A little deeper on amp power please....

Best analogy I can think of is:

Think of it as a high-power race car that can go 200 miles an hour. Even at the required speed limit, it will display an effortless response that you can feel.

ozzy

I'm using a 2 wpc amp right now with great success.  I've been told that the "rule of thumb" is that you need 20dB headroom which means that if I'm listening at 70dB, I should have enough power available to go to 90 dB just for the extremely loud parts.  I think they are especially referring to classical music that has tremendous variations in volume.  That said, my 2 wpc amp will indeed play much louder than I need it to.

So I've been thinking about it quite a bit and critically listening to my system, trying to find a weakness at higher volume. And I can't.

I'll keep reading this thread but I think there are 2 reasons people say this:  1.  It is natural to feel more secure having more power than you need.  2.  Cheap amps that were overrated in the first place and probably has nothing to do with us here.

I'm sure my skepticism will inspire replies.  I'm open minded.

Jerry

As a low volume listener with hard to drive speakers, I think total wattage is less important than how the amp handles changes in impedance. It is a good sign if the wattage doubles as the impedance halves. To use the race car analogy, I think of this as having lots of torque even at low RPM’s. 

I forgot to mention, I agree current capability of the amp is more important. --Jerry

Speakers are a much more complicated load than the fixed value resistor used on a test bench to determine the power spec rating for an amp.  First, speakers typically have a resistance curve that varies widely with frequency. An 8 ohm speaker may vary from below that number to way above it. Second, many speakers also present a capacitive load which can be difficult for some amps. And, the amp also has to deal with the signal it gets back from the speaker. Once a speaker starts moving, particularly at lower frequencies, it has inertia, so continues vibrating momentarily after the input signal stops. This turns the voice coil/magnet assembly into a power generator which can return power back to the amp. How the amp reacts to this signal can be an issue.

So, you can see this issue is more complex than just how many amplifier watts are generated. Some amps handle complex loads better than others. The specs alone rarely cover all the bases.

I think the reality is a little complicated, but you are right that most of us never exceed 20 Watts listening.

I think it may best be explained that some speakers are a lot harder to drive than you would think, and an amp that can handle them without sagging tends to be bigger, heavier and higher power.

By sagging I mean both reduced overall output but also changes in frequency response. That is, the ideal amplifier would output the same frequency response regardless of speaker, but as we get further and further from ideal the amplifier’s output tends to start to mirror the speaker impedance, even at low listening volumes.

One of the best sources of information about this are the Stereophile amplifier reviews.  Take a look at the measurements where they test with a simulated speaker load, and pick a few tube amps vs. a few solid state amps and you'll start to see a pattern.

Nothing is ever easy, and I will need to learn more. I get that power, impedence, and damping factor will all play, and then I need to learn more....

What inspired this is that I thought I had my end-game speakers in Harbeth 40.2. But then I got interested in playing around with low powered gear (I know it can be expensive too, but there are definitely interesting economies of design and $$$ in producing, say, 5 high quality watts vs 100). Some people say they have had good experience with low power amps and these speakers, others have said running SS monoblocks is a must-do! With unlimited space and budget, I would love to add a set of high sensitivity speakers for experimentation. (can somebody loan me their La Scalas?)😂

One of the things I learned along the way is that there is a really big difference between tubed amplification and solid state. For solid state, more power… of the same quality make most speakers sound better at most volume… even at nearly a whisper. This is also what is “said” in audiophile circles and it is typically very obvious unless you have very efficient speakers… then it can be a liability. Why? I have not really found investigating why results in a means of determining if speakers match amps other than at the most general way… like don’t pair a 2wpc amp with 82db efficient speakers.

Most speakers are much less sensitive to the overall power from tube amps… quality is more of primary concern… so I used to use a 350wpc SS amp to power my 90db speakers and dropping to 250 wpc would make a difference. I replaced it with a 140 wpc tube amp and they sound very different (the tube amp sounds way better… and different). I can operate my tube amp at 140wpc or 70wpc in triode mode and there is virtuallly no difference in volume or punch, but the 70wpc sounds better and more musical. Why? I have not really found investigating why results in a means of determining if speakers match amps other than at the most general way… like don’t pair a 350wpc amp with 100db efficient speakers.

So after fifty years of passionately pursuing high end audio what is my conclusion? If you are trying to come up with a set of specs that will lead you to the right choice of amp forget it. You must try them. However, if you are simply interested for the sake of learning… carry on… good luck. But still make sure you audition amps first… especially if you have a big spread in the sensitivity / amp power.

Also, one of the most amazing / interesting systems I have heard was a huge panel electrostatic powered by a 4 wpc Viva tube amp. Which was the most ridiculous pairing that would never be something that theory would have led you to put together.

I do have r3 red rose speakers , Iam using Norh 100 w monoblock, Iam very happy . Until I read one r3 owner said the t3 shines on 250 w amp above. I happen to have musical fidelity 550 supercharger mono, Indeed the R3 did sound amazing on more power.

@mathiasmingus Wrote:

A little deeper on amp power please....

See articles below:

Mike

https://sound-au.com/articles/amp-power.htm

https://www.firstwatt.com/index.html

I’m going to channel Ralph @atmasphere here because what I learned from him is that you don’t want your amplifier working toward the upper range of its capability.  I assume that’s true of both tubed and solid state amps.  Am I wrong on this Ralph?

@soix Got it in one 😁

SETs are where this is most important- they only have about 20% usable power out of whatever they are rated for.

You really don't want any amp to be asked for more than about 50% of full power even on peaks. The harder you work any amplifier, the more distortion you're going to get. That also means that the load for the amp should be benevolent rather than having odd phase angles in the crossover, low impedance or the like that tax amp's capacity. The more distortion, the less musical, since that distortion is likely to be the kind that the ear hears easily!

I have always found the sound from any given pair of speakers to be smoother (less grainy and more natural) when driven with larger power amps, regardless of listening level. This impression was less pronounced with tube equipment than SS gear, due to the "soft clipping" nature of tubes and their inclination toward even-order harmonics, but still apparent.

I owned a well regarded 255wpc powerhouse Hafler DH500 in 1986. My 2023 amps are ~ 17 wpc in triode, 30wpc ultalinear. Granted I don’t try to play as loud as I used to, but the system sure sounds better now.

More power doesn’t equate directly to better sound, unless the amp is straining to produce the needed power. Lower powered amps can sound abso-freaking-lutely amazing...,within their power range. Sound quality really boils down to the given circuitry of an amp, and it’s ability to power a given speaker at the desired volume.

And I do believe there’s an audible difference in the sound of different "well designed" amps, even at the lower volume ranges. If you don’t hear differences, start over, and try a different tact (if you’re interested in that sort of thing)....otherwise, enjoy the music. 😎

I happened to be using 2a3 monoblocks when I bought Klipsch Jubilees and was selling my KHorns. The buyers wife wanted to hear vintage rock on the Jubes which I obliged. She asked for “louder, louder” when at some point I indicated that’s all there is. Even with speakers rated at 105db 2-3 watts can be woefully inadequate when asked to do too much. Normal listening was divine, rockin’ the house was less than spectacular. 

Power or wpc do not translate into sound quality.  The Class A mono block tube amps we build easily outperform many higher wpc amps on many different speakers at various ratings or recommendations by the manufacturer.  We drove a pair of Vandersteen model Sevens 83.5db with a 28wpc stereo tube amp from ART Audio with no issues.  So you need to experiment.    

Audio peaks being measured can hit acoustic transient peaks up to 10 x the average from low to highs , That is why clean power for one is most preferable more so then just wpc ,that being said though more is better to have . transient distortions is what blow especially tweeters more times then anything else.

Amps to speaker equations do seem complicated. Having 100wpc tube amps on 93 dB speakers seems enough for me. Although not knowing the phase angles and impedance of the speaker calls into question the compatability of the pairing, headroom should compensate. Dynamic range is such an important factor and those pesky peaks demand so much, having as many quality watts as possible, to me seems a good idea. Trying to get by with borderline power is just such an annoying proposition when it's time to crank it up.

Here's an example that I hope sheds some light on your question. I have gone down the road of low sensitivity/difficult impedance speakers driven with a large amp. My Thiel CS6 speakers have an 86 dB sensitivity and the impedance drops below 3 ohms for a significant part of the frequency spectrum. They also present a difficult phase angle which requires a large amp that can supply high current. My Krell KSA 300S amp outputs 300 watts @ 8 ohms, 600 watts @ 4 ohms, 1200 watts @ 2 ohms, and 2400 watts @ 1 ohm. Very few amplifiers can do this. BTW, this amp weighs 185 lbs and when run loud gets hot enough so that the heat sinks glow in the dark (OK, I'm kidding about the glowing part but it's too hot to touch).

When you hook up an amp to these speakers that is not designed to provide high current into lower impedances the amp acts like a tone control. For the CS6 the lowest impedance is in the upper bass/lower midrange region and the speakers would sound thin in the bass with a typical tube amp. Will the amp work? Sure, it will make sound. You may even like the sound but you are not hearing what the designer intended.

The difference is not subtle. When I had my KSA 300S recapped I tried hooking up my CS6's to my 100 watt Onkyo AV receiver. I literally laughed out loud when I played the first cut. The sound was awful - the bass was anemic, the soundstage collapsed, and the dynamics were compressed - even at low volumes. BTW, this receiver sounds just fine driving a pair of Polk Audio speakers that have a more conventional sensitivity and impedance curve.

Your Harbeths are low sensitivity but present an easy impedance curve which would allow you to experiment with low wattage amplifiers. I suspect these speakers would sound better with more watts but it would be an interesting comparison.

Here's a good video by John DeVore about this subject. He designs high sensitivity speakers with flat impedance curves that are easy to drive. My Krell would be total overkill driving one of DeVore's models. DeVore Rant

It has been so long that I can't remember where the original thought of more is better came from.  Given that it's a good "more" not just crap.

I went tri-amp a few decades ago and will never go back.  At the moment using two Pass mono X260.8(s) and two Pass x30.8(s).  I must say the sound is effortless and yes I really let it have at times.

The beauty is that some thunderous bass doesn't change the mid or highs because they don't know what the hell is going on. Headroom is good for the soul.

YMMV

Regards,

barts

@8th-note 

Good post.

Think of it as low end grunt you have effortless power at low end instead of thinking it as a car with high horsepower think of it as one with a lot of torque that can push you off the line effortlessly... It does make a difference

+1 @carlsbad2 @8th-note @erik_squires 

Glad that current delivery came up soon here.

Although many other factors reside in dynamic response, current capability and response to impedance dips seem to rise to the top in my experience. Factoring in that all speakers are different in how they behave with any given amp or amp topology, that the type of music is also a factor, even cable guage for that matter.

But there really are so many things that can occur differently between systems of higher power with lower sensitivity and those with lower power and higher sensitivity. Even if dynamic response can be noted as similar in any given section of music (with volume matching), there can still a difference in tonal character and decay. There are reasons why many gravitate toward low power, high power, or something in between. But what seems to remain constant is that need for sufficient headroom, at least relative to the needs of any given listening environment (types of recordings, near field or large area etc.).

I think that in addition to power, there's the issue of feedback vs. output impedance. 

By using massive output stages (relatively) you reduce the need for feedback to achieve low distortion.  This tends to only work with high power amps, but surely not all of them.

I suspect that a lot of what we think of as the need for power is really the need for high current, and that with ideal 75-100WPC amps with many output devices and low feedback we'd be very happy.

@8th-note If I can add to that:

The power paradigm

I suspect that a lot of what we think of as the need for power is really the need for high current

@erik_squires Current cannot exist without power and vice versa. So if the amp can make the power into the load, it can also make the current.

There are a variety of issues with feedback. One is that the feedback node (where the feedback is applied in the amplifier design) usually is non-linear, meaning the feedback signal gets distorted prior to doing its job. This usually means additional distortion generated on that account, which is mostly higher ordered harmonics (and so IMD as well).

That is part of why feedback has gotten a bad rap in high end audio, but its not feedback's fault so much as shoddy execution.

The second problem is most amplifier designs lack the gain bandwidth product to support the amount of feedback being used as well as the gain of the amp. The result of this is that above a certain frequency, distortion increases. IME this phenomena is highly audible. Most THD measurements are made at too low a frequency to really show what is going on, but distortion vs frequency shows this problem with ease.

Class D offers a way around both problems. I think that is why there are class D amps now that are shutting down class A solid state designs for sheer musicality.

Current cannot exist without power and vice versa. So if the amp can make the power into the load, it can also make the current.

All these years and you still assume I'm typing in the least informed context, @atmasphere , give me a little credit?

I meant to say, being forced to now give an exhaustive/exhausting answer:

Our search for amps that are rated high into 8 Ohms may actually be a search for load invariant amps, that can sustain the voltage even as the impedance drops and the phase angle varies from resistive.   The modern day equivalent of the Krell KSA 50 (50w/channel into 8 Ohms) which was low rated power but able to drive 1 Ohm loads is a rare find indeed.

I think that is why there are class D amps now that are shutting down class A solid state designs for sheer musicality.

Ralph, others have posted here and there with the question "what is meant by musicality?". They do so because the word is vague and not particularly helpful. It connotes something positive and nothing more. When a food critic uses the word "tasty" I cringe and when a person near me claims something he or she is eating is  "tasty" I resist the urge to say "sh*t has taste, but I am pretty sure I don't like it". 

I would love to audition your Class D amps one day but I have heard others and to me they sound very clear, fast, but sterile and without soul. 

And where is the support for the "shutting down" portion of your statement? At the lower end of the cost spectrum certainly. We all know why. And for in-board speaker modules, sure and we again all know why. 

I actually do own and use two digital amps. One is inside a Rotel home theater receiver that replaced a nearly identical Class AB unit that became an expensive door stop due to not offering HDMI connectivity. It sounds markedly inferior to it's Class AB predecessor. The sound is dry and one-dimensional which is no easy feat when there are seven loudspeakers spread around. The other is installed under the roof of an outdoor pavilion covering our patio and powers low-end KEF loudspeakers. It sounds like good crap too. Good for the cost and intended use, but crappy. 

I think that is why there are class D amps now that are shutting down class A solid state designs for sheer musicality

There are some megabuck Class A amps I can't stand to listen to for more than a few minutes at a time, while I've never had a problem listening to ICEpower, so I could see some truth in this.

All these years and you still assume I'm typing in the least informed context, @atmasphere , give me a little credit?

I meant to say, being forced to now give an exhaustive/exhausting answer:

Ralph, others have posted here and there with the question "what is meant by musicality?". They do so because the word is vague and not particularly helpful. It connotes something positive and nothing more. When a food critic uses the word "tasty" I cringe and when a person near me claims something he or she is eating is  "tasty" I resist the urge to say "sh*t has taste, but I am pretty sure I don't like it". 

@erik_squires @fsonicsmith These are both valid criticisms!

Erik, I give you the credit, the reason I responded the way I did is because that sort of shorthand promotes confusion with those that aren't in on the shorthand. I see confusion about this topic all the time so its become a bit of a sore spot.

By 'musicality' I mean its easy to listen to all day, plays bass right, isn't harsh or bright, has good detail without being clinical, has an 'organic' presentation, yet is uncolored. Sorry for the shorthand on my part.

As far as 'shutting down'; most class A and AB solid state designs lack the Gain Bandwidth Product to support the feedback they use; in addition the feedback is misapplied so adds higher ordered harmonics (see the writings of Norman Crowhurst and Peter Baxandall on this issue; neither proposed a solution). The exception might be a zero feedback class A amp which can have ruler flat distortion vs frequency measurements.

Regarding your commend about lower end of the cost spectrum, I think one reason class D has fared so poorly has been insufficient power supplies. IME a power supply for a class D amp has to be robust but a lot of designs I've seen don't seem to reflect that.