tube watts vs transistor watts?

An excellent (although unnecessarily lengthy) thread on tube watts vs. solid state watts, from last March, can be found here. Several of our most knowledgeable members contributed some good insights, some of which (mainly on a theoretical level) have not yet been stated in this thread.

09-23-10: Xti16
I just traded my Bryston B100 rated at 100w 8ohm and 180w 4ohm for an Octave V70se rated at 70w 4ohms (can't find an 8ohm rating). Anyhow the Octave just crushes the Bryston in power and control. Not bashing Bryston for I had mine for 4 years and thought it couldn't be beat without spending a truck load of money. The demo in my home took about a whole 5 seconds to want the Octave (and that was with cold tubes). I think the caveat here is my speakers. They are Dynaudio C1's which are known to like current not necessarily watts. Rated at 85db 4ohms. Logically it makes no sense unless the power is calculated as E X I instead of I sq X R and Octave uses a lower voltage (E).

I hope Almarg chimes in on this one for he always has good technical answer.

So I'm guessing it has more to do with the speakers where tube watts may seem like more compared to solid state. I don't think that applies to all speakers, but some.

Thanks for the kind words, George.

In referring to "power and control" I assume you are referring particularly to the bass region. I can't offer a definitive explanation for the differences you found between the two amps, but my guess is that the differences are unrelated to tube watts vs. solid state watts, and are unrelated to the maximum power or voltage or current capabilities of the two amps.

What I suspect is that the major reason for your findings is the interaction of the differing output impedances of the two amps with the impedance vs. frequency characteristics of the speakers. As you'll see in Stereophile's measurements, the C1 has some wild up and down impedance swings in the bass region (which are not uncommon, btw), with a sharp rise to 18 ohms at about 60Hz.

I assume that your Bryston amp, like most solid state amps, had a negligibly small output impedance. I couldn't find specs on the Octave amp's output impedance (or damping factor, which equals output impedance divided into 8 ohms). However given that it is a tube amp, and one which, unusually, has only a single output tap on its output transformer, it is probably safe to assume that its output impedance is significant, perhaps 2 or 3 ohms or so.

Since an amp with near zero output impedance acts essentially as a voltage source (within the limits of its current capability), a load impedance rise will not affect the voltage across the load, but will result in decreased current draw, and therefore decreased power into the speaker at the frequency corresponding to that impedance rise.

In the case of an amp having a highish output impedance, the current decrease caused by a load impedance peak will be offset, at least partially, by an increase in the voltage across the load, since (oversimplifying slightly) the voltage the amp is trying to output will divide up between the load impedance and the amp's output impedance in proportion to the two impedances.

So I suspect that the low frequency impedance characteristics of the speaker, particularly the sharp peak at 60Hz, result in the speaker being synergistic with the output impedance of the Octave amp. And of course intrinsic differences in the sonic characters of the two amps could obviously be factors as well.

Best regards,
-- Al

Magfan, as you've frequently and correctly pointed out, phase shifts (reflecting the degree to which a speaker's impedance is capacitive or inductive at various frequencies) are a key factor in how difficult a load a speaker presents to the amplifier. Negative phase angles in particular (denoting that the load is capacitive), will increase the amount of current the amp will have to supply (i = C(dv/dt) for a capacitor), while simultaneously reducing efficiency (since capacitors (and inductors) cannot dissipate (consume) any power, apart from whatever amount of resistance may be present).

In the case of George's C1's, though, I suspect it is not too major a factor. As you can see in the Stereophile plots, the phase angle peaks and dips generally occur at points where the impedance magnitude is not particularly low, and the low points on the impedance magnitude curve generally coincide with benign phase angles.

Audioquest4life, thanks for supplying the literature on the Octave amps. I cannot understand the review in "Stereo" magazine, because it is in German, but I did read through the manuals, which I found to be well-written and confidence inspiring. They did not provide information on output impedance or damping factor, though, which is key to my previous post. One clue, though, is the statement on page 35 of the V70SE manual that "the oft-quoted damping factor is not normally a guarantee that an amplifier will exert tight control over the loudspeakers." That would seem to imply that the amp's damping factor is not particularly high, and therefore that its output impedance is not particularly low, which adds a bit of credibility to my previous post.

As far as the "black boxes" are concerned, as you realize their function is to increase the energy storage capacity of the power supply, and also to provide some additional noise filtering. That would seem to be something that can only help and can't hurt (aside from the additional cost), but without either running some careful with/without listening test comparisons, or having an intimate knowledge of the design, I don't think it's possible to say how much of a difference they would make.

Best regards,
-- Al