power/impedence reciprocity

What would you rather have for a medium sized room: an amp that does 125/250/500 watts into 8/4/2 ohms, or an amp that does 250/325/500 watts into 8/4/2 ohms?

It has been explained to me that the 125/250/500 will give more slam, low end detail and control than the 250/325/500. This is because all speakers will go below 1 ohm for short durations during drum beats and other low frequency percussive sounds, and the 250/325/500 falls off more quickly with the load that by the time you get below 1 ohm (or even below 2 ohm) it can't handle the low end nearly as well.

I have always suspected this to be the case, but would like to check my suspicions with some techies. Thanks for the advice.
MOST amps that "double down" are simply under-rated at higher impedances for sake of advertising purposes. Most will do WAY more at 8 ohms before clipping than what their rated power specifies. This gives them what appears to be "high current" status and a high "dynamic headroom" rating also. Both tend to add to their status as being signs that the amp is built quite sturdy and able to drive tough loads.

As such, you can see why they might take a lower overall power rating simply for the "doubling down", "slam" and "tremendous peak power capability" that it would offer. If you look at the actual measured power at clipping with various impedances, you would see that the figures were "fudged" somewhere along the line though. The figure for 8 ohm clipping would not be too terribly far off from the 4 ohm point of clipping in many cases. At least not the expected 50% ( or doubling as impedance is halved ) as one might expect.

The exceptions to this would be the amps that come as close as possible to true "voltage sources". These amps basically DO double down ( or come extremely close ) at the point of clipping as the impedance is halved. I know that the Sunfire's ( verified in independent testing and reviews ) do this and i would think that the Innersound might ( not sure ) according to their claims. Some of the other digital based power supply amps might be in the ballpark also.

The bottom line is that you want an amp that has the highest rail voltages and the highest amount of sustained current capacity that can respond quickly. If that sounds like a mouthful and a lot to ask, it is. Even if you were to get all of those things in one amp, that is not to say that you would like the tonal balance or other aspects of the amp. There is a LOT more to building a balanced package than just being able to drive the speaker to high volumes and try to keep it under control. Sean
OK, makes sense. Now where do switch mode power supply and class A amps fit into this? Or is that a totally different matter? Thanks for your advice.
That is, as you say, "a totally different matter." There is short discussion of what Class A is at www.audioasylum.com/forums/set/bbs.html Just look for posts within the last couple of days.

Two separate threads titled "Why class A" and "Is anyone interested in what class A means" explain it fairly well.

Also a recent thread at www.bottlehead.com within the last couple days.

Seems to be the issue of the week.

Personally, being a tuber, I would never buy a speaker that did not present a fairly consistent and benign impedance curve and was at least fairly effecient. Frankly, I think I would do the samething w/ SS where clipping is so much more of a wake up call.

I remain,
I agree with Sean at least in regards to lower cost equipment. I know one company that sells a 50 watt and a 100 watt integrated amp at 8 ohms. The 50 watt amp is rated at 105 watts at 4 ohms and the 100 watt amp at 135 watts at 4 ohms. The 50 watt amp is pretty typical; the 100 watt amp has a lot more features. Obviously, these two amps are really closer as far as power output than advertised, but they want you to think you are getting a lot more than features when you pay 50% more for the 100 watt model.
Sugar, cost is not a factor here. Look at what the amps clip at and then look at what they are rated for at the various impedances involved. The closer that they come to "doubling down" at the point of clipping, the truer the statement "high current" and "true voltage source" would apply.

The fact that the 100 wpc integrated that you mention does not go up drastically in power as impedance is halved tells me that this manufacturer is actually giving you "more legit" power ratings on this model. He could have rated the 100 wpc model at 70 wpc at 8 and then 140 wpc at 4 and marketed it as "high current", "high dynamic headroom", etc.. As i mentioned before, he would have simply de-rated the high impedance power to make the low impedance output look even more impressive. It would still be the same amp, just with different specs. This follows suite regardless of price.

As to the 50 wpc MORE than doubling as impedance is halved, i would surely tend to believe that the 8 ohm output level has been "deflated". I would venture to say that it is probably quite a bit more powerful than 50 wpc at 8 ohms. Only testing and measurements of the actual component would verify this though.

Some amps just come closer to REALLY doubling down than others. The power output at clipping tells you the amps REAL capacity and just how "beefy" the power supply and output section really are. Sean
That why I mentioned cost. Generally speaking many better made (higher cost) amps come closer to doubling.
This topic prompts me to finally ask a question that's been bubbling under my hat for years. Why is it, factors of preference aside, that both a SS amp that comes close to "doubling down" and a tube amp that doesn't can basically get the same job done in most cases? Average speakers, even if they don't drop to very low impedance levels, still present a load that flucuates markedly with frequency. Why don't tube amps as a catagory seem to suffer some pronounced sonic deficiency compared to "voltage source" SS types, as long as the minimum doesn't drop too far below the output tap rating? Could the ability to increase output as impedance drops not really be that important? Based on the "doubling" propaganda you read/hear, someone who had never heard a tube amp before would have to be excused for thinking that the breed must deviate noticeably from amplitudinal fidelity to the imput based on load impedance vs. frequency variations. On the Atma-Sphere website, Ralph Karsten argues in a white paper that what's actually desirable is for an amp to be able to maintain a *constant* output (watts) as frequency/impedance changes. His rationale is based on the inverse of the conventional wisdom - he criticizes SS amps for *dropping* in output power as impedance *increases*. (This, of course, is something his OTL tube amps supposedly don't exhibit.) But my own ears tell me that, within reason (and putting aside separate issues such as power vs. sensitivity, or the effects of output impedance on damping and frequency response), both tube and SS amps can drive most speakers possessed of average impedance curves and yield essentially comparable sound, despite the rhetoric. What gives?