A TVC will work its best when you think the music is plenty loud and the TVC attenuation is no more than 14 dB down. So, from zero attenuation (volume up all the way), dial back to minus 14 dB and tell us if it is plenty loud. If it is not, then you need an active preamp with that amp/speaker and source combination.
The reason why I say 14dB maximum attenuation is for a good reason. Most TVCs are in the "impedance friendly" zone and are transferring the energy their best when they are attenuating the most (volume way down). As you go above (louder than) the 14dB attenuation point, things start to fall apart and you get the symptoms of "no drive" and "wimpy dynamics", just as you describe.
I don't see why 14db is the magic number...
According to the tvc faq at Nick's site, the TVC is always reducing the output impedance of the input source (e.g. a DAC). So although you get a lowering of the impedance with greater attenuation, does that necessarily mean that the sound quality is *bad* when the attenuation is up to -10db or more?
I'm right on the fringe of that range--sometimes I attenuate at -18db, sometimes up to -12db--and, well, maybe I'm just not golden-eared enough to distinguish between these settings, but more to the point, it's literally impossible to make an a/b comparison since, well, changing the recording changes what you are comparing!
I've heard things that sound fine even if brought up to -10db, and I've listened to things that sounded mediocre even though they were pressed for a -18db setting. The quality of the recording seems to matter more than the setting of the TVC.
I wish Nick would clarify some of these issues, though. There are two that could probably be answered pretty quickly.
1. Is the attenuation in 3db steps for the first 12 steps, and 2db steps for the next 12 steps? Silly me, I had assumed that it was 2.25db per step.
2. When the tvc faq uses the example of 10% volume, 50% volume, etc., how does this correspond with the steps on the volume knob?! If I had to bet money on it, I would calculate the following "percentage" for, say, the famous -14db setting:
Voltage ratio = 10^(db/20)
= 10^(-.07)
= .1995 ~ 20%
That means that your boundary of -14db corresponds to a step-down of output impedance by a factor of 1/(.1995)^2 ~ 25, which seems pretty impressive. If I make it go an extra click to -12db (as I often do), this becomes
Voltage ratio = 10^(-0.6)
= .2511 ~ 25%
and the step-down of output impedance is now at 1/(.251)^2 ~ 15.8, which is not as impressive as 25 but still seems pretty darned good!
I suppose things may get downright magical if the attenuation is set at -30db and then we have
Voltage ratio = 10^(-1.5) = 0.0316 ~ 3.1%
with an impedance step-down of (1/0.0311)^2 ~ 1000. But just because it isn't magical, that doesn't mean that it sounds bad!