• Blown Tweeters

    Few audiophiles are aware of a lurking danger that may be very costly. I've know about this danger for several months now, but was re-energized to write about it over the weekend. I was listening to the Raconteurs through the new Bryston BDA-1 DAC at a pretty high volume. My music server was booted to Windows Vista Ultimate 64-bit. I stopped playback for a second, made a few changes to the Windows audio settings and hit play. As soon as I hit play a very loud tone blasted through my speakers and sent Chloe, one of the Computer Audiophile cats, running away very puffed up. I jumped up to turn down the volume on my preamp. My heart was racing and I couldn't stop saying to myself, "What the heck just happened?"
     

     

    The Danger

    Put simply, you can blow your tweeters using Windows if you're not careful! Fortunately my speakers survived the weekend. On one hand I totally knew what had happened, on the other hand I couldn't believe how it happened. The whole problem revolves around losing clock which causes either white noise to blast through the speakers or a tone like the emergency broadcast system here in the United States. Two verified instances that I know of happened when listeners tried to changed the name of a song while listening to music. My friend using TAD M1 loudspeakers was listening through MediaMonkey on Windows XP. He tried changing the name of a song and a blast of white noise came through the speakers. The very expensive TAD tweeters were blown. This same friend has reproduced the problem, at much lower sound levels, through many different procedures. In my opinion doing anything other than listening to music is dangerous. If you want to change settings in Windows or your playback application you need to be very careful. Another instance I am aware of where someone very respected in the high-end industry almost blew his tweeters involved Windows Media Player. The person changed the name of a song, lost clock and briefly experienced the white noise issue. The people I am talking about are not novices when it comes to this stuff. They are recognizable names in the audio world.

    Over the weekend I was testing Windows Vista Ultimate 64-bit with a number of applications and configurations. After several changes, that I certainly did not keep track of, I tested the output of Windows audio through my Lynx card. I heard a small blip instead of the typical chimes in the left and right speakers. By this time I was knee-deep in all the changes and was not thinking about the possible danger to my speakers. I started MediaMonkey with an ASIO driver. As soon as I hit play I heard the aforementioned very loud tone. This circumstance was vastly different than the previous two where the listeners tried to change the name of a song during playback. I have also heard of similar issues when using Windows Remote Desktop, but I haven't been able to reproduce the issues (fortunately).

    No matter what the causes of this issue are we can be certain that a number of listeners will damage their speakers this way. If the cause is user error, a software bug, a firmware problem or even by design this is a serious problem. In fact this issue may be one that old school audiophiles use to keep music servers out of their listening room or retail stores.

    It would be fabulous if we could find a definitive cause and spread the word about how to avoid damaging loudspeakers. I hope many of you can post comments about similar experiences and methods to avoid the potential damage.
    Comments 31 Comments
    1. Vincent's Avatar
      Vincent -
      I have had this happen 3 times using benchmark usb Dac and Alpha Dac running XP Pro. Some action in the OS does something to lose lock and white noise appears at the speakers. No blown tweeter yet. I'm careful to mute the system now when changing songs, albums, etc. Turn off any background activity like antivirus, internet activity, auto updates, etc. Otherwise you will be buying tweeters. I keep the vol control in my hand almost all the time to hit mute if it happens. It is a real problem and I saw the TAD 1 speakers that had the blown tweeters.<br />
      <br />
      Vince Fennell
    1. Poo's Avatar
      Poo -
      This alone was reason enough for me to buy my Mac Mini. Had the same thing happen using XP Pro, researched it and decided my system deserved better! Replacing a computer is nothing compared to the cost of new speakers...
    1. perborgen's Avatar
      perborgen -
      Hi,<br />
      <br />
      I am all new to computer audio and have no mediaserver / player in my system (yet). I am really impressed with your site and your enthusiam, Chris, and find your reviews and the comments from folks here very valuable. Look forward to dig into to this new world of audio. When I read about white noise / bursts possible of taking out tweeters I was kind of worried. What about Linux? Are there any risks getting unsuspected noise running Linux?<br />
      <br />
      Best Regards,
    1. The Computer Audiophile's Avatar
      The Computer Audiophile -
      Hi Guys - Thanks for the comments thus far! I feel like this is a support group of people coming out of the woodwork (ha, ha, ha). It's strange nobody has really discussed this potentially expensive problem. <br />
      <br />
      Vince F - Right on, Tim's TADs are the exact speakers I was talking about :-)<br />
      <br />
      Per - I've yet to hear anything like this on the Linux side, but I haven't used it much. One of my partitions runs openSuse11, but the audio I get out of it so far is very substandard. I'm sure I'll just have to work out the kinks because Linux itself is an ideal music server platform. Please let us know if you go down the Linux path!
    1. phunge's Avatar
      phunge -
      I always listen to music on my headphones while i'm working on the computer. I've had several cases where windows will emit a very loud beep that just about blows your eardrums and leaves you very disoriented.<br />
      <br />
      Now I always make sure that the device i'm using for playback is not the default sound device for windows. Even better is to use playback software capable of using vista WASAPI output in exclusive mode.
    1. carloscarr's Avatar
      carloscarr -
      I've never had this problem and I use applications (Foobar, XMplay) that use Vista's Wasapi exclusive mode. So that would imply that no Windows sounds or sounds produced by other apps will be heard (also I disable Windows sounds in the Control Panel). But doesn't the Lynx card that Chris uses use an ASIO driver that would allow equivalent exclusion of any sounds not from the audio playback application?
    1. The Computer Audiophile's Avatar
      The Computer Audiophile -
      Hi Carloscarr - Thanks for the comment. Using an ASIO driver with the Lynx card does not stop the sound from happening unfortunately :-(<br />
      <br />
      More careful experimentation is necessary.
    1. audioengr's Avatar
      audioengr -
      Some DAC designs neglect to put some type of muting function in place for the USB interface. This is critical. Even more critical for a direct I2S interface. This typically mutes the D/A output when the signal changes or is not valid. Initially Benchmark DAC-1 USB had a funky implementation of this, but when John Atkinson got hold of it and started doing measurements they were squirrely and it reflected in the review. As a result Benchmark removed completely the muting function for USB. Bad idea IMO. Most designers are used to this function being handled by the S/PDIF receiver chip. Well, there is no S/PDIF receiver chip when doing USB audio, so I believe most dont know what to design, or they just ignore it.<br />
      <br />
      All Empirical Audio DAC's and mods have muting function correctly implemented for USB. Also, we teach our OEM's how to do this properly as well.<br />
      <br />
      Steve N.<br />
      Empirical Audio
    1. weiss2496's Avatar
      weiss2496 -
      The new generation Minerva units have muting built in. Whenever the sampling rate changes, the output is muted until the unit is locked to the new rate.<br />
      Daniel<br />
    1. VincentH's Avatar
      VincentH -
      Hi Daniel, great to hear about the built-in muting on the Minerva. I wasn't aware of that, my demo unit did not have this. Are there any plans to include a volume control in future Minerva generations as well? That would put the Minerva back on my upgrade path.<br />
      <br />
      Cheers, VincentH<br />
    1. weiss2496's Avatar
      weiss2496 -
      Hi Vincent <br />
      No plans for a volume control in the Minerva so far. <br />
      I'd expect the (future, highendish) player programs to offer a decent digital volume control which is properly dithered. Dithering means that it can preserve the perceived resolution over a wide volume range. With such a volume control and with today's hires D/A converters it is possible to have like a 30dB range of control without noticeable degradation. 30dB is plenty for your everyday volume control.<br />
      Daniel<br />
    1. VincentH's Avatar
      VincentH -
      Hi Daniel, 30 Db would really not be enough in my situation, but maybe it's not so everyday ;-). <br />
      When I used the Minerva with the Focal Twin 6 Be and the Adams P33A's I needed 50 to 60 DB attenuation to get a moderate listening level (= something the neighbours would not complain about).<br />
      <br />
      Would 50 - 60 Db of digital attenuation result in a noticeable degradation?<br />
      <br />
      Triple-amped active nearfield monitors like the Focals Twin 6 Be's and Adams P33s have unattenuated 700 - 800 Watts (if you add a sub even more), but they are for nearfield use (1-3 meters ear to driver) and small spaces. I imagine more people will find themselves in similar situations.<br />
      <br />
      Of course one could always purchase a high-quality analog passive attenuator like the GoldPoint SA1X, but that would add extra interconnects and cost...
    1. weiss2496's Avatar
      weiss2496 -
      In the new Minerva we lowered the output level by about 7.5dB.<br />
      But if you needed 50dB of attenuation in your setup with the Minerva, that would mean that the sensitivity was like -32dBu which is about 0.02V or 20mV. Very sensitive. Such low levels are obviously required in order to tame the 700 Watt potential to a sensible listening level...<br />
      In that case you would go for an analog attenuator (can be a fixed amount) to get the digital level control into a sensible range.<br />
      Daniel
    1. VincentH's Avatar
      VincentH -
      Thx Daniel, that output level reduction is another plus for me. It's good to hear that you keep refining the Minerva.<br />
      (Btw my apologies to all for slightly off-topic, I'll try to control myself now...)
    1. David_H's Avatar
      David_H -
      Thanks for bringing up an important concern, Chris. I have been slowly making the conversion from media based to computer based audio for several months. Fortunately I set up on a Mac and have not had any issues along these lines. Have you heard of anything similar in the Macintosh realm? I was considering using an older PC for use as a second server but I have too much respect for my speakers to risk this kind of result. I find it disconcerting that both XP and Vista can create this kind of problem. Hopefully your cat has forgiven you by now. <br />
      <br />
      http://wanderingdavid.wordpress.com/
    1. The Computer Audiophile's Avatar
      The Computer Audiophile -
      Hi David - So far nothing "lethal" on the Mac side of music servers. The only weird thing I have witnessed and talked to others who have had the same experience is the audio playing back at about 8x speed. I call it the Chipmunks sound. It's only happened to me a couple times and this was a few months ago.<br />
      <br />
      The cat has started to venture back into my listening room. Slowly but surely she'll be back in here for some good tunes.
    1. audiozorro's Avatar
      audiozorro -
      This is a critical flaw and a show stopper for many interested in computer audio.<br />
      <br />
      “No matter what the causes of this issue are we can be certain that a number of listeners will damage their speakers this way. If the cause is user error, a software bug, a firmware problem or even by design this is a serious problem. In fact this issue may be one that old school audiophiles use to keep music servers out of their listening room or retail stores.”<br />
      <br />
      I assume this issue is repeatable and just not random. So why isn’t there an acknowledgment of this issue from Microsoft technical support? Are they aware of the problem? Does this problem persist in the newly announced Windows 7? Does it also occur with older versions of Windows such as Windows 98 or Windows 2000?<br />
      <br />
      I thought it pretty pathetic that Apple has not addressed the problem of 24/192 output from their toslink connection jack but this Windows problem is serious. Some of us have exquisite speaker systems such as Nestorovic speakers that cannot be replaced.<br />
      <br />
      If I used a Windows computer for computer audio I would be most appreciative of a big bold warning such as<br />
      <br />
      WARNING FOR WINDOWS USERS:<br />
      <br />
      1. DO NOT …<br />
      2. DO NOT …<br />
      3. NEVER …<br />
      4. UNDER NO CIRCUMSTANCES …<br />
    1. The Computer Audiophile's Avatar
      The Computer Audiophile -
      Hi Audiozorro - I agree that this is a showstopper for many people and should be taken more seriously by all.<br />
      <br />
      I think there will be a lot of finger pointing by software and hardware vendors and none of them will acknowledge this as their problem. I haven't tried any of the really old operating systems like 98 and 2000.
    1. carlseibert's Avatar
      carlseibert -
      This is a bit scary, simply because it isn't part of my normal routine. I wouldn't want to casually plug in my work laptop into the system one day and blow a speaker or headphone diaphragm (not to mention my head) to bits. <br />
      <br />
      So, do I have this right that SPID/F transceivers automute if they lose the clock signal and are thus immune to this phenomenon? And that some USB implementations will do this and some won't? What about TOSLINK? <br />
      <br />
      Chris, I assume that with all the USB DACs you've tested, you haven't seen behavior like this if you, say, pull the USB cord when the device is playing, right? <br />
      <br />
      In the meantime, I'll try to remember to be careful with Windows. Egad. <br />
      <br />
      -Carl<br />
      <br />
    1. The Computer Audiophile's Avatar
      The Computer Audiophile -
      Hi Carl - Yeah, this is a little disconcerting. <br />
      <br />
      I can't say whether all S/PDIF transceivers auto-mute or not or if any of the other types auto-mute without a solid engineering decision to implement this feature.<br />
      <br />
      As long as you pull the USB cable there is no sound going to the audio system so you couldn't blow any speakers. Wh, maybe the secret is to have a long re-routed USB cable at the ready for quick removal. Kind of like how fishermen set the hook when catching fish. Only kidding of course.