Cable "burning": Real or VooDoo ???

frap - I enjoyed hearing your take on the Ampzilla. I'm not a EE either but I have had the luxury of working with several. Here are a few things I've learned that relate to power supplies, voltage sources and damping factors. You may already be familiar with this.

It is a high damping factor that makes an amp a true voltage source. Damping factor is the measurement of change in output voltage as load impedance varies. In the case of the Ampzilla as the speaker impedance climbed from say 8 ohms to possibly 30 ohms at resonance (closed box), the output voltage of the amp increased. You got to have more bass without feeling guilty about using the tone controls :-) Naturally there is a price to be paid, two actually. One, the speaker impedance doesn't always change in a favorable way. This is probably why the electrostats sounded bad. The amp is changing the frequency response based upon speaker impedance. No one will argue that even tiny changes in frequency response are not easily detected. Second, when a dynamic driver overshoots or rings (as they all do) they generate a correction signal that acts like a shock absorber to dampen the overshoot. If an amp has a low damping factor much of this signal is lost. This may be the whoopyness you refer to. If you want to fool around and see what your amp would sound like with a lower damping factor, hook you speakers up with 24 or 27 gauge wire. Even that may not get the DF as low as the Ampzilla.

As I understand it, in SS amps stiff power supplies are a benefit to the designer but not to the listener. Suppose you want to get all the power you can from your output transistors. In that case you design a rock solid (expensive) power supply that parks the power supply voltage near the maximum voltage the output transistors can handle. As transistors got better it became possible to have a higher voltage spongy (less expensive) power supply. It would sag and deliver the rated power of the amp and under music conditions operate at higer voltages and deliver greater power to the speakers than a stiff supply of the same RMS rating. I'm not sure but it may be that this technique doesn't work so well with tubes. I think the tubes are biased to operate at a particular voltage and to have the power supply bouncing around one or two hundred volts could screw up the amps performance.

Stiffly regulated supplies in SS power amps typically DO NOT do well with low impedance loads, large changes in dynamics, severe impedance swings, peak power output, etc... UNLESS the amp is built like TWO tanks with a MASSIVE power supply reserve. The supply and regulators would have to be MUCH larger than an unregulated supply to achieve similar results under dynamic conditions. As such, it would cost about as much as a tank to build something like that.

Given that most people do not have the budget for such things, most "common" SS amps will perform best in an unregulated design with the "killer regulation" or "stiff" supply reserved for "price no object" and "esoteric" designs. As to which design sounds better ( tight regulation, moderate regulation or no regulation ), there are proponents in every camp that are quite vocal.

Tubes, on the other hand, typically don't pull much current and need higher and stable voltages to work best. As such, building regulators for a reasonable amount of current for tube use is much simpler and cost effective than building the monster circuit that would be required for a beefy SS design.

As to damping factor, that is determined by the output impedance of the amp vs the overall load impedance of the speakers. Given that fact, damping factor changes AS THE LOAD CHANGES. If you look at specs for amps, they will state the damping factor at a very SPECIFIC impedance for this very reason. As such, damping factor will ALWAYS be higher as impedance of the speaker is increased.

This is the very reason that tubes, which typically have a higher output impedance, don't do well with lower impedance loads. The damping factor is pretty low and the speaker can now "modulate" the load that the amp sees with greater ease. As such, you get less bass control with a reduction in definition.

Steve's suggestion of using tiny wire WILL alter the damping factor of the system, but not with the results that he speculates at. Putting more resistance at the output of the amplifier in effect RAISES the damping factor of the amplifier. You haven't raised the output impedance of the amp itself, you've raised the load impedance that the amp sees. Since bass requires higher current levels to reproduce than high frequencies do, the smaller gauge wire and higher series resistance will produce a much thinner, leaner and "taught" bass simply because current flow has been "choked". This is NOT due to a "reduced damping factor" as explained above, but due to lack of overall bass output.

Now compare this to an amp with a low damping factor and your results would be the opposite. Low damping factors are typically associated with an increase in round, mushy bass that rings. In effect, the tonal balance has now become warm and full, not thin and lean as would occur with the "thin wire" substitution recommended above. Sean
>

Sean - I think you are right that the output voltage of the amp will be more regulated with the small wire. However, I believe the voltage at the speaker terminals will not. Instead of small wire just imagine that an 8 ohm speaker has an 8 ohm resistor in series with it. Now half the amplifiers output voltage will be dropped across the speaker. AS the output impedance of the speaker falls less voltage is dropped across the speaker. As the speaker impedance rises the voltage across the speaker increases. The speaker is no longer connected to a voltage source. As seen by the speaker the 8 ohm resistor trashes the damping factor of the amp.

Low impedance loads are a problem for all amps that are not designed for them. Even springing power supplies optimized for 8 ohms will deliver enough current to smoke output devices if the impedance falls low enough. The designer either has to use more output devices and drivers or provide an effective protection circuit that limits the current vs. time the output sees.

Frap - I didn't mention the connection between damping factor and feedback. Basically, the more feedback the higher the damping factor. SS amps have higher damping factors and lower distortion because more feedback can be used. It is difficult to wrap much feedback around a tube amp because of the output transformer. Output transformers are a designer's nightmare. One of the transformer's problems is the unavoidably large phase shift at low frequencies. This will turn negative feedback into positive feedback if very much is used.

Steve, mostly all of the mid to late 70s high power SS superamps used tons of negative feedback. Whatever benefits derived from excessive amounts of it,were all negated by the God awful sonic results. The spec sheets were always the same; i.e. incredibly low static distortion figures at very high power...damping factors up in the hundreds!!! and low slew rates (remember when slew rates were the most important specs?). The tank amplifier that Sean was referring to, was, for sure, the Mark Levinson ML-2. There was simply no better made SS amp in that era.

Steve, your example takes things to an extreme. How much "small gauge" wire would you have to run to increase the line loss to the level that you mentioned ??? I took and measured appr 20' of 21 gauge wire. This would be equivalent to a 10' run of cable to and from the speaker. Series resistance was less than 0.3 ohm. While a smaller wire would be higher in resistance and a longer run would also contribute to this, it would take a LOT of wire to produce even just a FEW ohms of resistive loss.

Even with that in mind, i'm NOT discounting your theory that various speaker cables DO alter the load that the amp sees. As such, the amp CAN respond differently to specific impedance combinations that it is presented with due to different speaker / speaker cables / cable lengths required in various installations. As such, you have only helped me to prove that wires CAN sound different and ARE completely system dependent due to the stability of the circuit driving it.

THANKS for the help in clarifying this situation, as i never knew you had it in ya..... : ) Sean
>