Pass Amps Class A operation

http://passlabs.com/pdf/articles/leaving_class_a.pdf

07-21-09: Spatine
The power to SPL calculation is pretty much aligned with engineering understanding. Yet reality does not seem to support this kind of calculation. For example if we can all agree that 2W is all needed to produce 78 dB at 3m away, then adding extra cushion by a factor of 5 even (to bring the power requirement to 10W) is not bad an engineering cushion, really, to account for dynamic peak or other Murphy's law situations. Yet in reality, would we buy a 10W/channel amplifier to drive anything? Even a factor of 10, bringing the wattage to 20W/channel is a no-no.

Spatine,
my calc were meant to be a mere guidance to give the author of this thread an idea of the ballpark SPLs that he can expect in his room. it was not supposed to be 100% accurate. You can see that I've taken a swag at the attenuation factor for distance & furniture. So, take it as a guidance.

then adding extra cushion by a factor of 5 even (to bring the power requirement to 10W) is not bad an engineering cushion, really, to account for dynamic peak or other Murphy's law situations.

hmmmmmm...... I'm not sure that I'd agree. A factor of 5 in dBV terms is 14dB (we assume that the impedance the amp is working into remained the same, for argument's sake). Are you saying that allocating 14dB for dynamic peaks is "not bad an engineering cushion"?? If you look @ the spectral content of music you'll find that peaks can be much higher.

Yet in reality, would we buy a 10W/channel amplifier to drive anything? Even a factor of 10, bringing the wattage to 20W/channel is a no-no.

sure! the horn & single-driver speaker guys do it all the time. I've personally heard several very, very dynamic systems in Atlanta that were all Jordan & Lowther driver based speakers that were powered by tube power amps that were in the 5W/ch to 8W/ch range. Nothing higher! In fact, flea-watt power amps is the paradigm that horn, single-driver folks stick to.
The reason this won't work for a Sonus Faber Grand Piano is 'cuz that speaker is so damn inefficient mostly in the bass region. You couldn't get a 14W or 20W amp to provide enough current into a low impedance that the Sonus Faber Grand Piano presents. If you remove the bass driver, I bet that a 14W or 20W amp could drive only the mid & tweeter drivers with high enough SPL to run you out of the room!

lastly, it seems that you might not have read my post properly as I wrote & I've cut & paste:

Of course, this does not include dynamics peaks for which the amp has no class-A capacity (probably by design) & it'll need to go into class-AB & then switch back (just like Elevick wrote).

you might want to re-read my orig post before you take up any issues....

bombaywalla - is your comment about the inefficiency of the Sonus Faber GP also applicable to the more recent Grand Piano Home model? My speakers actually are the GP Homes, different from the GPs, with the addition of a third driver for Bass. Do you have a sense of where they fall on the efficiency scale and what impact that has on performance with an amp like the Pass 150.5? Thanks.

07-23-09: Gipp
bombaywalla - is your comment about the inefficiency of the Sonus Faber GP also applicable to the more recent Grand Piano Home model?

No! I would not extrapolate my comment w.r.t. the SF Grand Piano to your particular speaker. Almarg was kind enough to provide the efficiency number that i keyed off.
This is my opinion:
The SF GP Home specification is 90dB/2.83V/1m into 6 Ohms.
From what Almarg wrote & what I found in Stereophile, the SF Grand Piano is spec'd at 90dB/1W/1m into 8 Ohms.
Needless to say, since one is 6 Ohms & the other 8 Ohms, we cannot compare the efficiency #s directly.
For the SF GP Home, the 2.83V into 6 Ohms yields 1.3348W. So, of course, you end up with a more efficient # since you are inputting more power compared to 2.83V into 8 Ohms. According to my calc, 1.3348W into 6 Ohms is 1.25dB more power than 2.83V into 8 Ohms, so subtract 1.25dB from the SF GP Home efficiency #, which makes it 88.75dB normalized to 8 Ohms. (did I calc this correctly, Almarg??)
Thus, the SF GP Home is marginally more efficient than the original SF GP. IMO, then, if a X150.5 is good for the SF GP then it should be OK for the SF GP Home as well.

Yes, I agree with those calculations, Bombaywalla, aside from what amounts to a typo where you listed the older, non-Home version as 90db instead of 87db.

Here is a link to a review of the Home version, which confirms the 90db/2.83v/1m/6 ohm numbers you quoted for it:

http://www.enjoythemusic.com/magazine/equipment/1001/sonusfaber.htm

But I think that a better way to look at it (which results in essentially the same conclusion as you reached) would be to make the comparison in terms of output current requirements. I say that because the original question relates to the threshold at which Class A is left, which in turn is basically dependent on relations between output current and bias currents.

Analyzed in that way, we have:

GP (old version): 87db is produced at 1 watt into 8 ohms, which corresponds to a current of 0.354 amps.

GP Home: 87db is produced at 3db less than 2.83 volts, which is 2.00 volts, which corresponds to a current of 0.333 amps into its 6 ohms.

0.333 amps vs. 0.354 amps is a difference of 0.53db. So I would expect the "leaving Class A" threshold to occur at a volume level which is 0.53db higher for the Home version than for the old version. Which is an insignificant difference, as you essentially concluded.

Regards,
-- Al