Originally Posted by Ivan Beaver
Yes I have run into a couple of systems that had an oscillation in them causing all sorts of problems (burnt crossovers-blown speaker protection-blown HF drivers)-HOWEVER all the oscillations have been BEFORE the power amp-so it was just being amplified. The only time a speaker wire has caused an issue is when the wire had a nik in it and was shorting to ground. But remove the short and all was well.
Do you have any specific examples of where the wire has caused a problem? What was the remedy?
I would love to hear some specifics so I can be more aware of a possible issue.
It's worth mentioning that designers of solid-state power amps almost universally use a network specially designed to prevent oscillations into capacitive loads. It is a parallel R-L circuit in series with the output.
Not many manufacturers provide schematics for their products, but one company that does is Bryston, who provide them here
. I'll just pick one at random, the 4B-SST2 (PDF file)
. On the second page, bottom row, you'll see a block labelled "SSTPWR525-R4 PCB". If you zoom in, youll find L1, a 2uH inductor, in parallel with two parallel 10 Ohm, 3W resistors R11 and R12 for a total of 5 Ohms. Solid-state power amps will tend to oscillate at frequencies from around 800kHz - 2MHz, with 1MHz being very typical. So if the amp without such a network "sees" a significant capacitive load at around 1MHz, that is a potential problem with oscillations. But with the network, the inductor L1 starts to look like an open at high frequencies, putting the 5 Ohm resistance of the R-L circuit in series with the load to ensure stability.
So, are there any power amp designs that omit this R-L network? Yes, and amusingly, these are "audiophile" designs, for which the designer believes this R-L network impacts sound quality. One such manufacturer is Naim, who at one time at least, required the user to use only Naim speaker cable. In this case, they used high-inductance speaker wire in place of putting an output inductor in the amp. The speaker wire with highest inductance is zip cord, so they were selling fancy zip cord to their customers (at "high-end" prices I'm sure, LOL).
The Parasound power amps designed by John Curl also have no output R-L circuit.
Any real-world examples of a user having problems? Yes, here is one
. The user has a Parasound power amp driving the low-frequency part of his speakers in a passive biamp configuration. In this configuration, the part of the crossover providing signal to the midrange and tweeter is out of the picture, and the amp is only driving the low-pass filter portion of the speaker's passive crossover. So the first element the amplifier "sees" is the large series inductor of this low-pass network. It's probably close to an open at 1MHz. In combination with this, the amp has no output R-L circuit, and he's using low-inductance, high-capacitance cable
. So the high-capacitance cable is terminated with an open at 1MHz, and the amp "sees" a moderately high capacitive load at this frequency. I don't remember the details, but I think the amp just shut itself down and did not blow up. So at least the amp had good protection circuits. Another amusing corollary to this is that many "high end" amplifier designers omit protection circuits for fear of degrading sound quality.
Another example is here
. Same scenario: amp driving high-capacitance cable into woofer section of loudspeaker in passive biamp configuration. I don't know whether or not the particular amp used had an R-L output circuit though.
Please note I'm not advocating for "wire sonics" or other such audiophile nonsense. I just wanted to point out that the potential oscillation mentioned by Ugly1 can happen in the real world. But the funny thing is, the only real cases of this I've seen stem from a combination of "audiophile" amps and high-capacitance "audiophile" speaker wire. The usual "properly designed" disclaimer applies here.