Why no DC blocking caps on speakers?

[xtian]

Our bass player bought a new used rig, a GK 1001RB II and Peavey cab with two 8s and one 18. He was excited to show us. So he arrives at practice, plugs it in, and BZZZZZZZZZZZ-ffftt. 1.5 seconds later, the amplifier and the18" Peavey 1801-4 Black Scorpion is toasty toast. I could smell the toastness. It was so toasty, I cannot pull the coil completely out of the magnet structure.

So. Why no DC blocking caps on speaker cabs?

[tubeswell]

Usually, DC blocking caps are in the output stages of amp circuits (on transistor amps).

They aren’t needed on OT secondaries (because these only see AC). (Which ain’t much help to you now).

Good suggestion tho’.

[xtian]

We're talking about solid state amps here, so no OT.

I'm reading a pair of 10,000uF electrolytic caps, arranged in series as a single non-polar unit, will allow full bandwidth to pass. For my friend's 500w amp, that means 100V caps are needed, and that's $30 in parts. So I'm starting to see why this isn't common.

[xtian]

No DC blocking caps on the 1001RB

Gallien-krueger_1001RB_Service_Manual.pdf

[xtian]

And here, someone warns against using electrolytics as NP caps in high current applications:

https://www.diyaudio.com/forums/parts/1 ... c-cap.html

[pdf64]

With the amp topology used, the amp's output should have tiny Vdc levels. The development of silicon technologies that facilitated the design of high power output stages that could do that was a major breakthrough, as noted blocking caps are expensive, and compromised performance.

Of course if something in the amp fails, it can then put a lot of dc onto its output; ideally there should be a protection circuit to detect that and save the speakers but that adds expense and things to go wrong (nothing more annoying than an overly sensitive protection circuit that keeps cutting the output).

[didit]

Recommend recommending upgrade to class-D.

Best .. Ian

[R.G.]

Largely because the required value of capacitor is very expensive.

Assuming a speaker is actually its nominal impedance value, then the capacitor to pass 82Hz (low E on guitar) with no more than half power lost (-3db) is C = 1/(2*pi*F* 8 ohms) = 241uF. Not too bad (assuming I did the math right, and you're using an 8 ohm speaker).

But it has to be non-polar and high (50-60) volts, and losing half your power to the output capacitor has generally been taken to be a Bad Thing. And that's not good enough for bass, by a factor of two, nor for 4 ohm speakers, by another factor of two, nor for 2 ohms by another factor of two. So four or eight times more capacitiance, 1000uF to 2000uF, non polar, and 50-60V.

Not impossible, but this is getting expensive, folks. We haven't even talked about the hifi tweakos' yelling about capacitor distortion.

So if you can do without an output cap by being clever and sensing DC as opposed to low bass and opening a relay (or, recently, MOSFETs saturated to a few milli-ohms), it's probably cheaper and more reliable than inserting a cap.