Avoiding computer power-up surges passing through USB audio I/O to speakers?

[n13l5]

There are many USB and Firewire audio interfaces, that have no circuitry to protect your speakers from the power surge when the computer gets turned on or off, to pass through the audio I/O system and on to your speakers, creating loud pops that can blow the tweeters of expensive monitors.

Matter of fact, the audio interfaces that have such protection seem to be in the minority. Because of this, I'm too scared to use my monitors with my newly set up recording system.

Since turning on the stereo never makes your spakers pop, and there's also a pretty cheap Logitech USB audio interface+speakers combo that never does pop on cycling the computer's power switch, I know there must be some way to prevent this by adding the right kind of circuitry.

While I'm not bad at soldering together circuits, I just can't figure out where I could intercept this.

Would be nice if it was something you could just build into the speaker cable, but if I have to open up the device, that's fine too...


Trying to logically figure out where the pop originates, I already fail. The audio device takes a digital audio stream from the computer's USB port. I doubt if the pop comes from there, it seems unlikely that the computer would create a pop and convert it to a digital stream to pass to the audio device...

So it seems more like the USB device's power supply must be lacking some form of protection? But then, the computer isn't very inductive like a fan or refrigerator, so it shouldn't really be sending a surge through the whole power circuit in the house??


If anybody has an idea where to attack this problem, I'd much appreciate it :)

 

[westom]

Matter of fact, the audio interfaces that have such protection seem to be in the minority. Because of this, I'm too scared to use my monitors with my newly set up recording system.

There is no protection on most devices because that loud noise is not a surge. It is only noise as destructive as loud music. Circuits to avoid that pop are to make you feel better; nothing more.

A most common way of doing this is to put a relay on the speaker wire that does not close for a few seconds after power is turned on. That does not work on USB, et al audio devices that are already powered when other items are connected or powered on.

If it was a surge, the much larger current could be detected and cut off. But that will not work. Because the pop is not greater than any other 'pleasant' sound. Other pleasant sounds would also be cut off.

Speaker must be large enough to handle any current output by the amp. If current is too great, circuits already inside amp will limit it to protect the amp. Your have invented a hardware threat that does not exist because you do not like the sound.

 

[n13l5]

Thank you much for your response, some interesting points...

There is no protection on most devices because that loud noise is not a surge.

That may well be the case. However, a signal producing the pop gets generated somewhere. where?

It is only noise as destructive as loud music. Circuits to avoid that pop are to make you feel better; nothing more.

That statement is quite incorrect:
1) any recording studio technician has it drilled into them that speakers get turned on last and turned off first on power down. For good reason, cause tweeters easily get destroyed by potential pops generated.

2) you can even destroy your tweeter by playing back certain synthesized wave forms that are hard for the tweeter to render, overheating it after a short time, without playing any louder than you could play normal signals on it for hours.

So, just because its 'merely' an audio signal doesn't mean its not dangerous to your tweeters.

A most common way of doing this is to put a relay on the speaker wire that does not close for a few seconds after power is turned on.

This seems to be the solution that they use on most hifi gear (amplifiers, receivers) on the audio outputs.

That does not work on USB, et al audio devices that are already powered when other items are connected or powered on.

To make it work, the USB device would either have to be the variety that actually gets powered through the USB bus, or on those not powered from the USB bus, would need your aforementioned relay to also be hooked to USB powerline, so it basically detects when the computer gets turned on (since the USB powerline has no power with the computer off)

This leads me to believe that my next step should be to find out how to properly implement such a relay; to not close for 30 seconds after power is detected, as well as not close at all if no power is detected.

If it was a surge, the much larger current could be detected and cut off. But that will not work. Because the pop is not greater than any other 'pleasant' sound. Other pleasant sounds would also be cut off.

Yes, that makes sense, the cutoff would have to be only activated for so many seconds after power was first detected.

Speaker must be large enough to handle any current output by the amp. If current is too great, circuits already inside amp will limit it to protect the amp.

You would imagine on expensive active studio monitors that have individual amps for each speaker built into the speaker cabinet, they would match the amps perfectly, (which is usually what the advertising brags about) So by your logic, nobody should ever need to replace any tweeters, other than to fatigue from old age. This isn't the case though. I've lost both tweeters in Alesis active studio monitors, due to the nasty pop/cracle from my dell laptop on reboots.

Your have invented a hardware threat that does not exist because you do not like the sound.

Trust me, the hardware threat exists... ask anybody in any recording studio / television studio around the world.

 

[westom]

That may well be the case. However, a signal producing the pop gets generated somewhere. where?

First, what changed when the pop occurs. For example, when a signal producing device first powers on, all active components are operating in non-linear regions. Non-linear is how Kettering car ignition created hundreds of voltage from a 12 volt DC battery. Non-linear is how oscillators work. Non-linear is how harmonics are created. All those can be unpleasant sounds. Then DC voltages stabilize. And electronics begin working in their linear regions.

Another non-linear affect occurs when a cable makes or breaks. Two examples of how a pop might be created.

Speakers on an amp must be sized to take any power the amp might output. Amps are further limited internally so as to not self destruct. If any output from the amp destroys a speaker, then the speaker was too tiny for that amp.

Recording studio techs are taught to turn on speakers last so that nasty sounds do not harm human ears. If any tweeter is destroyed, the amp was too large for that tweeter. And yes, many audio techs do not know how to size speakers. It worked today, so it must be sufficient?

If a tweeter is too small, then better designs can include circuits to clip peaks or disconnect based upon a 'moving average' of too much power. Disconnect before tweeter damage could happen. But often, its just easier to blame the studio tech rather than an inferior or insufficient design.


You might do a search on the term 'time delayed relays'. But I suspect most will be 24 or 120 volt types. I have done this using 555 timers. It creates a weird effect. Without warning, suddenly a voce or music appears as sharply as a pop. It actually startles people even when they know the sound is coming.

 

[n13l5]

Thanks again for the further details...

Speakers on an amp must be sized to take any power the amp might output.

This would ideally be done everywhere, and especially in the design of self contained active monitors...

And yes, many audio techs do not know how to size speakers.

It appears the manufacturers of active monitors don't know how to size speakers either...

I'll wager that, given the constraints of competition and price sensitive customers, almost no active monitors are designed 'properly' by that standard. I should possibly call KRK and see what they say in that regard... Maybe they did it better than Alesis.


What I'm looking to do is take things in my own hands and fix at least the effects of the improper design... so I'll do a search for the relay circuits you mentioned.

Thank you for the pointers!

 

[westom]

What I'm looking to do is take things in my own hands and fix at least the effects of the improper design... so I'll do a search for the relay circuits you mentioned.

Time delayed relays are a single part. Sold in places like digikey.com, alliedelec.com, newark.com, mouser.com, and jameco.com