Question

Netwok Question

  • 17 December 2016
  • 13 replies
  • 467 views

Userlevel 3
Badge +8
My Sonos configuration consists of one Boost and two ZP90's (downstairs and upstairs). The downstairs ZP90 is on an Ethernet cable, whil the upstairs ZP90 relies on SonosNet for its connection.

My questions are:

1. If the downstairs ZP90 is connected using an Ethernet cable, and is also receiving the signal from the Boost, which connection is the ZP90 actually using, and is there an indication to confirm which connection is in use?

2. My wired connection at the ZP90 is ~70Mbs. Given this robust connection speed, why would the network matrix be showing the connection to the ZP90 as "Red"? I tried three connection methods: a) wired connection with Boost off, b) wireless connection with Boost only, and c) both the wired and the Boost at the same time. All three resulted in the "Red" status. The Phyerr log for the downstairs ZP90 has some issues, while the logs for the upstairs ZP90 and the Boost are clean.

Any advice would be appreciated.

This topic has been closed for further comments. You can use the search bar to find a similar topic, or create a new one by clicking Create Topic at the top of the page.

13 replies

We've been here before, and 8 months ago you said the 'case was closed' by wiring the ZP90 to a 5GHz bridge in the local equipment rack.

1. If the downstairs ZP90 is connected using an Ethernet cable, and is also receiving the signal from the Boost, which connection is the ZP90 actually using, and is there an indication to confirm which connection is in use?
The ZP90 is getting its signal wirelessly from the BOOST, so if it's still wired to the 5GHz bridge something's not working as it should. You can see in the matrix that the BOOST's cells are coloured between it and the two ZPs, and in the opposite direction hence 4 coloured cells in total in columns 2+.

2. My wired connection at the ZP90 is ~70Mbs. Given this robust connection speed, why would the network matrix be showing the connection to the ZP90 as "Red"?

The left column colouring reflects the amount of ambient RF noise seen by that device. In the case of your Living Room ZP one of its antennas has a lousy noise floor, possibly because of adjacent equipment.

I tried three connection methods: a) wired connection with Boost off, b) wireless connection with Boost only, and c) both the wired and the Boost at the same time. All three resulted in the "Red" status. The Phyerr log for the downstairs ZP90 has some issues, while the logs for the upstairs ZP90 and the Boost are clean.

Which connection method is it at the moment? As noted, the matrix shows a wireless connection to the BOOST.

The phyerr log you posted is quite benign. Nothing of any significance there.
Userlevel 3
Badge +8
OK, so here is the matrix with the Boost and the upstairs ZP90 powered off. The downstairs ZP90 is hard-wired. The matrix shows the same noise levels. What exactly is the matrix reporting on if the wireless network is not in use? Does the "interference" register anyway? I am confused.
The SonosNet mesh isn't carrying any live streams, but the ZP's radio is still active.

If it would make you happier, disable the ZP radio and that red cell will go white with the word "undefined" against the RF information.

Doing so is not officially condoned, but instructions for using the appropriate URLs can found easily if you google for them.
Userlevel 3
Badge +8
So what you are essentially saying is that the tools I am using do not reflect what is happening when using a wired connection. I can't be the only Sonos customer who prefers a hard-wired connection to a wireless one. Are there tools to analyze performance in this scenario?

And I have no desire to disable the disable the ZP radio.
So what you are essentially saying is that the tools I am using do not reflect what is happening when using a wired connection.
I'm saying that (a) your ZP is in a noisy location, (b) its radio remains enabled so the matrix diagnostic shows the noise, and (c) it doesn't matter one jot given that the ZP is wired.

I can't be the only Sonos customer who prefers a hard-wired connection to a wireless one. Are there tools to analyze performance in this scenario?

Of course users wire players all the time. How do mean, analyse performance? Ping times and packet loss would be the usual things to check if a wired connection was suspected of being flaky. Are you actually experiencing any problems now? I thought the 5GHz bridge arrangement was successful?

And I have no desire to disable the disable the ZP radio.

As a matter of interest why not?
Userlevel 3
Badge +8
I have no interest in disabling the wireless radio because you said that all,this will do is result in a blank network matrix. I don't see any value in doing that.

I have disconnected the hard-wired connection and am now relying on the SonosNet connection from the Boost. My objective now is to eliminate the "noisy" placement of the ZP90. The only other wireless device in the vicinity of the ZP90 is the Linksys WES601N wireless bridge. There are other pieces of equipment in the equipment rack, but they are not hooked up wirelessly.

I have two choices:

1. Move the ZP90 out of the equipment rack and place it off to the side, away from any equipment that might be contributing to the "noise".

2. Leave the ZP90 in the equipment rack, and move the wireless bridge off to the side, away from the ZP90.

Based on your experience, would you recommend one of these two options? And how far away is "far enough" that any noise will be reduced significantly?
Ambient noise can be caused by adjacent digital electronics of any sort, not necessarily wireless. For example a TV/monitor in close proximity can have this effect.

I'm still unclear what problem you're trying to solve. In months gone by you suffered from dropouts, and migrated to using a 5GHz bridge evidently with success. Have the dropouts returned? Are there occasional problems with reaching the living room ZP90 from a controller?
Userlevel 7
Badge +26
The red square showing there on the noise floor is just showing that the wireless antennas are seeing a lot of noise on them. It doesn't indicate any traffic on the wireless itself. I've seen 20 room systems, all wired up in a rack and all showing bright red on the diagnostics with a noise floor in the -50s, working perfectly fine. It's just that the antennas are always "listening", the playback is coming through the wired network connection.

As ratty said, is there any problem you're seeing? Audio cutouts, or playback issues perhaps? If you aren't, I wouldn't worry about seeing red on the network matrix.
Userlevel 3
Badge +8
Sorry for the delayed reply. I have not been experiencing any issues, per se. Since this discussion started, I switched over to running the ZP90 with just the SonosNet connection from the Boost. During this time, there have been no dropouts or difficulties connecting to the ZP90 from an iOS controller.

Several things have prompted this recent discussion. First of all, several months ago I was forced to replace my Linksys router because it was showing reliability issues. I now have an Asus RT-AC87U, which has been preforming flawlessly. However, for some reason, the Asus and the Linksys WES610N wireless bridge don't interface well. I cannot get the WET610N to connect to the router on the 5GHz network. On the 2.4GHz network, I can sustain a 65Gbs download speed, which is reasonably good for things like Netflix streaming, but nowhere near the 300Gbs rated speed of my ISP service. My wireless laptop interface is able to achieve 130Mbs on the 5Ghz network. So, since I am by nature an inquisitive person, I have been trying to determine whether wiring the ZP90 to the wireless bridge, or using SonosNet, provides the faster connection. (Note I equate faster with more reliable, which may not be the case.) So far, I cannot figure out how to measure any difference.

Second recent change is that I have installed the Wyred 4 Sound modification on the ZP90 (https://www.wyred4sound.com/products/upgrades-and-mods/sonos-connect-modified). It may be my imagination, but I think the Matrix report looks worse since the modification, and I want to make sure the modification has not affected the wireless capabilities of the device. I know that a modified device probably voids any support from Sonos, but darn, the ZP90 sounds even better now!

So, I still haven't figured out whether the ZP90 is actually using the wired interface when it is plugged into the wireless bridge. Since I am not currently experiencing any issues regardless of how it is connected, this is probably a moot point. I am just curious, and want to provide the optimal connection. Thank you for your patience and feedback.
Userlevel 7
Badge +26
So, I still haven't figured out whether the ZP90 is actually using the wired interface when it is plugged into the wireless bridge.

If you go to that same page where you get the network matrix open up the name of your player. Then select Ethernet Ports Statistics. It'll let you know how much data is being passed on the Ethernet, that might help confirm that you're using that connection instead of the wireless.

Alternatively, if you're looking to test the speed, leave it wired and go to Http://IPofRoom:1400/tools.htm and run a traceroute to a known location, like google.com. Then, disconnect the wire so you're wireless and run another traceroute to the same source. That'll help you identify the connection speeds.

When it comes to reliability, you only need to reach a certain threshold on speed before both are as reliable as they can be. If Sonos is streaming at 320 kbps, and you're at a consistent 15 mb or 40 mb, both are going to be as close to 100% reliable as that factor can get you.
Ar least one of the aftermarket mods -- not necessarily Wyred4Sound, I don't recall which -- warned customers that wireless performance would be compromised.

As for determining whether the Ethernet connection is in use, it wasn't at the time of the first matrix snapshot.
Userlevel 3
Badge +8
So, I still haven't figured out whether the ZP90 is actually using the wired interface when it is plugged into the wireless bridge.

If you go to that same page where you get the network matrix open up the name of your player. Then select Ethernet Ports Statistics. It'll let you know how much data is being passed on the Ethernet, that might help confirm that you're using that connection instead of the wireless.

Alternatively, if you're looking to test the speed, leave it wired and go to Http://IPofRoom:1400/tools.htm and run a traceroute to a known location, like google.com. Then, disconnect the wire so you're wireless and run another traceroute to the same source. That'll help you identify the connection speeds.

When it comes to reliability, you only need to reach a certain threshold on speed before both are as reliable as they can be. If Sonos is streaming at 320 kbps, and you're at a consistent 15 mb or 40 mb, both are going to be as close to 100% reliable as that factor can get you.


I connected the ethernet cable, but the Ethernet Port Statistics is not showing any packets. I reset the player and looked again, still no packets. The Boost is swowing packets, of course. I'll try the other test later on.

Regardless, the wireless connection is showing no faults, so this is probably an academic exercise.
I connected the ethernet cable, but the Ethernet Port Statistics is not showing any packets.
On some devices they don't, for some reason.

Once again, the network matrix is the easiest place to look and see if the wireless is in use.