PDA

View Full Version : SB3 crash



Millwood
2006-06-20, 18:35
My SB3 was playing an mp3 stream from an internet radio station. Suddenly, the output turned to what I might describe as static - not quite white noise. Stopping and starting did not change this. When I turned the SB3 off, it put a high pitched tone into my system. (Analog connect). Turned back on, still static.

I unplugged its power and replugged and it is working again.

I'm not even sure how to report this as a bug - I have no idea what caused it or how to reproduce it.

Has anyone else seen this?

seanadams
2006-06-20, 20:08
Please excuse the technical nature of this reply - I'm not sure how much of an answer you'd like. :)

What we know is that it is a rare occurrence where the data stream between the CPU and the audio output circuitry gets out of sync. When this occcurs, you should hear nothing from the s/pdif, and something resembling white noise from the analog outs.

Although we haven't been able to reproduce it, I suspect after reviewing the design that it is caused by a logic glitch when switching between 44.1 and 48KHz tracks. I am working on a fix and some experiments to confirm. Anyway, there is no reason to suspect anything physically broken in your SB3, and based on the infrequency that this has been reported, odds are strongly against it happening to you again. However, we are working on it and should have a better idea soon.

Millwood
2006-06-21, 05:55
Thanks for the update and explaination. I love to make believe I understand the technical stuff :-)

You are doing what I liked to call "debugging by thinking", one of the most rewarding activities when it works.

Jacob Potter
2006-06-21, 12:45
On 6/20/06, seanadams
<seanadams.29qf6n1150859401 (AT) no-mx (DOT) forums.slimdevices.com> wrote:
> Although we haven't been able to reproduce it, I suspect after
> reviewing the design that it is caused by a logic glitch when switching
> between 44.1 and 48KHz tracks. I am working
> on a fix and some experiments to confirm.

That's interesting; I got the White Noise Of Death a few days ago
right at the beginning of a track, but both tracks were at 44.1 kHz.
It stopped when I stopped the music (actually a soft power-off, but
same thing) and played perfectly afterwards, no power cycle needed.
This was on the analog outputs (I don't use the digital.)

First time it's happened to me...

- Jacob

MrC
2006-06-21, 12:59
FYI. The WNoD visisted us once a month ago. It happened as I was bumping the volume up and down via web interface while a song was playing (I was testing something from another room and bounced the volume to hear it).

All our tracks are 44.1, and we're using analog only.

seanadams
2006-06-21, 13:43
Interesting.... I guess that debunks my theory (as the sole possibility), so I'll pull at straws here: can you correllate it to an obvious electrical transient, eg lightnighg, or fridge, dryer, AC, or pool pump starting up? The reason this is so tricky to identify is that the CPU->audio interface is really pretty simple and there's not much that can conceivably go wrong. However, if we had a reproducible event, it would probably be trivial to fix.

MrC
2006-06-21, 14:00
Sorry, no I don't recognize any correllation w/power spikes or glitches. We have very clean power, as reported by my APC monitoring graphs, so wouldn't expect that.

One thing that might be useful, I wouldn't have described the problem as "white noise", but have followed the lead of others here; rather, what I heard was hints of the song boosted to such a high volume that it sounded more like massive clipping across the spectrum. This probably is of no use, but figured I'd share it.

Jacob Potter
2006-06-21, 17:48
On 6/21/06, seanadams
<seanadams.29rs0z1150922701 (AT) no-mx (DOT) forums.slimdevices.com> wrote:
> Interesting.... I guess that debunks my theory, so I'll pull at straws
> here: can you correllate it to an obvious electrical transient, eg
> lightnighg, or fridge, dryer, AC, or pool pump starting up? The reason
> this is so tricky to identify is that the CPU->audio interface is really
> pretty simple and there's not much that can conceivably go wrong.
> However, if we had a reproducible event, it would probably be trivial
> to fix.

I didn't notice anything at the time.

Just out of curiosity, what is it that gets out of sync to cause the problem?

(I have a 5 MHz analog scope that I could poke around with if it
happens again, but I'm sure there's some variant of Murphy's Law that
says it won't...)

- Jacob

tyler_durden
2006-06-21, 19:38
A few days ago my SB3 went all-noise for the first time since I got it in December. I was listening to something on the squeezenetwork at the time. One channel had wall to wall noise, the other a high frequency buzz. A soft reset (holding the power button on the remote for 4 sec) made it go away. It's been AOK since then.

If it happens again anytime soon I'll hook my 60 MHz HP scope up and see what it looks like.

There wasn't any "musical character" to the noise. It was just noise.

TD