TOSlink optical cable problem

[jacob]

Does anyone have experience with these things?

We have an old LG TV that's connected to a ZVOX 220 soundbar by an optical cable via the digital out on the TV. The TV is fed by a Roku.

This used to work. However, over the last year or so, the soundbar failed to detect the sound input from the TV for a few hours and would then spontaneously come back again. This happened every few months. Lately, it happened every few weeks, then every few days and now nothing.

The soundbar works fine with the RCA inputs from another source. Unfortunately, the only sound output from the TV is the TOSlink and I have nothing else that takes that as an input. Nor do I have another optical source.

I read that optical cables break easily. I also read something about a red light, but it was not obvious whether that was just the optical cable or whether there's a red dot on the cable when it's working. There is a red light inside the output port. There's also "red" coming through the cable at the cable-end going into the zvox input.

The only other issue I came across was that ZVOX demands the signal to be in PCM format. This should be settable from the TV. However, our TV does not know what PCM is. This setting is also not available on the Roku. All I could do was to switch from "Auto" to "Stereo". The theory here would be that Roku somehow had a software update that fouled the protocols. But this does not explain the intermittent->permanent behavior of the glitches. That sounds more like a bad connection.

Normally, I'd just try to isolate by switching sources, cables, and output devices, but in this case, since optical cables/connections are kinda rare, I can't.

Any ideas?

[Riggerjack]

The fiber optics I played with were custom spliced glass telecom fibers.

The opinion at the time from such "pros" was that the audio-visual fiber was a large diameter (comparative) cheap plastic manufactured product, and the expected failure would be that the plastic would degrade, and eventually signal loss would overcome the large signal loss the short transmission distance allowed. Such cables had been tried, of course, they were cheap. But they slowly gained attenuation with time.

If this were in fact the case, I would expect it to fail exactly as you described. A weak signal making an amplification threshold, or not, under nearly identical conditions.

I don't know anything of value about these AV fibers, but I bet some AV forums will have someone who has dealt with this. Audiophiles keep equipment, forever.

[jacob]

The cable is 10+ years old and looks very cheap/sketchy. I still see red light coming through it, but I have no way to gauge/know if it's bright enough.

The cheapest attempt would be to replace the cable. They're $3-8.

[fingeek]

Audiophiles would say there's a qualitative difference in the optical quality. Personally, I can't see that that's the case as the bitrate rate is low, as is a CD/BluRay ultimately (vs a 10Gbit fibre, for example).

That plus the error correction led me to conclude that a cheap fibre is good enough.

At least, it's good enough for my ears - And I get irritated intensely by any hint of autotune or (dynamic range) compression.

I also suspect the fibre as the most likely degraded point. It's possibly dust or a scratched lens, but unlikely if you've kept the fibre in consistently. It's possibly a degraded laser but I'd expect that to me more all-or- nothing when they fail.

[Sclass]

Audiophiles would say there's a qualitative difference in the optical quality.

Audiophiles say ridiculous things. Like they can “improve” CDs with green sharpie markers. It’s digital. 1’s and 0’s can take a lot of abuse before they become unintelligible. And there are no lasers, it’s just an LED over a fiber that’s more like a nylon fish line. If you see red light coming out of the connector it’s good. That’s the whole idea behind the system. They know a typical consumer has no way of knowing the cable is good so Sony built in a cable test. These are forgiving systems. It’s not gigabit Ethernet.

Okay I just aimed my Toslink TV audio cable into an LED and probed the LED with my scope. I can just make out the data pulse in the noise. It’s less than 1 MHz. The LED is a poor man’s photo diode centered at 650nm which is close enough to receive the red pulse from Toslink. If you have an oscilloscope and red LED you can likely see your signal. I can barely make out an edge in mine likely because I don’t have a collimating lens. Bare fiber into a detector doesn’t really work well…you need a lens.

An easier method is to buy a $2 DLT1150 Toslink receiver and plug it into a breadboard. It has a lens and amplifier all built into the female connector. Power up its integrated amp and read out the digital signal on a scope. I think SDIF protocol is NRZ IIRC so you’ll have to do clock recovery if you want to see it on an Arduino serial port.

Alternatively take your soundbar apart and probe the Vout pins on the Toslink receiver and see if you can pick up when a helper disconnects the cable at the TV. This will test your receiver.

These are very robust and forgiving systems. You can make the cables with scissors unlike glass fiber. Your problem is likely in your CODEC in your soundbar or the TV.

[fingeek]

LED? Thank you, TIL

[Sclass]

Yeah. When was the last time an LED burned out? They are really tough. So are the cables that I believe are just 50 lb test nylon fishing line with a black jacket over them. So the bad news is that the problem likely lies elsewhere in a more delicate part of the system. Unless a spider made a nest in the socket. But it looks like they anticipated that in the design of the receptacle. It’s bulletproof for consumer electronics.

I happen to have the same LG TV with the crummy speakers that needs a soundbar. The sound quality of the integrated speakers is so bad we couldn’t understand a lot of the dialog in movies. My wife wanted the soundbar but when I saw the price I said no way. I’m too cheap. I bought a $5 SPDIF to analog converter on eBay. It’s a little box that you plug the SPDIF into and power using a USB port. Out the other side is stereo analog on RCA jacks and a headphone socket which I plugged into some old PC speakers. Problem solved for $5.

I guess I missed the whole THX surround sound thing for TV years ago. I just wanted a TV to plug in and watch. It seems they design them with cheap speakers assuming everyone will hook them up to fancy audio systems. Perhaps it’s a consequence of making the screen so flat.

[jacob]

Well, it's not the cable @#$%@#%@ (replacing it didn't make a difference)

I'm going to haul the soundbar over to a neighbor and try a different TV. That should tell me whether it's the TV output or the soundbar input. If it's the input, I'll get the converter. If it's the output, someone else probably just scored a free TV.

@Sclass - One of the great things about the zvox (<$100 on craigslist) is https://zvox.com/pages/accuvoice

[Sclass]

If it's the output, someone else probably just scored a free TV.

@Sclass - One of the great things about the zvox (<$100 on craigslist) is https://zvox.com/pages/accuvoice

Not so fast. As you theorized the Roku can be messing up the sound track. My LG TV has a Bluetooth audio out as well. There are a ton of super cheap Bluetooth to RCA adapters boards out there. They should cost less than $2. Just another way to get audio out of your TV.

That ZVOX is actually pretty cheap. I just looked at their Bluetooth model and it’s only $70. I’m tempted to get that for my workshop. Right now I use a Bluetooth receiver board ($2) connected to some junk pc speakers from a yard sale to listen to the news while working.

If the Roku is working properly my gut feel is the problem is likely a connection issue. Consumer electronics rarely blow out semiconductors if they are designed properly. After manufacturing and testing thousands of boards the most common culprit is bad solder joints. A sound bar may actually see a lot of fatigue cycles from vibration. All it takes is one digital line in the ZVOX with a cracked solder joint to stop it from working. The fact it flatlines instead of making noise suggests something like a serial data bus has become disconnected. Just a feeling. It may just need a single solder joint reheated.

The probability it’s the digital audio output on the TV is unlikely given what it’s made of. LG’s design practices are not bad given what I’ve seen in my washer and TV. That stuff can potentially last forever again if it is designed correctly. My gut feel is the soundbar has shaken something loose. It may be a tiny SMT pad and very hard to locate but sadly it may be something that minor. One technique I use to find such joints is get it all apart on the table with the sound going through it and gently tap every component on the board with a yellow pencil with a rubber eraser. Listen for the anomaly.

ETA - yeah the more I think about it I like the fatigue induced solder failure in the soundbar theory. That box sees bazillions of stress cycles as the sound waves ripple through it. Likely a weak solder joint just gave up.