OFDM ANI Level 9

  • 1 February 2014
  • 35 replies

Everything works and I shouldnt worry but......

My BT HH2 expired last week and I replaced it with a Netgear D6200 (dual band). After getting everything setup and SONOS working, I thought I would have a quick peek at the matrix and was surprised to see a red box - on my Play1 due to an OFDM ANI level 9.

The noise floor is around -100, and the physical error count is consistently in the 1000 range, and more importantly it works fine. The other 7 sonos components show green and yellow and 1 amber.

The play 1 is positioned in my study , along with the router, the digital phone base , the QNAP. the bridge, a PC witha wireless keyboard and mouse - there is a lot of stuff going on and i have spread them out as far as I can, but its no Tardis in there.

So I can live with a red box on a matrix, as long as the music sounds good which it does, but with the play 1 being 5 feet away from any potential source of interference and still suffering so badly, I wonder whether I can afford to be so complacent. Is this a sign of other conflicts and problems elsewhere - with that level of interference should I be concerned about other non-sonos devices ? - could my router performance be impacted by sonos ?

Maybe I shouldn't worry if it all seems to work, but its a red box and I'm male - its my instinct to fix it

Any thoughts ?

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.

35 replies

Is your Sonos system actually performing correctly? If you're not suffering from dropouts -- and with yellow/green in the matrix the signal strengths are more than adequate -- then whatever the ANI is doing to reject local noise it's succeeding. I suspect it is quite sensitive. There are times when I can see elevated ANI levels which nothing in the house would explain. Interference could well be coming in from the outside, maybe even mains-borne.
Userlevel 2
Badge +1
Interesting... I do have a few computers with their bluetooth on, some with wireless keyboards. I see a sharp narrow and small spike on my scan when I use them. It could contribute to the high ANI but... when I see the power generated by the SONOSnet transmission, I think that noise is negligible... The Sonos is actually the largest source of noise (pun intended) to all other devices by several orders of magnitude. My ubiquity wifi APs are on 40mHz width and is spread on channel 11 and 7 and they don't come close to intensity and frequency of the sonos...
ANI is representative of the noise rejection doing its work. I can immediately tell if there's a 2.4GHz wireless mouse active in the house, and roughly where, on the basis of the pattern of elevated ANI. In a quiet environment all my ANIs are zero. A long-range Bluetooth transmitter drives the ANI levels up to 9.
Userlevel 2
Badge +1
No. The noise you are seeing is actually the sonos. It is sonos playing and idle. Without the sonos (unplugged), it is a flat. I also changed the sonos channel and it clearly follows. Again, not very surprising but the point I was trying to make is that the OFDM ANI is not representative of anything.
There looks to be substantial noise in the same passband as Sonos' channel 1, at a similar RSSI.
Userlevel 2
Badge +1
This is without. Needless to say that microwave makes a mess out of this. But the bottom line is that I am not seeing any correlation between the noise and the ANI number.
Userlevel 2
Badge +1
See comparison with Sonos and without. Sonos is obviously on channel 1. I started looking at it because my channel 11 zigbee were failing
Which channel is SonosNet using? Which channel(s) -- and bandwidth -- is the WiFi using?
Userlevel 2
Badge +1
I use a RF scanner. much better than Ubiquity which I tried too. I can see the full spectrum in intensity and hit frequency. My band is extremely clean and found that the ANI correlates to nothing RF related. Or if it is, it is very oversensitive to something I cannot see.
Non-WiFi users of the 2.4GHz band can push the ANI to the highest level. Look for Bluetooth, proprietary 2.4GHz mice/keyboards, wireless headphones, etc.

FWIW some WiFi scanners will detect SonosNet if any Android devices have been configured to connect to it.
Userlevel 7
Badge +21
Are you using a WiFi scanner that only shows signals with SSIDs or are you using an RF scanner that looks purely at RF energy in the WiFi spectrum?

The WiFi type scanners can't even see SonosNet as it has a hidden SSID so they aren't very useful.

I use a Ubiquity access point and their control software to run RF scans here and things often look very different there than on my tablet's WiFi monitor app. For closer range detection I also have a microwave oven leakage detector that does a fair job of finding 2.4 GHz RF energy at a couple inches range. I wouldn't recommend one but I had it in my tool box for other reasons.
Userlevel 2
Badge +1
Looking through the forum for threads on this topic and I am still quite puzzled by what I observing after reading them all:

I have a fairly large setup with 3 pairs of Play1, 2 standalone Play 1, 2 Play 3s and 1 Play 5 spread through the house. When I go and check the OFDM ANI they are all red at 9 and a couple at 8. Most of my units are standing on their own with no electronics around. If they have some, they are running wifi at 5GHz. I even have a 2.4GHz band scanner which comes out pretty clean. I have an older Connect Amp in the system which is the only yellow unit but is in a cabinet full of electronic with a receiver, an apple TV and a router... and it is showing an OFDM weak signal of 7 which is supposedly pretty good. All the noise floors are reporting at -100 to -120 so I have pretty low noise. All the STP forwarding are green or yellow. (yellow for the more distant ones so this makes sense). It makes me wonder if there is something wrong with the OFDM ANI reporting.
That is just silly for a wired Ethernet connected hard disk enclosure.
The means of communication are, in this case, not relevant. Digital electronics operates at high frequency and can leak RF noise.
Userlevel 7
Badge +21
Found this old topic when I did a web search for Sonos OFDM ANI level and it hit the nail on the head. Powered down my WD Live Drive and the network matrix shows a much improved connection. Lots of troubleshooting time saved. --- Thanks!

Doesn't say much for WD quality though, my WD is 3 feet from my Play:3 and ran the connection into the red. That is just silly for a wired Ethernet connected hard disk enclosure.
The system could have been on the edge due to the WD interference. With few extra water bags in the area, the edge was crossed.
Looks good so far. The 0.5 metre separation is just about manageable, and leaves the Boost in green too.
Hopefully this will improve music play stability beyond the present decent levels.
A high ANI level combined with low signal strength could spell trouble. If the WD drive was right next to the BOOST see what difference 0.5m separation makes.
The Boost changed to Green right away, with ANI reading of 3. My question above is relevant...
Ahh ok. I will check this out by turning the WD off and seeing what happens to the readouts on the Boost.

If the Boost then drops into the green, is it worth adding some distance even if not convenient to do so? Can reading of 6 and 7 on the Boost affect music play?
There can't be any effect from the WD My Cloud sitting next to the Boost, can there?
There could. It contains digital electronics, which inevitably leaks HF electromagnetic radiation.
None of those are in the vicinity. There can't be any effect from the WD My Cloud sitting next to the Boost, can there? It doesn't have any wifi in it...
As I've often remarked, ANI level is the canary in the coal mine when it comes to broad spectrum noise from non-WiFi 2.4GHz devices: proprietary wireless mice/keyboards, headphones, cameras, Bluetooth, etc.

An elevated ANI level is an indication that the radio's 'immune system' is at work, and I've had successful operation at level 9, albeit with signal strength in the high 40s or above.
No change after reboot. The Boost flips back and forth between levels 4 and 6, yellow to orange. Other units are steady in green.

It was a small party, just five water bags, so that can't have been the reason. No bag was anywhere near the Boost!

There were 3-4 bottles of alcohol close to it though, but not in the line of sight to the coordinator unit.

Systematically power down something and check the Matrix.

Rather than interference, perhaps there were lots of bodies at the party. Humans are mostly bags of water and water absorbs 2.4GHz energy very well. Bodies grouping at an inconvenient location could cause this.

The first member of a SONOS Group becomes the "coordinator" for the Group. In your case if the coordinator became isolated by water bags ... .

Maybe I shouldn't worry if it all seems to work, but its a red box and I'm male - its my instinct to fix it

Any thoughts ?

For somewhat similar reasons, my situation with the first vertical column in the Matrix for the Boost.
Phyerr counts show single digits in 5000 and the rest under 1000.
Noise floors are -105, -108.
The box tips over into orange quite often, when the OFDM ANI level goes to 6. Even when noise floor drops to -115. When the ANI drops to 5 or 4, the box goes to yellow.
All other units in the system tend to almost always be green boxes in the first column.
The Boost is placed next to the WD My Cloud NAS and three feet vertically away from the router.
What is causing the Boost to be different? Could the being next to the WD be a reason?
I looked at this because yesterday night during a party, one zone dropped out of the group and did not come back quickly. I don't remember all I did then to get all zones to play again. I think I stopped all, ungrouped them, reloaded the playlist and regrouped them. It was late into the party hence I don't now remember all the steps! And I have no idea if this is related to the orange box thing for the Boost.