2. tech-talk
  3. 132841-3105

Which watts are the right watts in SS amps?

Hello Sports Fans!

More than a few people over the years on these pages have said only those SS amps which double down in output power as impedance drops are truly special or worthy amps. Eg., 200 @ 8ohms; 400 @ 4 ohms; 800 @ 2 ohms; etc.

Not every SS amp made does this trick. Some very expensive ones don’t quite get to twice their 8 ohm rated power when impedance halves to four ohms. BAT, darTZeel, Wells, and Ypsalon to name just a few.

An amps ‘‘soul’’ or it’s ‘voice’ is the main reason why I would opt in on choosing an amp initially and keeping it. Simultaneously , I’d consider its power and the demands of what ever speakers may be intended to be run with it or them.

I’ve heard, 80% of the music we are listening to is made in the first 20wpc! I’m sure there’s some wisdom in there somewhere as many SS amps running AB, are biased to class A Only for a small portion of the total output EX. 10 – 60 wpc of 150 or 250 wpc.

After all, any amps true output levels are a complete mystery when anyone is listening to music anyhow.

I suspect, not being able to actually measure true power consumption, the vast majority of listening sessions revolve around 60wpc or so being at hand with traditional modern reasonably efficient speakers.

Sure, there are those speakers which don’t fit into the traditional loudspeaker power needs mold such as panels or electrostats, and this ain’t about them.

The possibility of clipping a driver is about the only facet in amp to speaker matching which gives a person pause while pondering this or that amplifier.

I feel there is more to how good an amp is than its ability tou double output power with 50% drops in speaker impedance.

However, speakers are demanding more power lately. Many are coming out of the gates with 4 ohm ‘nominal’ IMPs which lower with fluctuations in frequency. Add in larger motors on larger drivers, multiple driver arrays, and on paper these SOTA speakers appear to need more power.

IMHO It is this note which introduces great concern.

I’ve read every article I can find on Vienna Acoustics Music. Each one says give them lots of watts for them to excel.

Many times good sounding speakers I’ve owned sounded better with more power, albeit from arguably a better amp.

I tend to believe having more than an adequate amount of cap power is indeed integral. … naturally the size and type of transformers in play possess a strong vote for an amps ability to successfully mate with speakers.

Controlling a driver’s ability to stop and restart is as well a key to great sound and only strong amplifiers can manage this feat. Usually this gets attributed to ‘damping’ factor, but damping as I read it is more a shadow than a tangible real world figure as it depends on numerous factors. Speaker cable length alone can alter damping factors.

A very good argument exists about those mega watt amps voices. Each 500 or 600 wpc amp or amps, I’ve heard have had stellar voices too, not merely more watts.

So is it predominately these mega watt power house amps souls or their capacities that fuels the speakers presentation?

Would you buy an ‘uber expensive’ amp based more on its voice or soul, than on its ability to output loads of watts, even if you feel the amp may be somewhat under powered for the application?

Choosing this latter option also saves one money as the more powerful amps do cost more than their lower outputting siblings.

Please, share your experiences if possible.

Tanks muchly!

blindjim

Showing 6 responses by atmasphere

atmasphere's avatar
atmasphere
12,138 posts
 
First 5 – 7%? I’ll presume this initial area pertains directly to the scenario you presented, otherwise, isn’t the first 10 – 20% of a class AB amp where one finds Class A functionality, and many adore that area of operation.
@blindjim  Well they do say its all about that 1st watt, but in the case of less efficient speakers it might be about the first 5 watts in the case of some behemoth amps.

(If you've ever heard the maxim that lower powered amps tend to sound better, this is one of the reasons why. IMO, there are some fairly musical low- and mid-powered solid state amps, but there aren't any that are really high powered. As for tube amps, there are very few high powered amps that sound like real music either!)

Class A is helpful for reducing distortion. Most AB solid state amps go from class A to AB within the first watt or so. But even class D amps which operate on a different principle (as well as many push-pull tube amps) have this quality of increased distortion at lower power levels. So even if they were all perfect voltage sources, you would still want to match the efficiency and power handling of the speaker against the low power distortion issues of the amp.

For this reason, I suspect our grandchildren will still be talking about equipment matching- we're not seeming to be solving this issue anytime soon.
atmasphere's avatar
atmasphere
12,138 posts
 
Go ahead and spend gobs of money on those pure class A amps that burnout after 7-10 years
Any truth to this? (i.e. is a class A amps life span known to be less than other designs)
No. We've been making class A amps for the last 41 years, and recently serviced one that had been in the field since 1988.

Just make sure that they can get adequate cooling and they are fine.

I'd like the process of component pairing to be less guess work and more analysis.
Me too- but its not going to happen unless the audio industry cares enough about quality of reproduction to address the shortcuts (including measurement issues) that have been taken so often in the last 50 years. Until then, equipment matching is a thing :/
atmasphere's avatar
atmasphere
12,138 posts
 
I'm not sure I've ever seen any transformer coupled amp show damping over 100 into 8 ohm. I'm also not convinced that the drawbacks of feedback are so bad it should be completely avoided. I feel like thats a knee jerk reaction like many people have about guns. It's just a tool to be used.
The thing about feedback is there are more variables than the formula taught in school! You have to deal with RFI being injected into the amp via the speaker cables (and most designs don't...) and the amp has to have sufficient phase margins prior to application of the feedback... here is a great article about how to apply feedback properly:
http://www.normankoren.com/Audio/FeedbackFidelity.html

Since feedback is rarely applied with this sort of care, its incorrect to say that eschewing it is a 'knee jerk' reaction. Its misapplication is audible!

With regards to speakers, there is never a need for 100:1 (or more) damping; almost any speaker made is overdamped with a damping factor that high. Here is another great article that despite its age has stood the test of time, written by the head engineer of Electro Voice:
http://www.dissident-audio.com/Loudspeakers/CriticalLSDamping.pdf

Note that at the time of this article, there were speakers that needed far less damping than even 5:1, and there are still speakers around today that barely need that or even less (as an example, Nelson Pass demonstrated a small yet very impressive open baffle speaker at RMAF about 10 years ago that was driven by a current source amplifier, so the damping factor in that system was less than 1:10 and you did read 1:10 correctly, yet it played bass quite well).

It was Electro Voice and MacIntosh that led the way in championing the idea that the amplifier should be a voltage source and the speaker voltage-driven, back in the late 1950s and into the 1960s. But the simple fact is that this model simply did not and does not encompass all speakers and speaker technologies (any speaker that does not employ a cabinet would be an example, as well as many speakers that do have cabinets..). In addition, all amplifiers that are capable of acting as a voltage source are push-pull, and such amps (with rare exception) have distortion characteristics wherein the distortion actually **increases** below a certain minimum power level, typically about 5-7% of full power. So if you use a speaker that is too efficient for the amp, you will not be getting the best sound out of it.

The voltage model is intended to eliminate equipment matching issues, and if you are only interested in box store mid fi, none of this is a problem. But in high end audio there is quite a bit of diversity such as planar speakers, horns, full range devices as well as more conventional designs, SETs, OTLs, class A, class D, etc. because everyone is approaching the goal of electronics sounding like real music in a different way. So if you apply a generalization such as 'all amps should be voltage sources' right away it is belied by very notable exceptions!

The bottom line is that equipment matching is still very much with us and won't be going away anytime soon.
atmasphere's avatar
atmasphere
12,138 posts
 
However, transformers do have drawbacks. Many speakers don't expect the low damping factor transformers provide and they drive a lot of speaker hot in the highs. There's a reason DC coupling is the standard in solid state amps. It gives the output stage maximum control over the load.
We direct-couple our amps and they are vacuum-tube. An output transformer does not mean that you get a low damping factor. Its far more complicated than that!

Tube amps, including OTLs, can operate at voltage sources much like solid state. It all depends on the design and intention.

The thing is, with tube amps you get linearity, but usually not a low output impedance. So you add loop feedback and then you have the low output impedance. With solid state, you don't (usually) get linearity but you usually have a low output impedance. So you add loop feedback to obtain linearity.

The problem is that loop feedback, while suppressing distortion, adds some of its own in the process (this fact has been known for decades- see the writings of Norman Crowhurst). Some of it is IM distortion (which is highly audible) and some is higher ordered harmonic distortion (which is also audible). This is why amps with feedback tend to sound brighter (and also harsher) than amps without.

To this end, loop negative feedback is eschewed by many designers as they don't want coloration. Contrary to what has been stated elsewhere on this thread, its possible to design a speaker for an amplifier that has a higher output impedance (which it will have if it has no feedback). Such a speaker does not have to have a flat impedance curve either! Its also possible to find speakers that will obtain flat frequency response even though the amp has a high output impedance, despite the fact that the speaker was not designed for such an amp.

Its easy to see in the specs of the speaker when it will work with amps of higher output impedance. We've stayed in business for over 40 years doing just that. But there are also speakers that are designed with this expectation- Audiokinesis, Coincident Technology, Merlin, Altec, JBL, Tannoy...and many more. Intention also plays a big role.

So there is more than one way to obtain flat frequency response. The thing is, higher ordered harmonics are very audible to the human ear as the ear/brain system uses them to sense sound pressure. So if the system has more higher ordered harmonics than it should, it will sound bright and harsh despite the actual distortion being quite low. This is why amps with a high output impedance exist; its not because somehow we designers can't make them low impedance- we can! - its because we are trying avoid not just some colorations but **all** colorations, and a higher output impedance is a result of that. But no worries- if you want accurate reproduction that sounds like real music, look at it this way:

If the speaker requires that the amp employ feedback to sound right (regardless of why the amp has the feedback), **with today's technology** that speaker can't ever sound like real music- it might sound like a really good stereo, but it won't have the feeling of real music.


atmasphere's avatar
atmasphere
12,138 posts
 
I’m reading thru amp measurements done here and there, and often the terms get conveyed as ‘db’ losses or gains of voltage, I suspect. When the DB comes up, I am lost.
@blindjim Decibels are a logarithmic expression. They are handy because our ears are logarithmic as well. Voltages and audio equipment in general are linear expressions, so sometimes conversion is needed.

1 decibel is the least difference the ear can detect.
3 db is a minimum easiest change in volume that we hear. +3db requires twice as much power, -3db is half the power
6 db is a doubling of voltage, not power. This can be a bit confusing!
10 db is what we perceive as 'twice as loud' and requires 10X more power
20 db is 100x more power.  30db is 1000x more power.

In amplifiers there is something known as 'golden decibels', an expression that comes from the radio broadcast industry. 3 db represents a doubling of power, and to hear any significant increase in volume you need 3 db, so this gets increasingly expensive and in amps, the more power the less likely it will sound like music. This is why speaker efficiency is so important.

Voltage gain can be expressed as a ratio and can also be expressed in decibels. Here is a handy calculator if you need to make the conversion:
http://www.sengpielaudio.com/calculator-gainloss.htm
atmasphere's avatar
atmasphere
12,138 posts
 
Most tube amps are the equivalent of a graphic equalizer - hence a no no for me.
If the amp employs about 20 db of loop feedback, it is capable of acting as a true voltage source, whether it can double power into half the impedance or not. IOW, it can be within 0.5db if there are no limitations in the bandwidth of the output transformer. The trick is not to look at things at full power, but to observe what is going on at 1/10th full power, where most of the amplifier power is likely spent. At those power levels you will see any voltage source acting like any other voltage source, doubling power when it needs to or cutting it in half when it needs to.

Most of the time the coloration that people really complain about in a tube amp is the 2nd harmonic, which adds richness. Solid state amps usually lack this on account of being fully differential. Its possible to build a tube amp fully balanced and differential as well, and if so built they too will lack even ordered harmonics as they are canceled throughout the circuit, not just in the load.

However there is a price that solid state exacts of its owners- they too have coloration, only in this case it is higher ordered harmonics. They are certainly at a low level, but the human ear/brain system uses higher ordered harmonics to sense sound pressure (likely because pure sine waves are non-existent in nature and so are not part of our evolution) ; as a result humans are very sensitive to higher ordered harmonics and can hear them easily. The audiophile terms for this are 'bright', 'harsh' and similar turns. This explains why two amps can have similar bandwidth within 0.5db yet one might sound bright and the other not. IOW we respond to distortion by perceiving it as a tonal coloration.

This simple fact is why tubes are still around decades on after being declared obsolete. If you can eliminate the oppressive nature of the top end in a system, then the system might get more listening time.

IOW the real reason behind the tube/transistor ad nausem is all about distortion.
Check your reference… you must have meant someone else yet did not include their handle to indicate your words were directed to them..
 R
@blindjim , my remarks were quoting George from the other thread by the same name in the amps section.


More to discover