israel known to be an exceptional country in the whole world whose economic is fed by military development at a large percent(at least that's what I was thought in my local school when I was 16), I kinda understand why the question. Because of the limited info during dev phase... the CPU's are so mysterious and undocumented often. The OEM's just implement the features they need according to the RAW datasheet during the beta silicon phase. You may make a device, debug it for a month... it doesn't work... and it ends up being a CPU bug... new revision comes, and everything works... you wasted sweat for a month and everyone thought you are an amateur... (it wasn't my experience, a university friend). So... there are BUGS... any bug can be turned out as a weapon. If those were intentional or not... we don't know... but perfect device does not exist... and exploits will be there ever... even when AI supervisors are being implemented. But... the Snowden thing... kinda lacks body in my perspective. You could hire a team of talented hackers and have those triggers, exploits, but the manufactures ain't to blame there... at least that's my point of view.
 

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
Where I am in Israel, the price of two Asus 68's is 220$.
That is also the price of a SINGLE netgear...
Knowing that, would you still go for the nighthawk ?

They also have D-Link and all the other Links but they look meh... No firmwares available.


And also an unrelated question while I have your attention ! What a nice DSL Modem I can use with these ? Or is the ISP modem usually enough ?


*very interesing Ferrum Master !

Is it true that QC chips are all backdoored to 3 letter agencies ? I think it was Snowden who said that ?
Right, yes, that's the problem, the R7800 is generally only affordable in the US.

Stay away from D-Link, they're terrible with regards to software updates. Linksys can be ok, but it varies a lot between models.
There's a reason I suggest models supported by Voxel or Merlin, as you'll get good, patched software from them, without having to use something like DD-WRT or OpenWRT.

The ISP modem should be fine in most cases. My ISP at home provides a combo modem/router that I had them put in bridge mode, as I don't want to use their crappy router. As I live in a three story house, with only a couple of rooms on each floor, but with concrete and metal in the floors, I use a couple of cheap TP-Link devices as Wi-Fi APs on those floors. As those devices aren't connected directly to the internet, I'm ok with using them, but I wouldn't use their products as routers.

As for products with backdoors, who knows, too many rumours. Besides, who wants to watch us hang out here and talk tech?

I have no information about QC chips. You know why they are not used in routers that often? Because you have to have client base and net worth in dozen millions in order they speak with you and make an offer. The competition becomes crippled because of that actually... and here Mr. China comes in... they allow it to anyone, then who's fault it is? If you would have a backdoor, wouldn't you allow it to anyone?
This is not true at all. QCA (Qualcomm Atheros) is a separate business unit and they're happy to talk to you if you want to make 10k routers. The reason they're not in that many routers, is because Broadcom is cheaper, offers better support and have more local offices with FAE staff. As such, they've become the dominant high-end router chip maker, with MTK taking the lower-end of the market. Realtek, Marvell, Intel and QCA have been sort of minor players and it's gone so far that both Marvell and Intel has sold of much of their router business.

Qualcomm has a subsidiary company named Atheros for their cheap designs. But their development documentation comes all and only in mandarin... think about it, is not only black and white... or 3 letter agencies... it is often about the highest bidder and best offer and that's it. So in the end.. who's side it is?

Sorry, but this is wrong too. Atheros was a company that Qualcomm bought for their Wi-Fi know-how. Those guys are pretty much the ones running QCA today, which I presume is the subsidiary you're talking about. It's still part of Qualcomm and it's not a division for cheap designs, but rather for router SoCs, router and IOT Wi-Fi products, etc. i.e. for solutions that aren't for phones.
 
Last edited:
Joined
Nov 18, 2010
Messages
7,607 (1.47/day)
Location
Rīga, Latvia
System Name HELLSTAR
Processor AMD RYZEN 9 5950X
Motherboard ASUS Strix X570-E
Cooling 2x 360 + 280 rads. 3x Gentle Typhoons, 3x Phanteks T30, 2x TT T140 . EK-Quantum Momentum Monoblock.
Memory 4x8GB G.SKILL Trident Z RGB F4-4133C19D-16GTZR 14-16-12-30-44
Video Card(s) Sapphire Pulse RX 7900XTX. Water block. Crossflashed.
Storage Optane 900P[Fedora] + WD BLACK SN850X 4TB + 750 EVO 500GB + 1TB 980PRO+SN560 1TB(W11)
Display(s) Philips PHL BDM3270 + Acer XV242Y
Case Lian Li O11 Dynamic EVO
Audio Device(s) SMSL RAW-MDA1 DAC
Power Supply Fractal Design Newton R3 1000W
Mouse Razer Basilisk
Keyboard Razer BlackWidow V3 - Yellow Switch
Software FEDORA 41
That's transmit power, which by the way is no longer 200mW, you need to catch up on about 5 years of changes in Wi-Fi products. 1W is allowed in the US for the 5GHz band, with other countries having followed suite.

My eyes Swede... such a blurry pic... I took the math because it is an old router. There are even illegal and old 5W devices. But this is an Asus...

I kinda tried to accent the point of BOM cost and the target and why lower cost devices may have been used, thus causing these problems, and yes exactly because it is china. A really good PMU design may have need for 6 layer PCB design, thus making things more expensive. 4 are the main budget target(for example nowadays we only get 10 layer motherboards in ITX designs, no wonder they OC memory better often). I also agree that theory doesn't match experience... you have to stick in the lab for years and design and torture devices to understand what's what. School doesn't give that to you... simply it would be too expensive and time consuming... like learning for a surgeon... you finish the school and you are OLD, you spent your youth slicing dead people in the morgue.

I agree to the point you can have other great part suppliers, but from engineering stand point during dev phase, Murata provides a good datasheet, and that's gold. Then you can play around, you have values and can simulate things, then understand what's what. You can have the same IC and make two different boards for it, due to design they will operate like day and night in terms of noise, stability and performance. Coils like those rarely suffer from problems at such low currents, in GPU yea... there are dozens of A. It looks like a QC design, so obviously you have a PMU and drive Skyworks radio IC's and it is WAAAY more expensive than that ASUS Broadcomm you really cannot compare them... it like a Bentley vs Jaguar... (the latter one must be understood worse). I hope by inductors you din't mean those small, but higher inductivity ones for noise filtering before the VREGS thus we are talking about different things. ASUS design didn't even have those... and that's a basic design thing having a good power supply.

There actually are universal PMU's... for example like Raspberry Pi 3 has. Four output MxL7704... I have a funny story about it... I ordered a free sample for it... Mouser sent me a wrong order... complained, I am not a thief... and they did nothing about it... well... I kinda didn't argue sending me a roll of few thousand capacitors worth of few hundred bucks... but yea... ordered another free sample after making a complaint xD.

Btw I use kinda a shit old router... Linksys EA6300... I am lucky with the fact, they had to rebadge EA6400 and sell it as EA6300 for some reasons. It is AC1600 and they didn't cripple it slower, and in real life I just did a speedtest in my phone and it shows stable ~400Mbps upload and download... I am kinda happy with that... There is merlin and WRT for it... but it works like crap... the wifi performance is way slower, yes that sucks, I would like to have WRT... well this will kick the bucket... it is like 7!!! years old.

This is not true at all
I am not grasping the info as read on the internet but as insider info. There are rebadged ex atheros devices like used in cheap TPLINK made around 2010, and they are still fine... and that's what they will sell you, but not the bleeding edge ones. And... currently we are waiting for native pcie4 SoC anyways... to have few channels per device to be cost reasonable for mortal consumers, like you and I are wanting at least 2.5G-5G switches at home..

And Swede for the sake of conspiracy want something funny...? a Huawei E589u-12 it has WCN1314 that's a Qualcomm or who knows pre Atheros acquisition design... Older chummy days... so they kinda shared information... ah so lax...

 
Last edited:
Joined
Mar 10, 2009
Messages
192 (0.03/day)
Location
Mexico
System Name HTPC
Processor AMD Ryzen 7 5800X
Motherboard Gigabyte X570S Aorus Master
Cooling BeQuiet Dark Rock Pro 4
Memory Kingston Fury Renegade RGB 32GB Kit (2x16GB) 3600Mhz DDR4 CL16
Video Card(s) Sapphire NITRO+ RX 5700 XT
Storage Acer Predator GM7000 4TB / WD Blue SN570 1TB / TOSHIBA HDWD130 3TB
Display(s) LG Oled 42C3
Case Lian Li O11 Air Mini
Audio Device(s) Creative Labs Sound BlasterX G6
Power Supply EVGA SuperNOVA 850 P2
Mouse Logitech G900
Keyboard SteelSeries Apex Pro
Software Microsoft Windows 11 Pro
If you're looking at a new router, I would consider the Netgear R7800/X4S with Voxel's firmware. It's really great hardware in it and Voxel & Co are developing some really cool features for, if that's your kind of thing. Admittedly, anything supported by Merlin would be nearly as good, although, personally I do prefer QCA to BCM when it comes to the hardware side of things.

The R7800 is admittedly quite an old router by now, so you might want to consider something else simply because of that. Stay away from 802.11ax for now, as it's not ready for mainstream usage as yet, imho.

As an affordable router, the RT-AC68x range isn't bad, but it's not great either. The one I'm using at the moment (as I'm stuck abroad due to the Wuhan virus) has pretty poor 5GHz range.

Not sure what else you're looking for in a router in terms of features.

I had the R7800, great hardware, but useless firmware, buggy as hell, had to use 1 old version just to get DLNA to work on some devices

Edit: RT-AC86U and R7800 side by side they performed almost the same, what i didnt like was netgear UI, its fugly and a mess OTOH Asus is more user friendly
 
Last edited:

TheLostSwede

News Editor
Joined
Nov 11, 2004
Messages
17,864 (2.43/day)
Location
Sweden
System Name Overlord Mk MLI
Processor AMD Ryzen 7 7800X3D
Motherboard Gigabyte X670E Aorus Master
Cooling Noctua NH-D15 SE with offsets
Memory 32GB Team T-Create Expert DDR5 6000 MHz @ CL30-34-34-68
Video Card(s) Gainward GeForce RTX 4080 Phantom GS
Storage 1TB Solidigm P44 Pro, 2 TB Corsair MP600 Pro, 2TB Kingston KC3000
Display(s) Acer XV272K LVbmiipruzx 4K@160Hz
Case Fractal Design Torrent Compact
Audio Device(s) Corsair Virtuoso SE
Power Supply be quiet! Pure Power 12 M 850 W
Mouse Logitech G502 Lightspeed
Keyboard Corsair K70 Max
Software Windows 10 Pro
Benchmark Scores https://valid.x86.fr/yfsd9w
My eyes Swede... such a blurry pic... I took the math because it is an old router. There are even illegal and old 5W devices. But this is an Asus...

I kinda tried to accent the point of BOM cost and the target and why lower cost devices may have been used, thus causing these problems, and yes exactly because it is china. A really good PMU design may have need for 6 layer PCB design, thus making things more expensive. 4 are the main budget target(for example nowadays we only get 10 layer motherboards in ITX designs, no wonder they OC memory better often). I also agree that theory doesn't match experience... you have to stick in the lab for years and design and torture devices to understand what's what. School doesn't give that to you... simply it would be too expensive and time consuming... like learning for a surgeon... you finish the school and you are OLD, you spent your youth slicing dead people in the morgue.

I agree to the point you can have other great part suppliers, but from engineering stand point during dev phase, Murata provides a good datasheet, and that's gold. Then you can play around, you have values and can simulate things, then understand what's what. You can have the same IC and make two different boards for it, due to design they will operate like day and night in terms of noise, stability and performance. Coils like those rarely suffer from problems at such low currents, in GPU yea... there are dozens of A. It looks like a QC design, so obviously you have a PMU and drive Skyworks radio IC's and it is WAAAY more expensive than that ASUS Broadcomm you really cannot compare them... it like a Bentley vs Jaguar... (the latter one must be understood worse). I hope by inductors you din't mean those small, but higher inductivity ones for noise filtering before the VREGS thus we are talking about different things. ASUS design didn't even have those... and that's a basic design thing having a good power supply.

There actually are universal PMU's... for example like Raspberry Pi 3 has. Four output MxL7704... I have a funny story about it... I ordered a free sample for it... Mouser sent me a wrong order... complained, I am not a thief... and they did nothing about it... well... I kinda didn't argue sending me a roll of few thousand capacitors worth of few hundred bucks... but yea... ordered another free sample after making a complaint xD.

Btw I use kinda a shit old router... Linksys EA6300... I am lucky with the fact, they had to rebadge EA6400 and sell it as EA6300 for some reasons. It is AC1600 and they didn't cripple it slower, and in real life I just did a speedtest in my phone and it shows stable ~400Mbps upload and download... I am kinda happy with that... There is merlin and WRT for it... but it works like crap... the wifi performance is way slower, yes that sucks, I would like to have WRT... well this will kick the bucket... it is like 7!!! years old.


I am not grasping the info as read on the internet but as insider info. There are rebadged ex atheros devices like used in cheap TPLINK made around 2010, and they are still fine... and that's what they will sell you, but not the bleeding edge ones. And... currently we are waiting for native pcie4 SoC anyways... to have few channels per device to be cost reasonable for mortal consumers, like you and I are wanting at least 2.5G-5G switches at home..

And Swede for the sake of conspiracy want something funny...? a Huawei E589u-12 it has WCN1314 that's a Qualcomm or who knows pre Atheros acquisition design... Older chummy days... so they kinda shared information... ah so lax...

It's not that old and even so... 5GHz PA/LNA's are power hungry and produce a lot more heat than you'd think.

Asus is Taiwanese. Yes, they produce some things in China, but at least their products are designed in Taiwan.
I think they cut far few corners than their competitors, especially D-Link and even more so TP-Link and even cheaper brands like Totolink and Tenda (btw, what is it with all these "link" company names?).
Asus has a pretty good reputation in the router market, but they've had a few duds as well. These days they seem to churn out a lot of cheap (cost, components and made) products that they don't seem to support, which is surprising after the FCC slapped them on the fingers a few years ago. Maybe the FCC agreement to provide software upgrades have ended?

A PMU doesn't require more than a four layer PCB, why should it? You can't compare super compact phone PCB's with something like a router where you have ample space to place the components and don't need more than four layer PCBs.
I mean, if you can make a PC motherboard using a four layer PCB, why should a router need more?
Sure, more layers CAN be good, but it's not always better to have more layers, as it just makes a product more complex. In PC motherboards, a lot of the layers are also just extra shielding.

When I worked at Securifi, they hired some guy in his 50's that was supposed to be a really experienced engineer. He was given a task that took him months and he just kept coming up with excuses why he hadn't found the problem and suggested it could be this, that or something else at every meeting about it. In the end he was fired, as he couldn't solve the problem and one of the other engineers figured it out in the end. It was something fairly simple, but it only happened sometimes, so it actually required some out of the box thinking to solve, which clearly the older engineer didn't have.

Datasheets are nice and all, but what I've learnt over the years is that sometimes datasheets are wrong and sometimes the reference designs are wrong and sometimes the hardware if flawed, so no matter, they don't help you solve the problem. This is why companies with local FAE engineers (good ones at least) often end up getting business as the device makers know they can get local support when there are problems.

That router was not a QCA design, at least not as far as the main SoC went, it was something entirely different. If you do some digging you can find out what it was, if you care. The Wi-Fi was indeed from QCA (with Skyworks PA/LNAs) and so was the switch.

Inductors are coils/chokes... Not sure what you're talking about. The Asus has at least four of them from what I can see.

A lot of routers are like that, one PCB, multiple products with either a removed front-end or some other tweak, such as less RAM or something stupid. I guess that's how you make multiple SKUs on the cheap.
I think you mean OpenWRT and DD-WRT for it? Merlin is exclusive for Asus routers.

QCA still sells older chips and they just simply didn't bother making a new silkscreen for those old chips, big deal. A lot of companies do that.
And older hardware doesn't always mean it's bad hardware, it just means it's no longer cutting edge. For a lot of industrial stuff, these companies need to provide the same parts for up to a decade. I guess they sell some of the spares or the ones that don't meet certain temperature requirements as general usage parts on the cheap.

Not sure routers needs PCIe 4.0 SoCs yet, most only have PCIe 2.0 to date, as it's fast enough. However, as we're moving beyond 2.5Gbps, unless it's built into the SoC, we might need something faster. Most router SoCs have RGMII or something similar for the Ethernet not PCIe. PCIe is often only used for the Wi-Fi, or in some cases for additional features such as SATA or USB 3.x.

I already have a basic 10Gbps switch at home, so I'm not waiting. Only two ports, but I only have two 10Gbps devices so far, so not really a problem.

I had the R7800, great hardware, but useless firmware, buggy as hell, had to use 1 old version just to get DLNA to work on some devices

Edit: RT-AC86U and R7800 side by side they performed almost the same, what i didnt like was netgear UI, its fugly and a mess OTOH Asus is more user friendly
They've released a lot of firmwares for it. I have had exactly zero issues with mine, but I wasn't an early adopter of it either. By far the best router I've ever owned. Even better with Voxel's firmware for it.

I do agree on the UI though, Netgear is way behind Asus there. Asus has done some really good work on their UI, IF you're a techy user. For the average consumer, both are a mess.
 
Top