REVIEW: The best yet most affordable network switch (TPLink WR902AC)

[keithc]

I haven't' gone down the rabbit hole further with the 3.3v hack, LifePO4 supply, or tried different 5v battery packs. Those days are behind me!

Amen brother. I have a slew of TPLink carcasses here from my failed attempts before I got it to work... but hey, at $30 a nut I just keep cracking them open...

 

[seeteeyou]

However, access point mode seems to bring more significant improvement on WR802N model.

Just wondering if you're happy to provide both Firmware Version and Hardware Version as shown below?

https://www.tp-link.com/us/user-gui...r-6-configure-the-router-in-access-point-mode

Much easier to replicate the results if we're following everything to the letter, many thanks.

 

[Exocer]

Keith,

I've tried some of these experiments, and can summarize my findings as follows. First, let's define additional scenarios:

E) Home network wifi / wifi to TPLink+Anker -> Taiko Router -> Taiko Switch -> Olympus server
F) Home router -> Olympus Server

My findings:
D >> (i.e. much better than) F — pretty amazing improvement over a straight connection for a sub-$100 solution!
A >> D — the Taiko stack brings different, and more substantial enhancements than the travel router's
E > (i.e. better than) A — the travel router adds SQ on top of the Taiko stack
B < (i.e. not as good as) E — removing the Taiko switch still has a negative impact on SQ

I personally have settled on E (full Taiko stack preceded by battery-powered travel router) as my preferred configuration for my Olympus.

I haven't' gone down the rabbit hole further with the 3.3v hack, LifePO4 supply, or tried different 5v battery packs. Those days are behind me!

I do not have the Taiko stack but I do have a diy clock modded ubiquiti edgerouter 4 and pf clock modded buffalo switch.

Air gapping my PF Ultra modded Buffalo Bs GS2016 with the Tp-link902 AC, connecting the tp-link to my home Wifi (also connected to the modded Edgerouter 4) sounds significantly better than the wired connection (edgerouter connected directly to buffalo).

I opted for the smaller Anker 313 because it comes with a pair and I have two travel routers. One will come out of the edgerouter in wireless ap mode, the other connecting to it in client mode, with output to the buffalo (i am going to conduct this test later, only using one presently).

I intend to use the 2nd Anker to try using my existing subnet isolated network with an air gap. So I would be connecting two Tp-Link902s.

It will be easier to confirm if power to the source wireless AP is impactful to SQ with two equivalent batteries.

In the interim, I have tried tp-link902 directly to server and have come to the same conclusion: Incorporating the Buffalo between the wireless AP and server still adds value.

I swapped out the buffalo for a stock QNAP QSW-2104-25-A to determine how impactful the switch quality itself is, and the Buffalo sounded significantly better than the stock QNAP switch here. I am trying a modded Teradak FMC with ocxo(not pink fuan level) with the same power source as used with the Buffalo and QNAP and this also added value when placed between the travel router and server.

Will report back once I test two travel routers on a dedicated separate subnet. If this brings improvements I will most definitely delve a bit deeper into the mods where deemed necessary. I suspect mods to the travel AP closest to the server will be the only one worthwhile to mod and feed premium power.

 

[kswanson27]

I do not have the Taiko stack but I do have a diy clock modded ubiquiti edgerouter 4 and pf clock modded buffalo switch.

Air gapping my PF Ultra modded Buffalo Bs GS2016 with the Tp-link902 AC, connecting the tp-link to my home Wifi (also connected to the modded Edgerouter 4) sounds significantly better than the wired connection (edgerouter connected directly to buffalo).

I opted for the smaller Anker 313 because it comes with a pair and I have two travel routers. One will come out of the edgerouter in wireless ap mode, the other connecting to it in client mode, with output to the buffalo (i am going to conduct this test later, only using one presently).

I intend to use the 2nd Anker to try using my existing subnet isolated network with an air gap. So I would be connecting two Tp-Link902s.

It will be easier to confirm if power to the source wireless AP is impactful to SQ with two equivalent batteries.

In the interim, I have tried tp-link902 directly to server and have come to the same conclusion: Incorporating the Buffalo between the wireless AP and server still adds value.

I swapped out the buffalo for a stock QNAP QSW-2104-25-A to determine how impactful the switch quality itself is, and the Buffalo sounded significantly better than the stock QNAP switch here. I am trying a modded Teradak FMC with ocxo(not pink fuan level) with the same power source as used with the Buffalo and QNAP and this also added value when placed between the travel router and server.

Will report back once I test two travel routers on a dedicated separate subnet. If this brings improvements I will most definitely delve a bit deeper into the mods where deemed necessary. I suspect mods to the travel AP closest to the server will be the only one worthwhile to mod and feed premium power.

I've been running the TP Link 902 for about two hours. It's a step up in smoothness, clarity and seems to definitely have a lower noise floor. No drops on 192 or DSD 128 (DOP). I've been running my Aardvark Link from the TP Link to my server so far. Tomorrow I'll substitute a Blue Jeans Cat 6A cable for the link and report back. We're leaving for Christmas week Monday. When I return I should have the QSA stuff I ordered so I'll try one of their Ethernet cables as well. At this point the sound I have now is the best I've heard from this system.

 

[audioblazer]

View attachment 162829


I recently stumbled upon the parenting milestone of dealing with Gen Alpha slang. My son, when asked to put down his iPad, used language which left me perplexed and uneasy. What is this idiocy and which damn kid did he learn that from? Why can't they just talk in a respectful way and use normal English especially in front of their dad? Later that night, after I calmed down and had 2 glasses of wine, I started researching how to deal with kids and their distorted lingo and learned, much to the realization of my ignorance, that slang can be appropriate and useful among the younger generation in how they communicate.

Which brings me to this review. Just like language I take for granted everyday, for almost three decades I have increasingly gone deeper down the audiophile rabbit hole and religiously believed that "everything matters" -- components must be shielded and grounded from environmental noise, isolated from mechanical vibrations, supplied with the cleaning possible power via cables that minimize interactions with anything else around it. In our collective pursuit of uber excellence in the era of digital streaming, this means dedicated audiophile-grade network cables, routers, switches, servers and streamers floating on isolation platforms and driven by exotic power supplies augmented by equally-exotic power conditioners, fuses and power cords, all sorts of signal isolation, clocks and conversions to make sure those little 1s and 0s are interpreted as perfectly as possible. I drank the Kool-Aid and even contributed towards this effort with my review of the Network Acoustics tempus, and have owned 3 different switches (Innuos, Network Acoustics, and Taiko) and several power supplies in this never-ending pursuit. I would put my listening room in a Faraday cage if it was possible.

What if something came along and just makes you question if all that we have accepted in high-end audio is.... well, maybe not-so-absolute?

Courtesy warning: The rest of this post will be heresy to some audiophiles. It is so against the grain that perhaps I will be dismissed and shunned, and therefore to protect the credibility of other WBF members involved I will defer to their willingness to post on their own. Oh well -- here it is...

First, I cannot claim credit for any of this discovery. A saavy audiophile has a German website with electrical measurements of various switches and found that common mode noise is inherent in physically-connected switches, and that while Wifi can theoretically eliminate this interference, the key is in the implementation: using a low-energy minimalist wireless router in client mode under pure battery power. This was introduced to me by a WBF member and over the past two months an increasing number of us have been trialing this Wifi network setup in our various systems and have found results that I felt necessary to share in the spirit of the audio community.

For background, the incoming Comcast cable goes to my living room where the Motorola modem directly connects to an Orbi Mesh router. This then connects via a switch and in-wall cabling to my listening room where there is another switch -- at this point a QSA-Lanedri ethernet cable goes to my Taiko Audio Extreme Switch or Network Acoustic Tempus switch and a Network Acoustics Muon Pro filter and then my streamer which relies on Roon / Qobuz and sends music to my Lampizator Horizon. This has sounded so enjoyable that I have preferred streaming over locally-stored playback for the past year. While my peers have additional upstream audiophile routers, switches and optical moats, I am absolutely satisfied with this setup and felt I was done. Until this past month.

Instead of all the above, I am now using a $30 TPLink WR902AC wireless router powered by a $50 Anker battery pack to supply streaming data to my server. You read that correctly. This is the kind of device that people use at a coffee shop when working on their laptop... except with some easy software options used to tune it for optimal audio streaming use.*

keithc.just checking . are u listening thru headphone ?

 

[keithc]

keithc.just checking . are u listening thru headphone ?

Haha yes sir.
Speakers are in storage at the moment.

 

[Exocer]

Right on. You had better luck than I, but it might be because I opted for the travel router that supports gigabit. The instructions Keith provided can’t be performed exactly as described. Frustrating as heck so I will likely bail on it.

I did also purchase the recommended device so I will likely try configuring that instead.

Did you ever get this working? 100mbps isnt an issue with Roon and I don't do DSD streaming from a NAS but I would like to know how the 1000mbps version sounds in comparison to the 920AS.
Cheers