AnalogSteph
Major Contributor
Probably not all that great. Spec for residual noise is <= -70 dBu (20-20k), that's 245 µV or -81 dB ref. 2.83 V (or by the metric Amir uses, -85 dB ref. 5 W / 4 ohm). So with 104 dB speakers you can expect about ~23 dB SPL @ 1 m anechoic. That's about 19 dB more than you'd ideally want. Just from the power amp at minimum gain.i expext to use my Yamaha p2500s , while not high end sound it Sounds more than okay, but i havent heard it with a 104db speakers so not aware of how noisy it will be.
Not to mention that a 250 wpc amplifier is vastly oversized in general, but you know that. For 110 dB SPL peak, you'd only need about 4 W.
If you can dial in the levels (or use a line-level attunator) you don't actually need a super duper high-strung DAC, although the ones that comes with a volume control knob will generally be plenty good enough.
So these are the speakers?
You know, instead of (or before) buying another amp you could also make a passive attenuator (L-pad). About 7.5 ohms at ~250 W total and 1 ohms at ~25 W total, so for the former you would no doubt have to combine a number of fat power resistors while providing good cooling.... not exactly pure canned efficiency I'll admit.
I had a look at the schematic for the P2500S, and the limiting factor in terms of noise - aside from relatively high gain - seems to be the inverting opamp circuit used as a level pot buffer and as an inverting summing stage for the limiter circuit. That circuit alone would seem to be costing them about 10 dB or so. (it generates about 4 µV, times 40 = 160 µV. The power amp alone would be at little over 20 µV. So room for improvement = up to 17 dB or so.)
Not exactly an easy mod though, unless you can live without the limiter and with inverted absolute phase (in which case you could remove R431/432 and R507/508 on the "DC" board and run a jumper wire from the input end of R431/432 (facing the PCB connector to input board) to the output end of R507/508 (facing the coupling caps)).
Not sure how much gain reduction one could risk in addition to that. If really needed, I might change R304/308 in the input stage instead. While you're on the input board, it looks like XLR jack wiring is not yet AES48-2005 compliant, though it seems there might be a star ground point to chassis right next to them, which would make this a minor concern at best.