Here's a plot of the impedance curve of the Planetarium Alpha main module:
For my dastardly comparison purposes, below is the impedance curve of a generally well-regarded speaker (I'll admit that I cherry-picked one whose curve will illustrate the point I'm going to make):
The scales are different, and the scale on my curve is harder to read. The horizontal line that most of my curve is hovering above is 8 ohms, the line above that is 12 ohms, and that's a 60-ohm peak at about 65 Hz.
The first thing I'd like to focus on is the spread between lowest and highest impedance above 100 Hz. In this region, the well-regarded speaker (WRS) has roughly a 10-to-1 impedance spread (5 ohms to 49 ohms), while the Planatarium Alpha module has a roughly 1.5-to-1 impedance spread (8 ohms to 12 ohms).
The problem arises when the WRS is connected to a different amplifier type than what it was designed for. In this particular case, the WRS was designed with a solid state amplifier in mind. A solid state amplifier typically acts as a voltage source - that is, it puts out constant voltage regardless of the load. To illustrate, such an amplifier might be putting out 1 watt into an 8-ohm load, but 2 watts into a 4-ohm impedance dip and 1/2 watt into a 16-ohm impedance peak.
So, when connected to the WRS and at an output level that would give us 1 watt into 8 ohms (for the sake of convenience), a solid state amp will put out about 1.6 watts into the 5-ohm dips at 100 Hz and 6 kHz, but only .16 watts into that 49-ohm peak at 2.5 kHz. But since the speaker was "voiced" on this type of amplifier, the frequency response curve will be flat (or whatever the designer intended).
Now, let's look at what happens when the WRS is hooked up to a tube amp that approximates a power source (many tube amps have this characteristic) - in other words, it delivers pretty much the same wattage regardless of the load. So, theoretically we get 1 watt into the 5-ohm dips and 1 watt into the 49-ohm peak. Unfortunately, the speaker was voiced to have correct tonal balance with 60% more wattage into those dips and only 1/10th as much wattage into that impedance peak. Well... not quite; in the real world, we might get 4 times as much wattage it was voiced for into that impedance peak, rather than ten times as much. So... with a tube amp we might get a (rather insignificant) roughly 1 dB dip at 100 Hz and 6 kHz, combined with a 6 dB(!) peak at 2.5 kHz, which will be highly audible. Obviously, the tonal balance of the WRS will be way off with a tube amp that acts as a constant-power source!
What types of tube amps approximate a constant-power source? Those with low damping factors, which typically are those with very low or zero global negative feedback (which is a desirable feature itself, but that's another topic for another day).
Okay, let's back up a bit. What if the WRS was designed to have correct tonal balance when connected to a constant-power-approximating specialty tube amp, but we hook it up to a typical solid state amp? With the solid state amp, the WRS will be getting 60% more wattage than it was voiced for at 100 Hz and 6 kHz, but only about 1/10 of the power it was voiced for at 2.5 kHz. The mild bumps at 100 Hz and 6 kHz would probably be inconsequential, but the 6 dB dip at 2.5 kHz will be audible as a loss of presence and upper harmonic richness (though it won't be as objectionable as that 6 dB peak would have been a couple of paragraphs up).
Alright, so a speaker with an impedance curve like the WRS should be used with the type of amplification it was designed for. What makes the Planetarium Alpha (and my other speakers) any different?
Above 100 Hz, the impecdance curve of the Planetarium Alpha will give us only about 1 dB variation as we go from solid state amp to specialty tube amp. That means the tonal balance won't change significantly as you go from one amplifier type to another.
Now wait a minute, you're probably saying - what about that 60 ohm peak in the bass region? That's gotta be just as bad if not worse when you go back and forth between amplifier types!
Well not really, because of where it occurs. You see, at the frequency where that peak occurs, the bass response is already down by about 5 dB, assuming a constant-voltage (solid state) amplifier. Connecting the speaker to a low damping factor tube amp results in perhaps 3 or 4 dB more output centered on 65 Hz, so the net result is greater bass extension but no significant peak. In practice, what you'd do is simply set the low-pass filter on the bass amp to a lower frequency.
So with a solid state amp, the Planetarium Alpha module goes down to about 70 Hz. With a low damping factor tube amp, it goes down to about 55 Hz.
Duke
