Bugzilla – Bug 9262
Power flicker triggers Boom factory test tone
Last modified: 2009-09-22 08:58:08 UTC
We just had a series of power flickers. The power did not go out completely it just flickered off and on a bunch of times. Then my Boom (not sure if it was one or both) emitted a strange tone that sounded like musical scales of 5 ascending notes. (almost like Close Encounters!) After that both Booms displayed "Obtaining IP Address..." on the screen and appeared to be stuck there. I went through network setup on both of them and they both appeared to be stuck on "Connecting..." Then I did a factory reset on one of them and it still appeared to be stuck on "Connecting..." During the setup after the factory reset, it emitted the musical scales tone again. I went to the forum to post this experience, and when I checked the Booms after several minutes, they both had the clock on screen and then worked fine. Not sure how long it took them to eventually connect, but more than 5 minutes and less than 30, I'd guess. It was suggested I file a bug report for this since having the factory test tone sound if there is a power flicker during the night would be an unpleasant experience.
More info... The Boom that I am certain emitted the tone was on a surge protector, and is one of the older hardware revisions. I don't think the newer Boom made the sound because it was in the opposite direction when I heard it. Unless they both sounded the tone at the exact same moment, it was probably only the one. So possibly it is something that only happens in the older hardware but I can't be sure.
Caleb: what do you think?
We need to set up a line simulator to induce line transients at the high voltage side, and see what kind of problems we can induce. Perhaps they have a simulator in Fremont we can use. Alternatively, we could just rig up a relay and flick it on and off quickly. I'm marking this as 'Future' since I don't see this as a critical issue for MP. -Caleb
Caleb: Anything that was fixed in the later builds that could explain why sue only saw it on her earlier unit?
Not sure. The input protection circuitry did change from the older units to the newer ones. Also, the bulk capacitors are bigger on the newer units. So now that I think about it, I'd say yes, it's quite possible that those differences could explain it. -c
I was woken up by the fw alarm tone this morning, but my Boom appeared fully connected to SC. The fw alarm tone went off at exactly the alarm time. See http://forums.slimdevices.com/showthread.php?p=332201#post332201 for more info. I wondered if this could be related in any way, but I don't think we had any weird power behaviour.
Felix, check out comment 6... Any idea?
I've since realised that I had two Booms with identical alarms set, so at 7:30 this morning two Booms in different areas of the house started playing. My wireless network isn't that hot, so it may have struggled a bit with this. Could this have caused one of the Booms to disconnect from SC for a very short period of time, and hence trigger the RTC? The odd thing though is that I saw absolutely no hint of the Boom being disconnected on the display.
Well, if you have your time format set identical for when connected to SC and when offline, you probably would not notice a very short disconnect.
I don't think it is identical - the clock normally looks very different when it appears on an SC restart. Anyway, I'll watch out for it happening again and report back.
Both Booms (PQP1 and PQP3) went off at 7:30 this morning, as they did yesterday, but this time I was not woken by a test tone on my PQP3. Instead, Radio 4 played as intended. Hopefully yesterday's behaviour was just caused by a temporary drop in connection and me being too asleep to notice what was going on. I'll report back if I hear the test-tone again, but otherwise let's just assume it was a red herring. Sorry for the confusion!
Is anyone still seeing this issue? if not I'd like to close the bug.
I'm sure Elliot has some kind of power transient testing if Fremont doesn't.
bug 11010 could be related to this, same error, different symptom
We haven't see this behavior in ages.