condenser versus direct coupled?

DC absolutely WILL burn up your speakers faster than AC, especially if of the same peak voltage level. That is because DC has a much higher RMS level than the AC waveform generates.

As the AC waveform vacillates from positive to negative and varies intensity*, the DC signal remains consistent in amplitude. Unless the DC has a high level of ripple, the peak power is the same as the average power. With AC, the peak is measureably higher than the average. This means that the heating power of DC voltage is FAR greater than that of AC. Due to this factor, DC tends to melt voice coils and / or the glue that secures the voice coil to the coil former much faster and at lower levels. Sean
>

* With AC, the intensity varies due to the "hill and valley" shape of the sine wave. Not only is the voltage varying as it climbs up and down those "hills", the duration or "thickness" of the peaks and valleys vary too. On the other hand, DC is pretty much a steady state signal that remains consistent. If you took the same peak amplitude of both signals, the DC would look like a straight line running across the peaks of the AC "hills". If you look at the average power that the DC signal maintains compared to the areas that drop below that flat line at the peaks of the hills, AC is FAR less "potent". The reason that we don't use it for our power grid is that it doesn't fare very well when travelling long distances. As such, we generate AC to send the signal over a long distance and then convert to DC for it's greater work-load capacity.

Sean...You really ought to be more specific than the frightening "FAR greater" when describing the damage potential of DC vs AC. Actually, as I think you must know perfectly well, the rms voltage of a sinusoidal waveform (its heating power) is 7/10 that of its peak voltage, which is presumably what you are assuming for DC from a faulty amp. Furthermore, if the amp goes into oscillation, which is the most likely failure mode, the output will be a square wave, (instead of a sine wave) which has exactly the same power as DC. Both levels could cause damage, and the difference in time required for damage to occur would NOT be great. I think that, because there is no loud audible indication of a problem with DC, amplifiers that are DC coupled are more likely to include protective circuitry. (Mine do).

Oh, and by the way, DC transmits over long distances very efficiently. Better than AC. A lot of research work was done on that technology here in Pittsfield Mass by the General Electric Transformer business. (Alas, GE quit the large power Transformer business). The problem with DC is that there is no easy way to step voltage up and down. With either AC or DC long distance transmission calls for voltage of several hundred thousand volts. You can't have that going into your house!

"Actually, as I think you must know perfectly well, the rms voltage of a sinusoidal waveform (its heating power) is 7/10 that of its peak voltage, which is presumably what you are assuming for DC from a faulty amp."

7/10th's is the same as a 30% reduction. That means at least a 30% increase in steady state heat. I say "at least" because we are basing this on a linear sine wave measurement. Using anything other than that can skew the results even further i.e. even greater heat dissipation.

"Furthermore, if the amp goes into oscillation, which is the most likely failure mode, the output will be a square wave, (instead of a sine wave) which has exactly the same power as DC."

This is not "completely" true. A square wave is still AC and will work the driver as it normally would. The only difference is that the average power levels are higher than what one would see with a sign wave. The difference is that due to the driver still working properly i.e. moving fore and aft, you will dissipate energy as both acoustic output and as heat. If the driver has a vented spider or a vented pole piece, the motion of the driver will aid the natural convection of heat away from the voice coil and coil former. This has to do with the pressure changes within the driver itself caused from the fore and aft motion.

When you apply DC to the driver, it locks the driver in one position and holds it there. Depending on the polarity of the DC signal and how it is connected to the driver, it can be thrown out or sucked in. There is no acoustic output to dissipate energy and since the driver is stuck in one position, the natural convection cooling of the driver itself is reduced.

While both will end up causing driver failure, DC will do it faster than clipped AC.

"DC transmits over long distances very efficiently. Better than AC. A lot of research work was done on that technology here in Pittsfield Mass by the General Electric Transformer business."

I wonder if they ever published any information on this based on their test results. I'm sure it would be interesting.

"With either AC or DC long distance transmission calls for voltage of several hundred thousand volts."

Current day AC systems travel for many, many miles using voltage levels that never come close to 100+ KV. While i could be wrong, i think that our local lines operate at appr 24 KV or so.

"You can't have that going into your house!

You can't really have that going through the air either. In times of heavy humidity and the level of current available, the potential for arcing would be too great. Sean
>

Sean...From an electrical power point of view a square wave is the same as dc. The amount of power radiated in the form of sound would be trivial. True that cone motion with AC might tend to ventilate the driver. However, the oscillation would likely be at a frequency too high for the driver to respond, and therefore no significant cone movement would occur. I only claim that the damage potential DOES NOT DIFFER VERY MUCH, and that your "FAR greater" words were a gross exaggeration.

I was not involved with the DC transmission research, but I seem to remember hearing about 450 KV. If you are really interested in power transmission research results I will see what I can dig up and Email it to you. I guess we could check the web also. Some of the GE work was probably proprietary.

Sean...A quick check of the net suggests that 15 to 50 Kvolt transmission is considered "medium" voltage. 200 to 400 Kvolt is "high voltage". And this is for AC. I know that the research on DC transmission was at higher voltages than used for AC.