DAC as preamp w/subwoofer

Is this feasible with the W4S or with any other DAC for that matter?

It depends on several factors. Let us know the make and model of both the sub and the power amp, and how long the cable lengths will be between the DAC2 and each of those components.

Regards,
-- Al

I see that the DAC2 is specified as having a nominal output impedance of 100 ohms; the RB991 is specified as having an input impedance of 32K, with Stereophile's measurements being 36K unbalanced and 45.2K balanced; and I couldn't find an input impedance spec on the Totem sub.

Given that, I don't think you'll have any problem, unless the Totem's input impedance is extremely low, which I doubt.

I would guess, however, that using y-adapters to split the rca outputs of the DAC2, and connecting via rca cables to both destinations, will give better results than connecting via xlr to the power amp and rca to the sub. I'm basing that on the suspicion that the DAC2's rca output is driven by the same output stage that drives the positive polarity signal of the xlr signal pair. If so, connecting via rca to the sub and xlr to the power amp would, to some degree, unbalance the impedance levels of the balanced signal pair. You might want to ask W4S if my assumption is correct, or if the rca and xlr outputs are driven by separate output stages.

Also, it would be preferable to use cables to the sub that have reasonably low capacitance (e.g., 35 pf/ft or less), as capacitance that is excessively high on the cables to the sub, in relation to the DAC2's output impedance at high frequencies, might result in a slight rolloff of the extreme upper treble on the signals going to the main power amp.

I note that the sub has speaker-level inputs as well as rca inputs, so driving them from the power amp's outputs is another possibility, although for some reason Totem indicates that using the rca inputs is preferable.

Regards,
-- Al

Good posts by Eric and Cymbop, with which I agree 100%. Given that the DAC2's rca output is common with the positive polarity output signal on xlr pin 2, using an xlr-to-rca adapter for one of the two rca runs, and connecting the other run to the rca output, would indeed be electrically equivalent to using a y-adapter on the rca output.

In that situation, I second the caution about making sure the adapter leaves xlr pin 3 unconnected, as the Cardas adapters apparently do. Most xlr-to-rca adapters connect pin 3 to pin 1 (ground), which would short the DAC2's output signal on pin 3 to ground. A lot of equipment can tolerate that, but some cannot. Why pin 3 is commonly shorted to ground on xlr-female to rca adapters, which would be used on xlr outputs, has always been beyond me. (It is desirable, though, and often may be a necessity, on xlr-male to rca adapters, that would be used on xlr inputs).

Regards,
-- Al