If a device truly is "high current", it will produce twice the power output at or very near the point of clipping as the impedance is cut in half. If the device doesn't do this, it is NOT "high current". While some might say that having a "big power supply" is what makes a design "high current", having the big power supply and having an output stage that is capable of passing all of that current are two different things. This is why all amps should be tested at the point of clipping. This tells you more about the entire amp in terms of "brute strength" than any other spec. Sean
PS... Most amps will fall flat on their face in such a test. That's because most amps aren't really "high current", they are just "high hype".
The amplifier design determines the maximum voltage that it can put out. The load (speaker) impedance, 8 ohms, 4 ohms etc. determines how much current will flow if this voltage is output. That is, if the amplifier can deliver that much current. It would be possible to design an amplifier which did not put out a high voltage, but could maintain that voltage despite a low impedance load which requires a lot of current. The power rating of the amp specifies that it applies at a particular load impedance. If a lower impedance speaker is connected, more power will be delivered.
The "wattage" can decrease and the current can increase. Look at your formula - if the current decreases by a factor of 10 and the voltage increases by a factor of 20, the wattage doubles. As far as the the power rating of the SimAudio i-5, it's determined by the voltage output and the current output into an 8-ohm load. If the output current is fixed to a certain maximum, then that will determine the power rating at a given load regardless of the voltage. If the voltage level is fixed, then that fixes the power rating at a given load regardless of the current available (the extra current available will give more power as the load decreases).
The term "Ultra High" current is a marketing one and there's nothing "notorious" about the amp's output current - it obeys the laws of physics to determine the power rating.
El: what you are basically saying is that it is possible for a low powered amp to be "high current" even though it isn't "high powered" and i agree. It is actually much simpler and less costly to do something like that than it is to make a high powered high current amp. Regardless of the rail voltages used, so long as the amp can supply enough current into the various impedances that it encounters, it will be able to sustain the same voltage levels. This is what makes the amp a TRUE "voltage source". Sean
All good answers so far, IMO. Do understand that solidstate amps, because of their VERY-low output impedance, are 'constant-voltage' devices, varying current delivery as load impedance changes. 'ANY' SS amp can double power as impedance halves, again and again...if one starts at a LOW-enough point. True-high-current amps can do that at least once while starting at rated ('maximum') power. The 20-year-old class-A-biased Lazarus hybrid poweramp is rated at 50Wpc into 8 Ohms, 100Wpc into 4, and 200Wpc into 2 Ohms. THAT is 'high current' indeed. IMO, any amp that won't double 8-Ohm rated power once (into 4 Ohms) is NOT 'high-current'.
BTW, it takes lots of expensive engineering to build a HC amp. The relative sizes of the power transformer(s), rectifier, wire, circuit-board traces, driver and output transistors, etc. all affect the amp's ability to deliver high current flows.
To answer your middle-paragraph question, an amp's power is limited generally by its ability to deliver VOLTAGE into the test impedance. Eventually, ANY amp will clip its output waveform at some voltage, and its rated power will be calculated at some point below that clipping voltage.
On your paragraph-3 questions, I have no idea about the 1st half, but my example of the Lazarus answers the 2nd part.
Jeffreybehr: As discussed in another relatively recent thread, power ratings can be manipulated by the manufacturer. Even if they are rated to "double down", most amps DO NOT really do this, the ratings are just "fudged" to look good to the consumer. That is why i specifically stated "power at clipping" as it is a much more thorough, grueling and revealing test. With this type of testing, the amp either does or doesn't "double down". There is NO guessing involved, regardless of factory ratings at various impedances. Unfortunately, very few reviews actually perform or publish this type of testing, so getting good information becomes harder when trying to make an informed decision. Sean
I'm surprised by your comment:
Unfortunately, very few reviews actually perform or publish this type of testing
for this is certainly something that Atkinson at Stereophile does regularly test and publish.
Metralla: The audio world does not revolve around Stereophile alone in terms of equipment reviews. How many other reviews / reviewers do this? That was the point that i was trying to make. Having said that, i'm glad that Stereophile does perform such tests and it is one of the reasons that i still subscribe. Sean
Sean...This issue of measuring power "at" clipping is new to me. I usually see specs quoted at some specified level of distortion (presumably an acceptable level). But isn't it true that distortion typically begins to increase abruptly just before clipping, and the manufacturer takes the power value just before this happens? I suggest that power measurements at the two points will differ only slightly.
As an aside, how do you deal with an amp that doesn't clip, because it has clip limiting. In that case I think you will just see increasing distortion as the signal peaks are rounded off.
All of this comes back to how the manufacturer wants to market the product and how the marketing department thinks will be the most effective manner at achieving those goals.
Given that i've seen amps rated at 200 wpc @ 8 ohms clip at 280 wpc, there's obviously some discrepancy and sometimes sizeable differences between ratings and reality. Given that the same amp was rated at 400 wpc @ 4 ohms and clipped at 418 wpc may shed some light on the subject.
While some might say that the manufacturer was just being "conservative" with their 8 ohm rating, others might say that the "doubling down" figure of 200 / 400 looked more impressive to the audiophile than a 250 / 400 rating. This is especially true since we are only talking about a percentage of one dB in rated power @ 8 ohms.
I think that a common question / statement in this type of situation, and one that i've seen here on Agon before is, "How much more would an extra 50 wpc get you when you've already got 200 wpc to start with?". Given the average audiophile's way of thinking, and this was already expressed in this thread above, isn't it easy to see why the manufacturer would "lie" about how powerful the amp is?
By down-grading the power rating of the amp by a "mere" 50 wpc, they can now pick up the "prestige factor" and "reputation" of building a "beast of an amp" that "doubles down". If it actually DID double down as impedance was halved at clipping, we would know for certain that it IS a beast of an amp with a very ample power supply and output stage. Seeing that it actually doesn't come close to doing that at clipping tells us that it isn't nearly as "beefy" as they want us to believe.
This is called marketing and it is pretty much what high end is all about. The more successful marketing is, the more product that they can sell and the more they can charge. The approach that i'm talking about removes the gloss of marketing and allows one to see the actual build quality of the entire product at a glance, hence the "newness" of a rating like this to you. They don't want you to know the truth about their product, they want you to be fooled by the hype. Sean
Sean, IS it possible to manufacture an amp for home use that doubles down at clipping to, say, 2ohms??
I'm not sure, but i think it's theoretically very dificult?
The Sunfire amps can do this with enough AC and the right internal fusing. I've got test results from an old magazine review on one of the original Sunfire's that verifies this. I'm not saying that i consider the Sunfire's to be the "best built" ( NOT by a longshot ), "built like a tank" ( they aren't ) or "the best sounding" ( no way ). What i am saying is that if Bob Carver could do this 10 years ago, i don't see any reason why others can't now. Especially "big name" amps that cost WAY more money and weigh two or three times as much.
One should bare in mind that this is NOT a "little" amp either. We are talking about a product rated at 300 wpc @ 8 & 600 wpc @ 4. For the record, this amp EASILY beat the factory ratings too, producing well over 700 wpc at clipping into 4 ohms and almost 1500 wpc into 2 ohms. I can provide the specific figures if someone is really interested, but i'll have to dig them up.
How much easier should it be to achieve similar results when starting off with a much smaller amp that needs less current, not as much of an output stage and even less heatsinking? There really is NO excuse for something like this other than cut-corner production and less than optimal design. Sean
Sean...To achieve the "doubling down" at clipping which you describe requires a substantial increase in the power supply cost. It is not a characteristic that is particularly useful in the home audio situation, where amps are not driven to clipping. (Pro sound is a different situation). While you are technically correct, I think that the law of diminishing returns comes into play here.
Again from my military electronics experience...when I went to work in 1961 "built to mil specs" really meant "overdesigned", and the poor taxpayer paid the bill. Recently we have built even the most complex military systems (such as a Fire Control System to prepare and lauch 24 ICBM) by assembling Commercial Off The Shelf (COTS) hardware. That old custom designed stuff was interesting work for the engineers, but common sense prevails.
Two equations tell the entire story. Watts=VoltagexCurrent, and Watts=Current(squared)x Resistance.
So as has been stated earlier an amp that is not current limited will double its power as the impedence is halved. So if an amp is rated at 200 watts into 8 ohms it must produce 400 watts into 4 ohms and 800 watts into 2 ohms.
But that really is only half of the story. To answer the question about how this impacts the sound you have to know what the impedance curve is for the speakers you have. I used to own Thiel 3.6s. They are nominally 4 ohm speakers but there is a point in the lower midrange where they dip to about 2 ohms as I recall. That means if you are driving a current limited amp (say 100 watt into 4 ohms rated amp) close to its limit, the speakers will be demanding 200 watts in this midbass region. If the 100 watt amp is only rated at 150watts into 2 ohms you could be running into clipping and a clear audio distortion. So high current designs are not necessarily required for many well behaved speakers. There are lots of high end speakers some cause trouble other wont.
The Krell Class A "Full Power Balanced" cx-series "double-down"
all the way to 2 ohms.
Take the "baby" of that line-the FPB-300cx: 300 wpc at 8 ohms,
600 wpc at 4 ohms, and 1200 wpc at 2 ohms. See:
Dr. Gregory Greenman
Morbius: Those are typical factory specs, nothing like what we are discussing here. On top of that, the voltage swing on that specific Krell is pretty low at their rated 138 v peak to peak. My Perreaux PMF-3150, which is also rated at 300 wpc @ 8 ohms, is rated at 180 v peak to peak. I would like to see what the Krell REALLY clips at into various loads. I have NO doubt that it is a sturdy amp, but exactly how sturdy, i don't know.
As a general rule, so long as current levels are reasonably up to snuff, the amp with the highest rail voltages will typically sound the cleanest and least compressed when driven hard. This is why tubes, with their very high plate voltages, can work quite well. This is true even though they are typically quite current limited in comparison to a "big brute" SS amp. The reason? So long as they have "enough" current to deal with the speaker load at hand, they've always got gobs of headroom due to the high rail voltages. Sean
The reason current is the issue is that high rail voltages collapse under demand and low impedences. Hi voltage measured with high impedence has nothing to do with an amps performance.
Keis: Your comment makes it look like high rail voltages have the potential to suffer more than a lower rail voltage. That simply is NOT the truth. ALL voltages collapse under demand IF there is a lack of current. The use of a higher impedance load simply reduces the need for as much current, but one could still run into sag IF the circuit was still not "beefy" enough OR the circuits response time was not quick enough. As such, rail voltages have little to do with sagging whereas they have a LOT to do with the dynamic capabilities / headroom of the circuit.
If one needed 1 amp, 10 amps or even 50 amps of current, so long as the circuit could deliver it on a timely basis, there would be minimal voltage sag. The idea behind a "TRUE" high current amp is that it can handle any load at any time due to always having an abundance on reserve. That "any time" part takes into account that the circuit is responsive enough to deliver the needed power as needed, not lagging and therefore sagging under demand.
This is why i said that an amp with high rails, high current, high speed, high stability and wide bandwidth is the ultimate amp. It can deal with any load or signal related condition thrown at it and do so with finesse. Sean
I really have to disagree with your remarks on the Krell
FPB-300cx. [ Are you sure you are familiar with the "cx"
I have access to some pretty sophisticated test equipment -
with pico-second resolution - and the Krell 300cx matches
its factory specs - even with heafty reactive loads.
Dr. Gregory Greenman
Morbius: You either don't understand what is being discussed here or aren't paying attention. I never said that the Krell won't meet spec's. What i did say is that we are discussing a different type of spec that Krell doesn't publish. For that matter, most manufacturers don't publish this spec as previously mentioned. While i'm quite certain that the Krell can deliver its' rated output at those impedances, i would be interested to see what it clips at into those loads.
Using the published specs of the Krell as an example, let me explain what i'm talking about. An amp that delivered 400 wpc @ 8, 700 wpc @ 4 & 1250 wpc @ 2, all measured at clipping, it would EASILY meet the 300 / 600 / 1200 wpc spec's. Having said that, it would not be a voltage source or a "true" high current design. It is simply an amp that is somewhat de-rated at higher impedances in order to look "more impressive" or "beefier" on paper.
A voltage source amp ( TRUE high current design ) using the same 400 wpc clipping figure would theoretically produce 800 wpc @ 4 ohms and 1600 wpc @ 2 ohms. While nobody could deny that the "made up" power output ratings at clipping of the amp above weren't "solid", such an amp truly isn't a "high current design". If it were capable of delivering all of the current that it needed, it would "double down" at clipping, not just double the "factory rated specs".
In this regard, the theoretical yet "beefy" amp that easily meets "factory spec's" diverges some 12.5% at 4 ohms from being a TRUE "high current" or theoretical "voltage source" amp. At 2 ohms, the amp nosedives to being 21.9% away from being "theoretically perfect" or a TRUE high current design. One should bare in mind that these "power levels", albeit completely theoretical and made up, are still VERY, VERY respectable compared to most commercially available products. Most amps made today don't come anywhere near these types of figures i.e. they are much worse / further away from being a true high current design.
For sake of comparison to the "theoretical amp" above, here's the factory ratings of the "baby sized" Sunfire amp:
300 wpc @ 8 ohms
600 wpc @ 4 ohms
1200 wpc @ 2 ohms
As you can see, these are the same as the Krell's published spec's that i referenced for the "theoretical" amp referenced above. The actual measured results of the small Sunfire amp at clipping as published in a review dated April of 1995 were as follow:
387 wpc @ 8 ohms
768 wpc @ 4 ohms
1460 wpc @ 2 ohms.
Theoretically, based on the 387 watt clipping figure @ 8 ohms, the "baby" Sunfire should have been able to deliver 774 watts @ 4 ohms compared to the 768 watts that it tested at. VERY close to theoretically perfect as you can see and less than a 1% discrepancy. On top of this, it is 28% more powerful than the published spec, which was already nothing to sneeze at.
As we move to the 2 ohm theoretical output based on the 387 watt clipping figure at 8 ohms, the amp should have been able to deliver 1548 rather than the 1460 that it produced. This is a difference of 88 watts and a divergence of appr 5.6% from theoretically perfect. Obviously, this is still a phenomenal figure but a greater deviation from "theoretically perfect". Even though the amp deviates further from being "theoretically perfect" or a "TRUE high current design", it still surpasses the factory rating by 21.6%. Again, a pretty whopping percentage, especially given the staggeringly high factory rating of 1200 wpc.
Outside of that, one has to wonder if the mains feeding the amp and / or the original small gauge factory hard-wired power cord weren't responsible for the reduced output @ 2 ohms. This is quite possible due to the amount of current draw required from the wall to output this type of power. If the AC system itself were sagging or the factory hard-wired cord were restricting current flow, it would therefore lower the power potential of the amp itself.
Given that these test results were generated on a first generation Sunfire with a somewhat "scrawny" ( 16 gauge ) hard-wired power cord, and the newer models have had several upgraded revisions ( increased power supply reserves ) performed to them with the option to use your choice of detachable power cords, it is possible that they could come closer to being "theoretically perfect" in terms of "doubling down" at clipping.
As such, one can see that most amps truly are NOT "high current" designs. In fact, most are FAR from it. Some amps are "higher current" than others, but as a general rule, "high current" is simply another marketing term used as a buzzword for audiophiles to be seduced by. Sean
PS... For sake of clarity, the figures used on the first "theoretical amp" were only based on the rated figures of the specific Krell amp mentioned above. The actual figures have NOTHING to do with the actual performance of that specific Krell product. It may perform better or worse at any given impedance, i don't know.
Sean...To put it briefly...we don't operate our amps at clipping. What we care about is distortion-free power, and that's what is usually quoted.
BEAR this in mind :-)
El: I think that you missed the point. The ability to deliver the amount of "clean" power as needed in a timely fashion on a dynamic basis at any given frequency or impedance is what i'm talking about. The closer that one comes to achieving that goal, the more refined, musical and natural their system will perform. The lack of smearing, strain and ringing combined with the improvements in liquidity and harmonic structure that one experiences is an eye & ear opening experience. The music is no longer coming from a stereo system, sounding canned, compressed and "box-like", but actually spreads out and presents a very dynamic yet subtle panoramic view into the recording.
Much of this comes from the improved response times and control that a non-current limited wide bandwidth high dynamic headroom system provides. If you actually measured the peak power required to reproduce specific types of dynamic transients, you might be pretty shocked. When you factor in that the impedance of the speaker varies over the frequency range and that more / less power & ability to control and respond to the signal and load may be required simultaneously, one begins to understand that you can never have "too much", so long as it can respond on a very timely and dynamic basis.
Obviously, one can avoid some of these pitfalls by picking speakers that present a gentle impedance curve with lower levels of reactance, but such designs typically tend to sound somewhat "stifled" to me. That's because the manufacturer has typically added quite a few parts to the crossover to tame specific resonances and problems. This ends up sucking the life and harmonic structure out of the music and reducing the quality of the amplifier / driver interphase.
I guess that it boils down to the fact that we are damned if we do, damned if we don't. There are obviously compromises to be made in all but the very most expensive, well-designed systems out there. I always strive to try and reduce the potential for technically related problems, but when it comes down to it, the bottom line is that it has to sound good. Luckily, resolving most of the technical issues ends up sounding markedly better, so the two goals seem to work hand in hand. Sean
Sean...How could I have missed your point? All those beautiful audiophile words..Clean power, refined musical and natural, lack of smearing strain and ringing, harmonic structure, yada yada,yada. I just don't agree with all your conclusions.
Until one has heard / experienced a system that is capable of such things, their standard of reference will be lower than someone that has had such an experience. Based on common sense, my previous education and a lack of exposure / experience, i used to think that a power cord couldn't produce either measurable or audible differences in audio components. As such, i clung to that belief quite adamantly. After opening myself up to such ideas and conducting some simple tests, i experiencing quite the opposite first hand. This not only changed my point of view, but helped me come to understand how / why this was possible.
What i'm getting at is that man doesn't know nearly as much as we think we do. Life is a learning experience and audio is no different. One can learn / unlearn as much as they want to, based on how much effort they are willing to put forth in doing so. There are things that i now know to be true that i would have worked hard to refute just a few years ago. Those things that i believed to be accurate up until a few years ago were based on many years of first hand experience and education. The bottom line is that i was lacking both education and experience in certain areas and until i opened myself to learn more in those areas, i was stifling my own personal growth and that of my audio systems.
If this sounds like i'm on a pulpit preaching, it's not meant to. It's simply meant to explain why some folks may have different points of view than others do, even if they may have similar qualifications and / or backgrounds. The same can be said for audio components. They might measure similarly in certain areas, but be quite different in others. Thinking that they all sound the same would be equivalent to thinking that all people are the same. This is obviously not true and both cases are quantifiable. Sean
With all due respect to all, what Sean is saying is hardly revolutionary -- and thereby, hardly controversial. All he's describing is an amp where the manufacturer is (truthfully) saying: "this amp is stable down to 2ohms and I guarantee that at that (resistive) load this amp will produce 1200W of energy; similarly, it will produce 600W if the load is 4ohm and 300W if the load is 8ohm".
What said manufacturer did NOT openly claim is that these are peak power/ energy ratings before clipping.
All SEAN is saying is that one shouldn't construe these specs as such. Some manufacturers may be refering to peak W at clipping while others NOT: I.e., what the manufacturer COULD be saying is "this amp will put out a MAX of 1200W at 2ohms, well over 600W at 4ohms, and substantially over 300W at 8ohms".
In the old times some manufacturers were even proud of quoting stability at 1ohm to show how well their product had been implemented.
Not just the old times. I love my 1 ohm speakers. It is comforting to know there are amps specifically designed to mate well with my speakers.
Power measured at clipping and peak power are different things. Peak power will typically be higher than power at clipping due to the pulsed signal that is very short in duration. The high intensity, short duration signal does not stress the amplifier nearly as much, reducing sag in the power supply and resulting in greater momentary power output.
The test for peak power output is typically referred to as "Dynamic Headroom" and isn't all that uncommon of a spec. The dynamic headroom spec simply tells you how much power above the rated output that the amp can deliver for a very short period of time. Most manufacturers are glad to provide such a spec and as Greg pointed out, it is NOT what i'm talking about. The specs as posted for the Sunfire were taken at clipping ( max sustainable output ), not using momentary short-term tone bursts. Sean