Re: Suspect amp power ratings
And unfortunately, when it comes to musical low frequency power, the watts and voltage specifications don't tell the customer (who is always right) what they need to know.
As i think I already mentioned, there is no accepted quantification for duty cycle. Since we entered the age of 2 ohm operation (like 4 ohm bridged) there are no continuous operation amps sold to consumers that I know of.
Using music and pink noise I tested all the amps I had in house at the time, a mono SpeakerPower SP1-4000 plate amp, a Crest CC2800, a Crest CA-9, an old Crown PSA 2 (weighs 8 times more than the SP-4000 !) and a QSC PLX-3602.
All the rack amps are capable of 4 ohm bridged mono operation, so a single four ohm load was used.
Bridged mono 4 ohm the CC2800 is rated for 2800 watts, the CA-9 at 1800 watts, the PSA 2 at 1210 watts, the 3602 at 3600 watts, SP-4000 at 2400 watts.
With music and pink noise (30-100 Hz) into a B&C18SW115-4 loaded bass reflex sub run to just under clip or limit lights illuminating, the SP1-4000 equaled the SPL output of the CA-9 and the PLX 3602 (though one 3602 died during testing) put out about 4 dB more than the PSA 2, and 5 dB more than the CC-2800.
The 44 pound CA-9 could be driven another 3 dB or more louder than the other amps and still "sound good", while the limiters in the other amps kept level from increasing.
The CC-2800 limiter made the speaker sound like it was flapping as soon as it hit, horrible.
As i think I also mentioned (recently if not in this thread), all clip limiters are not created equal. IIRC and I am not an expert on Crest, but I knew them as a competitor (before PV bought them). Their version of clip limiting used a somewhat slower attack time, that actually delivered a louder signal, with the mostly inaudible leading edge or complete very brief transients passing unmolested and allowed to clip. In my experience consumers always liked louder better, so the Crest variant was successful in the marketplace while not preventing as much clipping as other more conservative clip limiters.
Worst clip limiter I ever heard was one Yamaha amp model (no I don't recall which one) we were bench testing, but I wouldn't expect all of them to sound that bad (they couldn't).
Although all the amps bench tested according to their specs, they didn't "sound" like their specs would indicate, and the difference was rather huge.
Art
When operating at the limits of amplifier current capability (2 ohm) this is not very surprising. Add the variables of loudspeakers that are not pure resistive loads, and some amps with VI (secondary breakdown) limiting protection. This secondary breakdown protection is a variant on current limiting that involves both current and voltage across the power device, so different than simple current limiting. Audible perturbations could occur related to protection circuits tripping when driven hard with marginal loads. Not to mention different duty cycle capability. This may not show up with resistive loads.
This is yet another reason I don't advocate routine 2 ohm operation... Spec your system to work well within the capability of the amps (4-8 ohm) and use the extra drive capability to use in emergencies if an amp failure occurs so you can double up a load to finish the gig. If all amps are loaded to the max you have no option to limp along. (Back in the day I've heard anecdotes of finishing shows loaded to 1 ohm but it wasn't pretty.)
Customers are attracted to driving amps at 2 ohms since it seems like free power. Nothing in life is free... That said I expect the modern generation class D amps to be somewhat less affected by load impedance (ignoring frequency response interaction with output filters).
It is difficult to make listening evaluations of amps driving loudspeakers. One test we tried at Peavey during development of one amp series, was to load up an amp with some PA speakers, in parallel with a second pair of low efficiency studio monitors in a different isolated sound space, so we could drive the amps hard but listen over the studio monitors without hurting ourselves (I still wore plugs).
We kind of explored this duty cycle issue (or at least I raised it) back when Bink did his amp shoot-out years ago.
JR
