I got the HDMI cable today and all problems seem to be solved. Thank you so much!!

All - I just want to summarise and clarify my workaround.

The audio drop-out is assumed to be when the Beam Gen 2 communicates with the TV, causing the TV to trip over and drop the audio for moment. This only seems to happen on HDMI 3 or 4 and when playing 5.1 PCM audio. The most common sources of this audio type include the PS5, Nintendo Switch and the Apple TV 4K 3rd Gen.

Although I linked a specific one way fibre-optic HDMI cable that I bought and confirmed fixes the issue for me (or more correctly, provides a work-around), I believe that any similar HDMI cable should work.

My understanding is that these cables do have some copper wires in them for certain elements of the HDMI signal, but some parts of the signal go down the fibre-optic part. The fibre-optic part is one way only. The TV needs to be at the source end of this cable and the Beam Gen 2 at the other.

It seems the signal the Beam Gen 2 sends to the TV which is assumed to cause the drop-outs, is trying to go the wrong way down the one way part of the fibre-optic HDMI cable and therefore doesn’t get through to the TV, resulting in the audio signal from the TV not being interrupted. I can’t really explain exactly what is happening as I don’t really understand how HDMI works (there are others on here that know much more than me and my interpretation of the fix may be completely wrong anyway), but as an engineer that has experience troubleshooting stuff, I thought this had a chance of working, so I tried it and was successful.

It is great to see people posting on here that have tried this and also had success. Hopefully this won’t stop Sonos from implementing a proper fix in the future as I’d wager a lot of Beam Gen 2 sound bars have already been returned due to owners thinking they are faulty - most people aren’t going to try and troubleshoot or search for fixes online. At least we now have this workaround to keep us going. Many thanks to other posters on this thread that also helped narrow this down.

One thing I don’t understand though is why HDMI 1 is unaffected. While this may be an unpopular opinion, I think this fact suggests the problem may be the TV after all…..

Purely speculation, but maybe HDMI 1 and 2 share a bus on the board, and 3 and 4 share another. 3 and 4 are affected as they are on a different bus to the Beam Gen 2. So the signal from the Beam Gen 2 to the TV is normal, but the TV deals with it incorrectly on HDMI 3 and 4 or it has an issue passing it between busses. Again… this is purely speculation.

This has been speculated before but then the question is why isn’t the Arc affected if the TV is at fault?

So the cable works but this isnt an ideal solution because its an added cost as well as cables long term reliability being one of the point of failures. 

I have tried the same HDMI cable as suggested here and I can confirm this solution works for me with the Beam and LG C1: no more dropouts with the Nintendo Switch and Xbox Series X. I have also upgraded my Apple TV from a Gen 2 to a Gen 3, which fixed the issue of LPCM 5.1 being send as 7.1 (and thus lowering perceived audio quality).
 

Quite a costly solution, but at least everything is working as it should. I want to thank all the wonderful people in this thread for finding a solution to a problem which had been driving me crazy for over a year.

I only recently realised (when looking into how optical HDMI cables are wired) that my logic analyser can decode both the DDC and CEC HDMI signals, so I found a cheap HDMI breakout PCB on e-bay and now have it connected between the TV HDMI2 and Beam2 using the original Sonos cable (not the reversed optical).

When I restart the Beam2 I can now read “Sonos Beam” being spelt out in ASCII in the CEC SetOSDName command and the EDID being transferred over the DDC I2C channel during the initial boot up. However, when a drop out happens I see no obvious immediate activity on either channel, so I’m thinking the HDMI2 connection is not collapsing every 59s. I haven’t put much thought yet into how I could monitor the TMDS/eARC lines as I think my test equipment is too basic.

The Beam2 could possibly be sending a valid (but slightly different to the Arc) format over TMDS that triggers a bug when eARC is routed from HDMI3 (maybe HDMI 2, 3 and 4 are sharing resources with HDMI1 having the better isolation). The bandwidths involved shouldn’t be maxing out the HDMI 2.1 circuits though if my A1 can handle the same configuration? So maybe not load related. I was reading about InfoFrames in the TMDS signal, so maybe something buried in there.

Alternatively though, the Beam2 could be sending an anomaly over TMDS and the TV is just handling it differently (but within spec) depending on which source the eARC stream is being sourced from.

But it could also be that both devices are operating reasonably, but so close to some tolerance that only when combined triggers an interrupt/underflow in the TV HDMI circuitry.

There are more things that I don’t understand:

• Why is the issue only audible on 5.1 if 7.1 also show an anomaly on the 1111 menu every 59s?
• Why did HDMI 1 become more robust?
• why did the dropouts change from every 39s to every 59s
• Why does cutting the CEC line stop eARC working, it shouldn’t need CEC to operate
• What did Apple change between gen2 and gen3 to make AppleTV correctly use 5.1 instead of 7.1 - and why wasn’t it back ported to gen 1 or 2 in the same TVOS patch

My C1 wanted to download a new firmware today, but I haven’t installed it yet in case the behaviour changes yet again...

The million dollar question.

I suppose a lot of us are feeling a bit more happy lately due to the discovery of valid workarounds. But don’t this influence us into absolving Sonos of accountability. The LG TVs may have their own set of flaws, but nothing suggests that this issue cannot be addressed from the soundbar end (PCM 7.1 works, other manufacters’ soundbars work, Sonos’ very own Arc works, etc).

Speaking of, a similar thing happened with the Bose 600. Basically at launch the Bose 600 soundbar had a no sound issue with LG TVs. Users even tried replacing the supplied HDMI cable to no avail. Until a redditor discovered that for some reason a ~$30 Spigen cable “fixed” the issue. Other users then tried and confirmed that it worked for them as well. But this development did not stop Bose from owning the issue and releasing an actual fix - a firmware update.

Of course if this isn’t actually a Sonos issue, they are free come out and issue an official statement. After all, it’s been 9 months since this has been reported.

I keep getting marketing emails for the Era speakers and just laugh each time as I doubt I will ever buy another Sonos product again after this.

Does anybody have any ideas about getting this picked up elsewhere, maybe the Verge or Peter Pee? Email to Patrick Spence? Does he still reply on Twitter?

I messaged Peter and he responded, but has his Beam connected to a different TV, so would need to move things around a bit to try and replicate the issue. He suggested trying a different cable for these sort of issues (which I had already tried). Fair play to him for at least responding though. 

To follow up on this, like others who’ve bought this cable, I can also confirm the issue is gone.

The issue had been gone for a while without me having to do anything, so I thought perhaps some update had fixed it. However today the TV received a software update and the issue returned immediately afterwards. I’m thinking software updates (as well as power outages) somehow “reset” the state of the tv, including reintroducing this bug.

Anyways, nothing I’ve tried so far worked.

It isn’t clear from your post, so I’m just checking - did you read the last page of posts where a ‘workaround’ was discovered? 

My oscilloscope only has a bandwidth of 30MHz which is inadequate for displaying even the TMDS clock signal (74.250MHz from the Beam2), however it does show some smoothed waveforms so is enough to detect gaps in the signal.

Whist I didn’t see any disruption on the TMDS clock, I finally managed to trap some physical layer evidence - on HEAC+ (pin 14). Unsurprisingly, the TV’s eARC output is showing disruption during the dropouts. Sadly this doesn’t really help narrowing down the cause more than any of the observations that have already been made on this thread.

HEAC+ is the upper waveform (red) and apparently is some S/PDIF-like format, the analogue audio is lower (yellow). Audio waveform was actually captured from a Sonos port line output in multi-room mode as this seemed easier than trying to hook up a microphone to the scope, so it won’t line up exactly:

I was wondering whether I could feed a test video signal in from the Raspberry Pi on the TMDS pins, but spliced with the eARC pins routed to the Beam2. This might isolate whether there's some problem in the Beam2 video stream or whether any video on HDMI2 will activate some buggy logic inside the TV (the Beam2 is still doing something different to the Arc, so maybe on the eARC control line).

If you mean the workaround with the cable then I’ve read it, unfortunately I don’t have such cable and Amazon doesn’t seem to have it available (or they’re not sending it here).

It may be possible to use a slither of insulation tape to cover pins 10 and/or 12 (TMDS clock) similar to others managing to block pin 13:

https://bonigopalan.medium.com/how-i-solved-hdmi-arc-problem-between-a-set-top-box-tv-and-sonos-arc-using-scissors-54e3d0f23ca6

Looks quite fiddly, but if the TMDS clock is blocked then the TV shouldn't respond to any video signal, which reduces the number of pins needing covering. It may actually be easier though to cover all the TMDS pins (1-12) as the extra width will be less fiddly.

I have a short HDMI extender (can't remember it it came with the Beam or not), which should save the TV/Beam sockets from ingesting insulation tape should it come loose, so may try this a bit later.

Not sure of the warranty situation of doing this sort of thing either...

It might be semi off-topic, but just a heads-up for anyone considering upgrading an Apple TV 4K Gen 3 to fix the LPCM 7.1 issue: the new model does output in LPCM 5.1 according to the Sonos app, but audio quality is noticeably worse compared to forcing Dolby Digital 5.1 on it. Especially voices sound more echoey somehow.

Masking pins 10/12 with a *sliver of tape does seem to have the same outcome, TV no longer recognises the video on HDMI2 (as per 11111 menu) and the dropouts are gone:

Just seen update 15.3 released - I can’t update until later however I have now changed to the fibre HDMI cable so won’t be able to tell if it’s fixed anyway. Anybody else able to check? 
 

I think if this isn’t it then it’s never getting fixed.

Just updating mine, will test shortly.

Set my PS5 to 5.1 linear PCM and drop outs as normal every 57s, so still not fixed in 15.3….

This problem just fixes itself randomly for me, but the moment the tv gets an update and does this kind of hard reset it returns, so I think the cause is in the TV. I tried talking with LG yesterday but all they offered me was to send a technician to check my TV and their local branch here in Israel said global LG has not notified them about this issue and it’s the first time they hear about it. They’ve suggested to me to contact the global LG on Facebook and try to get their attention. Maybe if we all try someone there will care…

Anyone tried an LG C3 with the Beam 2 yet? 

Oh forgot to mention, the LG support asked me for a video showing the issue, but it’s currently disappeared for me so I can’t send them. Any chance one of you could make a video recording of the issue occurring?

This is it happening on my C2.