HELP blown speakers

Here is my situation. I am using a set of Silverline Sonata Mk 1 speakers with Pass Aleph 2 amps and a Pass Aleph P preamp for the monitoring in my home recording studio. The Aleph P is fed from a MOTU 2408 Mk 3 Audio interface/sound card that is in turn fed by a Power Mac G4. I run all my signals through a single set of stereo outs on the 2408. All the mixing is done via software in the computer (no mixing board)The effects (reverb etc...) that the computer can generate can go from smooth and lush to downright hashy digital grungefest.

I was working on a piece of music yesterday and using an effect that can tend to produce some peaky loud transients at certain moments. Never would I say the volume exceeded 95 db. I do know that more than once I'm sure the output of the 2408 was overdriven. I was noticing that the sound had a definite distorted quality about it, but I thought that it might have something to do with the effect I was using. I started to try and figure out what was causing this and I ended my search at the speakers. It seems that both speakers have suffered damage. I confirmed this by hooking up another set of speakers, which sound fine. The sonata speakers have 2 tweeters each, and it seems that on both speakers they are damaged. The produce sound, but it is definately distorted, I don't think that the midrange and woofer is damaged, though I'm not entirely certain.

What would cause this? I didn't have the preamp volume to high, and I never played the speakers that loud. I'm pretty sure that I have caused major damage to the speakers. I'm sure I can replace the drivers, but I want to make sure this doesn't happen again. Is it possible to overdrive an input on a preamp to the point where it can damage the drivers even though the preamp is not feeding the amps that high of a signal?
Clipping begins to shape the waveform like a square wave, generating large products at ultrasonic frequencies. If it is the preamp clipping this high frequency garbage can send some amps into oscillation exacerbating the problem. Other amps can amplify it just fine, passing it right along to the speakers at great levels. Moving coil tweeters do not like substantial amounts of energy at ultrasonic frequencies, the rise time is too fast and the power too great. If you wish to persist in this I would suggest a simple ultrasonic filter at the input of the power amp with a knee just over 20K and a steep rolloff.

Don't use your nice stereo gear for your computer
use. The soundcard(even a good one) in a pc is vastly inferior to your stereo gear. You are using thousands
of $$$$ of equipment on a $100-200(if that) soundcard.

I agree with the above post. Clipping doesn't have to
be loud to cause damage. You can do all kinds of waveform
manipulation on pc's, and output waveforms that are very
unhealthy to your speakers, even at low volumes. I am
a loudspeaker builder(not a pro!), and a pc tech, with
a degree in electronics, so for whatever it's worth, that's
my opinion.

I have a receiver, Infinity IL120 sub(500 watts), and 2
Infinity towers which makes for a dandy PC sound system.
Not top class audiophile gear but very nice for pc. You can
get all these for $500-700 used, here at Audiogon.


Mark, your note is correct.

The beast is the computer interface sound card that can draw the significant amount of DC to the direct-coupled amp and certainly to the speakers.

Ej.., If you want to experiment with other set of speakers I bet you'll get the same "success".

Jitter, improperly smoothed and rectified digital samples, re-locking DAC errors(that i'm sure a-plenty in the computer interface card) do produce huge DC splashes onto the further analogue proccessing and amplification that theoretically can damage the speakers.

In most audio dedicated digital components they occur for the very short period of time and someway may be neglected. In the sound card you're now using, this problem might be on the significant level that you cannot neglect and should realy avoid.

In my case, to make a few points, I have an output transformer that won't likely pass any DC and certainly will save my speakers from such splashes but becides facing the problems with the rest of electronics ESPECIALLY power amp where everything in this case depends on quality of passive elements and the power supply as well is in effect.

The design philosophies of different manufaturers state its own truths and the right truth is the one you carefully select.
There are preamps that filter out DC and there are preamps that don't by its manufacturer's own stated reasons and they're all right. It means that if you have the direct-coupled preamp you MUST have the digital components near-perfect in terms of re-locking, jitter and certainly the high quality analogue output.

I don't want to drive you away from having Pass P preamp since this unit is truly music-dedicated piece AND has no dedication to the mixing boards or PC sound cards. There are pro-grade dedicated speakers along with preamps that connect to mixing boards and/or PC sound cards without hurting each-other.
DC would seem to not be a problem for tweeters because of the high pass crossover associated with them. The blocking capacitor in this crossover would not pass DC. It would be possible that the crossover was rendered inoperable by a DC pulse, thus passing it on to the tweeter, but again, DC in a tweeter would normally fuse the voice coil, causing total failure, not distorted sound.
I believe into my poor knowlege of the speaker's internal structure but in reality, the rectangular and saw-shaped splashes due to improper re-locking and/or jitter in digital-to-analogue domain have the huge spectrum from DC components way upto 10x harmonics. X-over caps might not even have enough time to smooth-out such splashes and pass this all junk onto the coil and it probably doesn't neccesarily have to be DC.

Right now I believe that not only tweeters suffer but woofers as well

There also could be solution to use cheap tube buffer such as Musical Fidelity X10 since tubes do not operate on such high freequencies as transistors.
Thanks guys for all the valuable information. I have been using the Silverlines and Pass amps for some time with no problems. Recently I did add the Pass preamp and got rid of my mixing board. The sound is so much better than the Mackie mixer I was using, I would hate to give up the what the Aleph P adds to the sound of my system.
I side with Mech and others on this. A relatively inexpensive audio card can produce unhealthy ultrasonic garbage, whether during clipping or simply as a byproduct of a poor digital filter. It can further cause oscillation or instability in wide-bandwidth amplification stages, like many ss electronics. Whether additional equipment instability occurs or not, this ultrasonic junk will slowly cook your tweeter and/or high-pass x-over components.

Also, don't assume that because a circuit uses tubes that it is limited enough in bandwidth to help in your situation. A good tube pre-amp or buffer stage can easily be flat to 30kHz, and some much higher. If the problem is with sample rate spiking or distortion products within its passband, they will easily be passed downstream to the next component. Mech's suggested solution is sounding better and better. Good luck.

I talked to Silverline and it appears that the midrange driver is what failed. Apparently these are rather fragile compared to the other drivers in the speaker. I'm gonna replace it and keep my fingers crossed. Thanks for everybody's help.

I'm going to correct a few of you on your statements on sound cards. While you are all correct that the average sound card will produce some harsh and undesireable digital artifacts, the cards that are used in recording applications, such as the higher end Turtle Beach and Delta cards are designed to filter this out. It would not do to record without filtering that sort of hash out.

Ejlif, I'd encourage you to look at the information that came with your card and see what it's input and output stages are spec'd out for. Find out if they are using op-amps or if they have more sophistocated preamps and processors in them.

Best regards,