Linux midi blew up my subwoofer!!!!!

SamuraiDan · 2005-10-12 16:51:31

I set up midi today so my girlfriend could use her Electric Piano with stuff like Rosegarden for composing etc. It was all working nicely as far as I could tell, I played several files and all was good.

I proceeded to go to the shop, leaving the box on, speakers about a quarter volume, nothing playing. When I got back an hour later, I have no sound at all. Hmm I think, mess around a bit with the mixer and shut down the box at which point hear a sort of loud feedback hum. Eventually narrow the problem down to the sub woofer box which I've just opened and seen one of the chips has blown :shock: (cant see what it is yet as need to open more things).

So is it related to alsa and installing midi, or just a horrible coincidence?

I'm also a bit worried about the output levels of the sound card as they seem very low, but I can get output with headphones/other speakers (not sure how to test this). If you are interested it is an Audigy (1) sound card and Cambridge Soundworks Desktop Theatre 5.1 DTT2200 speakers. The packages I installed for the midi were "alsa-utils" and "awesfx" from AUR.

Opinions?

phrakture · 2005-10-12 17:26:05

Um, no offense intended, but how did you diagnose a chip as "blown". Most components of that nature are able to handle a decent amount of power before they pop (and they do *pop*). Usually, due to signal clipping, the power can't reach these stages. It's a safeguard. The usual way something like that happens is from the +-limit values, which are controlled by the wall outlet input and the "volume" (or whatever it's called on a subwoofer) knob.

Basically, my point is that, if you're going to blow a subwoofer, it's usually not due to the signal being amplified, but the power input from a wall socket (or whatever) - your sound card would probably fizzle out before it was able to produce enough power to kill you subwoofer.
Also, if you're saying the sound levels are low to begin with, that's an indication that the output power is low as well.

SamuraiDan · 2005-10-12 18:17:51

>Um, no offense intended, but how did you diagnose a chip as "blown". Most components of that nature are able to handle a decent amount of power before they pop (and they do *pop*).

None taken. I diagnose it as blown by the lovely black scar on the underside of the PCB. I'll post a photo if you like... And now Ive got the board out I can see its only a Cap. Trouble its' a surface mount one without value markings... Trust me I know about ICs going pop - I think ive had one hit the ceiling before (due to my own mistake).

>Usually, due to signal clipping, the power can't reach these stages. It's a safeguard.

You cant really make this statement or assumptions without seeing the schematic in question.

>The usual way something like that happens is from the +-limit values, which are controlled by the wall outlet input and the "volume" (or whatever it's called on a subwoofer) knob.

No offence intended to you this time but you cant make these statements without knowledge of the device. I have an degree in electronic engineering and I know without the relevent circuit layouts and datasheets its very hard to diagnose something like this from sight, let alone from a short description.

>Basically, my point is that, if you're going to blow a subwoofer, it's usually not due to the signal being amplified, but the power input from a wall socket (or whatever)

There is an external transformer as in most of these kind of devices which would go first if it were a mains problem. As I had it on a surge protector I doubt it is the fault of the mains .The transformer is meant to output 13.5V in this case (which is a bit odd) and. The transformer is now putting out a steady 17V which although high is probably within tolerance of the internal regulators- if that had been a problem I would have expected the reg to blow - which incidently is what I though when I first opened it.

>your sound card would probably fizzle out before it was able to produce enough power to kill you subwoofer.

Well this is where I am not sure. Maybe the subwoofer could be killed by bursts at a high enough frequency (although this does seem implausable).
But the point is it may have gone too - see below - and took the woofer with it. Maybe, just maybe something to do with the midi software triggered it all off. And I did hear some short random hisses and pops from the speakers initailly but thought nothing of it.

>Also, if you're saying the sound levels are low to begin with, that's an indication that the output power is low as well.

I probably wasnt clear here. I mean that the sound levels seem low after this happening. But as I didnt measure them before it happening there is no way to know that I can think of.

When I have time ill swap out the cap with a pot (variable cap) and power it on a nice supply and see if I can make it sing again

Moo-Crumpus · 2005-10-12 18:22:05

Aliens ate my buick.

SamuraiDan · 2005-10-12 18:40:00

Moo-Crumpus wrote:

Aliens ate my buick.

Yeah mine too..

Here's the photo:

phrakture · 2005-10-12 19:28:32

SamuraiDan wrote:

>Usually, due to signal clipping, the power can't reach these stages. It's a safeguard.
You cant really make this statement or assumptions without seeing the schematic in question.
>The usual way something like that happens is from the +-limit values, which are controlled by the wall outlet input and the "volume" (or whatever it's called on a subwoofer) knob.
No offence intended to you this time but you cant make these statements without knowledge of the device. I have an degree in electronic engineering and I know without the relevent circuit layouts and datasheets its very hard to diagnose something like this from sight, let alone from a short description.

Aye, I never said "this is always the case", I simply said "usually". Without getting into a discussion outside the context of this case, subwoofers are nothing more than a low pass filter with an optional amplifier on the end. Both of these can be sufficiently modeled with an opamp, or at a finer grain, a big-ol' BJT schematic (or FETs if you paid the extra cash for a MOSFET subwoofer, who knows) - either way, amplifiers/LPFs are limited (Vcc/-Vcc on the OpAmp) and can take pretty high voltages in the normal case. FETs tend to handle high voltages mush easier than BJTs, which break down just like normal diodes, though I highly doubt low breakdown voltage BJTs would be used in audio applications.

Anyway, I was simply stating that, in most cases it is the limiting voltage which will cause the burnout, due to the fact that it can amplify the signal well past the threshold for the transistor.

SamuraiDan wrote:

>your sound card would probably fizzle out before it was able to produce enough power to kill you subwoofer.
Well this is where I am not sure. Maybe the subwoofer could be killed by bursts at a high enough frequency (although this does seem implausable).
But the point is it may have gone too - see below - and took the woofer with it. Maybe, just maybe something to do with the midi software triggered it all off. And I did hear some short random hisses and pops from the speakers initailly but thought nothing of it.

Hmmm, I cannot recall all the transistor info that once polluted my brain, but I don't think high frequencies will affect most applications. The main reason I say that is I've never seen a threshold frequency specified (example (in pdf)).

Anyway, more to the point. You say that *now* your soundcard is crapping out. If you're feeling brave, try "cat /dev/urandom > /dev/dsp" (dsp probably isn't right anymore, but there should be *something* similar under ALSA) and see if you can get it to produce some decent sound. I've done this for about 10 minutes before with no problems, so I don't think you should be concerned about constant output.

SamuraiDan · 2005-10-12 21:24:11

Well the thing wasnt that expensive. You're right though, all it consists of is a bunch of JFETs and Op-amps. Besides its a cap thats popped. If you have any suggestions as to why then thats great - always good to come at a problem from different angles.

And I don't really blame ALSA midi - it was just the coincidence that made me post this plus the irratation of going though my system+ALSA for ages only to find out it was the hardware - and I thought it may amuse!

> try "cat /dev/urandom > /dev/dsp"

This was actually one of the first things I did try (you do indeed have the devices right under udev and ALSA) and made me unsure about the gain on the output signal.

Arch Linux

#1 2005-10-12 16:51:31

Linux midi blew up my subwoofer!!!!!

#2 2005-10-12 17:26:05

Re: Linux midi blew up my subwoofer!!!!!

#3 2005-10-12 18:17:51

Re: Linux midi blew up my subwoofer!!!!!

#4 2005-10-12 18:22:05

Re: Linux midi blew up my subwoofer!!!!!

#5 2005-10-12 18:40:00

Re: Linux midi blew up my subwoofer!!!!!

#6 2005-10-12 19:28:32

Re: Linux midi blew up my subwoofer!!!!!

#7 2005-10-12 21:24:11

Re: Linux midi blew up my subwoofer!!!!!

Board footer