cause of 881s death: unknown

[kneadle]

Subject: Tangband W3-881S. Cause of death: unknown. Symptoms:

First, I did the Mike McKenzie Mod , seen here . For what it's worth, it's a beautiful mod.

This morning, I was testing different speakers with my BrianGT chipamp, and when I hooked my hohlflutes back up, one of the drivers suddenly went into full extension; I don't know why. I'm certain I didn't accidentally reverse the polarity. That was the first thing I checked. Upon examining it, it seems like the coil has changed size or shape, or the magnet is working incorrectly. I say that because there's resistance when I try to get the cone to move freely.

Then there was the dreaded smell of something fried.

I checked the amp. No strange or foreigh voltages. Other speakers work fine.

Condolences may be left at my e-mail address or as answers to this thread.

Can I save it? Or should I order a replacement? Any ideas concerning cause?

Thanks,
Inconsolable,
Dave

[markp]

You had DC go through you speaker! Either the power supply had only one polarity working or the input signal was DC. It does not take long for DC to fry a small voicecoil like in the 881. Check all your power connections from ps caps to the chip to see if one side is flakey. Do this with a meter not the speakers connected.

[kneadle]

Oh, I just figured it out...

My wife turned on the preamp side first, then the amp side. That may have invited unwelcome DC currents somehow. Your saying DC brought that back to mind. Thanks.

Poop.

Well, that's why I DIY on the cheap. Easy to replace.

Thanks,

Dave

(edit) PS...I have a few millivolts of DC going through anyway. What would I do to prevent that and maybe protect my speakers from a future mistake?

[BassAwdyO]

You can put a capacitor on them. Capacitors increase in resistance as frequency decreses, when it becomes DC the resistance becomes infinite. Therefore there will be no DC flow in the circuit. I'm no expert, but I've put large capacitors(1000uF+) in line with full range speakers and listened to find no noticable difference in the sound. The larger the capacitance, the lower your DC protection, but basically all you need is a capacitor that will only attenuate out of the speaker's usable range.

[kneadle]

As simple as that? How do I interrupt the circuit with the cap? One cap per pole (+/-) or what?

Thanks,
Dave

[BassAwdyO]

Well you need a nonpolar capacitor, one that doesnt have a + or - terminal, but one that can be configured either way. Either that or two polar capacitors per speaker. Wiring can go as diagramed below. I have shown sketches of polar and nonpolar capacitors and their schematic symbols. You said you have a chip amp? You might want to make sure that chip amps can handle capacitive loads. I've heard that tube amps are inherrently flawed with this problem, but most solid state has no trouble driving such a load. Make sure you use a large capacitor. 1000uF are needed to lower the cutoff on an 8ohm load down to 20hz. For a 4ohm load capacitance needs to double to achieve the same cutoff. I would imagine you'll probably need to buy four polar capacitors and use two for each channel. The only reason I say this is because most larger capacitors (100uF+) are polar. I dont think it matters if you hook up the cap on the + side or - side of your speaker because the signal is AC.

[markp]

You can put a very large bipolar cap in series(in line) with the speaker but you might notice a difference in overall sound quality due to all the signal going through the cap which does have an effect on the signal. You might not like the sound after the cap is in line.

[kneadle]

Okay, thanks. That's easy enough. Tomorrow is DIY day around here anyway.

What I really need to do is understand where the blasted spike came from in the first place.

Since that is my first amp/preamp project, and this is the loudspeaker forum, I'll try to keep this to a yes/no question:

Was it because the preamp, which is wired directly to the amp, turned on before the amp? Did that give me a DC spike to kill that poor thing?

Thanks
Dave

[kneadle]

And can someone direct me to a resource that explains what the DC current did to destroy my speaker?

I'd like to do an autopsy and learn.

Thanks,
Dave

[BassAwdyO]

"one of the drivers suddenly went into full extension"

Let us go back to the basic workings of a speaker. The speaker is driven by a coil and a magnet. When a current runs through the coil it creates a magnetic feild. The magnetic feild of the coil and permanant magnet react with one another to create some force which moves the coil and cone. When a sinusoidal AC wave is fed to the speaker the coil's magnetic polarity changes as the electrical wave's current changes. This will cause attraction and repulsion from the permanant magnet which moves the coil and cone in and out. The inward and outward motion of the cone cause pressure diferences which are transmitted as sound waves. When a DC signal is fed through the coil the magnetic feild of the coil does not change. It moves the farthest inward or outward(depending on the direction of current flow) that it possibly can until the restoring force of the suspension is equal to the driving force of the coil. For a strong DC current this will easily be outside the linear Xmax of the driver where BL is quite low and most of the power through the coil is converted to heat. Because the majority of power is converted to heat and the coil is no longer moving to cool itself down the result will be a fried voice coil. It's hard to say if you're preamp being turned on first actually caused the DC spike, but it is quite possible. You said you had a few millivolts of DC going through anyway, perhaps when you turned on preamp first it accumulated a larger charge and when the amp turned on it discharged