Jump to content

Some thoughts on power requirements.....


tube fanatic

Recommended Posts

7 hours ago, jjptkd said:

Well, I've been sitting here listening to the new RF-5's and figured I'd pull out the old SPL meter. It appears as though my normal listening volume at my seated position is about 65 db. With speakers rated at 98 db at 1 watt I'm guessing I'm barely using any of the 405 watts on tap but it sounds very good with plenty of detail and bass. I know the room plays a big part in that being only 13x14.

If we accept the 98 db rating as being correct, and you are sitting 8 feet from the speakers (just a reference distance), you only need 1.5 MILLIwatts to create your 65 db listening level!  This translates to needing only 150 MILLIwatts to handle a 20 db peak in the music.  These numbers are not an exaggeration, and this is why using flea/low power amps with Klipsch speakers is possible.  When I do demonstrations for the "uninitiated", they often refuse to believe me when I disclose the power output capability of the amps they are listening to.  With your room being only 13 X 14, and such a low listening level, the chances are the room isn't contributing too much to the resultant sound.  And, the best part is that such a level in a room that size allows you to experience imaging which is simply not possible at the extreme listening levels which most seem to prefer.

 

Maynard

Link to comment
Share on other sites

15 hours ago, jjptkd said:

Well, I've been sitting here listening to the new RF-5's and figured I'd pull out the old SPL meter. It appears as though my normal listening volume at my seated position is about 65 db. With speakers rated at 98 db at 1 watt I'm guessing I'm barely using any of the 405 watts on tap but it sounds very good with plenty of detail and bass. I know the room plays a big part in that being only 13x14.

This is similar to what I've been saying for years. PWK too.

 

The "First Watt" posse is arguably preoccupied with low power. In my view they're way off. Too much power. Further, people ask about a good amp. There is little information about distortion at 1 watt, much less for 1 milliwatt. Class A has to be a first choice or AB where it operates in A. 

 

For K-Horn, LS, and Belle's consider the numbers:

104 dB, 2.828 volts, 1 watt.

94 dB, 0.894 volts, 0.1 watt

84 dB, 0.282 volts, 0.01 watt

74 dB, 0.0894 volts, 0.001 watt. (i.e.  1 milliwatt) (under one-tenth of a volt) (0.011 amps)

64 dB, 0.0282 volts, 0.0001 watt. (i.e. 0.1 milliwatt) (under three one-hundredths of a volt)

 

Again, "The First Watt" ??? In a pig's eye! It should be "The First Milliwatt."

 

Smile.

  • Like 1
Link to comment
Share on other sites

On September 23, 2016 at 11:56 PM, jjptkd said:

Well, I've been sitting here listening to the new RF-5's and figured I'd pull out the old SPL meter. It appears as though my normal listening volume at my seated position is about 65 db. With speakers rated at 98 db at 1 watt I'm guessing I'm barely using any of the 405 watts on tap but it sounds very good with plenty of detail and bass. I know the room plays a big part in that being only 13x14.

Yeah, after shows I usually cool down with my LaScalas at a whopping 60 dbA at home. Any more I try to keep even pro PA systems around 90dbA at FOH (console) position, this is usually more than ample at the first 10 rows/dance floor. 

 

We all listen too loud. That's why the damned titanium tweeters sound good to so many of you guys - you're all going deaf. 

  • Like 5
Link to comment
Share on other sites

6 hours ago, colterphoto1 said:

Yeah, after shows I usually cool down with my LaScalas at a whopping 60 dbA at home. Any more I try to keep even pro PA systems around 90dbA at FOH (console) position, this is usually more than ample at the first 10 rows/dance floor. 

 

We all listen too loud. That's why the damned titanium tweeters sound good to so many of you guys - you're all going deaf. 

 

God Bless the sound man... who would do this, Michael. I pretty much quit going to live shows because the volume is way too high.

 

Bruce

  • Like 1
Link to comment
Share on other sites

Ultimately, power requirements have to reflect listening level requirements and distortion requirements.

 

Let the listening requirement be 85dB average.

Let the distortion requirement be for now, "clean"; un-clipped similar to values at rated power.

Let the crest factor requirement be +20dB.

 

Crest factor is a comparison of the peak level to average level with respect to recorded music. A crest factor of +20dB is sufficient to cover all recorded music except for some test records and demonstration albums. This +20dB is a convenient figure because it represents a linear factor of 100.

 

An amp rated at 100W subject to a +20dB crest factor results in a maximum average level of 1W. This means with the 100W amp the maximum average power level that does not exceed 100W when the +20dB crest factor is applied to the average power level is... one watt.

 

Since this 1W level is average, the crest factor above is reflected by a similar range below.

1W + 20dB = 100W

1W - 20dB = 0.01W or 10mW

 

So the 1W level is in the middle of the 10Mw - 100W range... half the music is above, half below.

At this point it is clear that even with a 100W amp, the "first watt" is half the music.

 

The listening level of 85dB is also a convenient figure because the average speaker has a corresponding sensitivity specification.

Continuing with the above instances, the 85dB speaker means the 1W average power will result in 85dB, and the +20dB crest factor applied means the level from the speaker will vary accordingly.

 

At 10mW the speaker will present 65dB

At 1W the speaker will present 85dB

At 100W the speaker will present 105dB (if it can)

 

There are some simplifying assumptions in the above, but the point is to compare this to using low power amps and high sensitivity speakers. So, we start again with the same requirements...

 

An amp rated at 1W subject to a +20dB crest factor results in a maximum average level of 0.01W. This means with the 1W amp the maximum average power level that does not exceed 1W when the +20dB crest factor is applied to the average power level is... one hundredth watt.

Since this 0.01W level is average, the crest factor above is reflected by a similar range below.
0.01W + 20dB = 1W
0.01W - 20dB = 0.0001W or 0.1mW

So the .01W level is in the middle of the 0.0001W - 1W range... half the music is above, half below.
At this point it is clear that with a 1W amp, the "first watt" is ALL the music.

The listening level of 85dB is also a convenient figure because the sensitive speaker of 105dB is +20dB above that, same as the crest factor.
Continuing with the above instances, the 105dB speaker means the 0.01W average power will result in 85dB, and the +20dB crest factor applied means the level from the speaker will vary accordingly.

At 0.0001mW the speaker will present 65dB
At 0.01W the speaker will present 85dB
At 1W the speaker will present 105dB (easily)

 

So at this point, both approaches meet the requirements at first glance. But looking further:

 

In the high power amp low sensitivity speaker combination

- class-AB has a constant level of non-harmonically related crossover distortion whose proportion of the total distortion increases with decreasing power output

- the need for global negative feedback for solid state amps to be stable is a known detriment to harmonic distortion spectrum (shifting to higher order)

- the harmonic distortion rises at lower power output levels (below the 100mW level is difficult to measure and distortion plots cut the line off below 100mW. The few instances when an effort has been made to examine distortion levels below 100mW has given a shocking result; in one case a highly regarded 250lb amp with "perfect specs" that sells for many thousands of dollars was measured to have distortion levels of 8% in the sub 100mW power output. The implication is that modest power amps may be even worse.)

- high power levels to the speaker heat the voice coils, increase their resistance, and decrease their marginal output, so resulting in compression of dynamics (so they may not make it linearly to +20dB)

 

In the low power high sensitivity combination

- class-A has no crossover distortion

- class-A SET amps typically have no global negative feedback

- class-A SET amps' distortion curves get cleaner with decreasing output levels

- low power levels to speakers are thermally insignificant to their operation (maximal dynamics)

 

Keep in mind that the SPL meters and meters on amps are not perfect. Those that show average level are generally OK in so far as the level shown is close to correct. Those that show peak levels actually are not doing so... the method for summing amplitude over time to get the canonical "peak level" was standardized for male voice early in the radio broadcasting era. The summation time is too long to represent an instantaneous music transient peak and falls short by about 13dB of true peak.

This means that when you see the biggest peaks on the meter bumping up to 87db, the actual peaks are up around 100dB.

 

 

  • Like 1
Link to comment
Share on other sites

85 db 'average' measured how, and over how long of time span? 

 

I'd probably be a good candidate for tube amps. I rarely utilize a watt I suppose. 

 

You can tell a sound man's car, it's the one with the stereo turned OFF on a Sunday morning after shows. We relish silence. As I sit here typing this on a quiet Thursday morning when you'd think I'd be enjoying my music on one of many stereos throughout the house - it's dead silent here. 

 

Yeah, I need to sell some stuff.... 

  • Like 2
Link to comment
Share on other sites

Mostly quiet here, too. Other than taking a break and listening to some YT tracks of some old Ian and Sylvia music. Don't you hate visiting someone and they have a couple of TVs going but no one is watching and it just adds to the noise?

  • Like 1
Link to comment
Share on other sites

I use an SPL meter. The average level over how long a time span is a good question. The meter itself is using a standard summation time for averaging that is about two seconds, so that average reading varies with the music... but if you watch it for a little while you can see what the "long term" average is for most music. Last night I checked and was averaging 75dB playing jazz records for a couple of hours; the loudest I ever listen is about average 80dB which I consider pretty loud, but many people consider that just the beginning. The standard level for control room monitoring in recording studios is slightly above 85dB, which means that the engineer set the tonal balance at that level.

  • Like 1
Link to comment
Share on other sites

I don't think it's been mention yet but, output voltage is one reason tube amps sound louder and more dynamic compared to a similar watts solid state amp.  Wattage is not the only factor that makes tube amps appear louder.  Distrotion is also on the side of tube amps.  Maybe someone else can explaing this better.

Link to comment
Share on other sites

18 hours ago, derrickdj1 said:

I don't think it's been mention yet but, output voltage is one reason tube amps sound louder and more dynamic compared to a similar watts solid state amp.  Wattage is not the only factor that makes tube amps appear louder.  Distrotion is also on the side of tube amps.  Maybe someone else can explaing this better.

 

A tube amp in overload conditions will increase the low order harmonic content of the reproduced music. 

 

Let's say we have a one watt SET.  We listen at a level that requires a fraction of a watt most of the time, where distortion is negligible.  But occasionally a peak will come along requiring more than one watt, and the amp will not reproduce the peak accurately, but will round it off and pepper it with low order harmonics.  Because of the brevity and particular low-order pattern, these harmonics are perceived as added loudness, much in the same way that the harmonic content of notes from musical instruments dictates their perceived loudness (J.C. Risset, "Computer Study of Trumpet Tones," Bell Telephone Labs, File MM-66-1222-2).   

Link to comment
Share on other sites

On 9/29/2016 at 4:08 PM, derrickdj1 said:

I don't think it's been mention yet but, output voltage is one reason tube amps sound louder and more dynamic compared to a similar watts solid state amp.  Wattage is not the only factor that makes tube amps appear louder.  Distrotion is also on the side of tube amps.  Maybe someone else can explaing this better.

Oh boy, lots of strange things going on here.  Ohm's Law governs wattage pretty clearly, 10 Watts from a tube amp and 10 Watts from a solid state amp into an 8 Ohm resistor is the same voltage in each case (about 9V), but Ohm's Law says nothing about distortion.  

 

There's also a sad generality that lumps all tube amps together.  I can tell you that mkane's 45 amp will sound absolutely nothing like a Dynaco Stereo 70, and that I could build you an OTL amplifier that would sound a lot like a solid state amp.

 

Distortion both is and isn't on the side of tube amps.  In a solid state amp, you generally take the signal at the input, apply an absolutely absurd amount of gain to it, then take some feedback from the output to compensate for the non-linearity inherent in most solid state devices, as well as to flatten out the bandwidth and reduce the output impedance (if necessary).  In a lot of tube amps, you do the same thing, but with the output transformer sitting in the way, you end up with stability issues that limit how much feedback you can use.  With these high feedback designs, you will tend to get less overall total distortion, but the feedback mechanism tends to create higher order harmonics which are not too pleasant to listen to.  

 

In some tube amplifiers, you can operate all the tubes in class A, and get away with no feedback.  This will result in reasonably low distortion at low power levels, and the distortion will increase as power increases with the nasty higher order bits only showing up as you clip.  In addition, for single ended amplifiers, clipping can be pretty soft, only flattening out half of the waveform, which can let you push an 8W SET amplifier way beyond 8W for brief transients that aren't super noticeable.  

 

Regardless of all of this, I'm a subscriber to the idea that you only need enough power to hit about 102dB peaks, and on a 98dB sensitive speaker, that's very little juice.  I also have seen 1000W amplifiers hooked up to high efficiency speakers, and have watched someone hot swap a cable and send the cone of a $4000 driver flying across the room.  On the CF4's that I had some years ago, 2WPC from a 45 amp would rattle the pictures on the wall without much effort.   

 

Link to comment
Share on other sites

  • 2 weeks later...
On ‎10‎/‎7‎/‎2016 at 8:36 PM, audiowize said:

 On the CF4's that I had some years ago, 2WPC from a 45 amp would rattle the pictures on the wall without much effort.   

 

 

 

5 to 10% distortion has a way of easily causing pictures to rattle off the walls...especially in the bass frequency range. 

Link to comment
Share on other sites

22 hours ago, NOSValves said:

5 to 10% distortion has a way of easily causing pictures to rattle off the walls...especially in the bass frequency range. 

The rattling of picture and other things in the house is related to their resonance frequecy and spl.  This will happen with a distorted or none distorted signal.  A speaker can rattle at high spl or with poor construction but, distorion is not the cause.  Of course a rattling speaker is a form of distortion.

Link to comment
Share on other sites

On ‎10‎/‎18‎/‎2016 at 8:30 AM, tube fanatic said:

Why so high?  It's easy to design SETs and SEPs with maximum distortion figures of 2-3% or less..............

 

Maynard

 

 IMHO that rating has nothing to do with what occurs during the wall rattle situation. In almost all cases when those pictures start rattling with a low wattage amplifiers were way past full power rating where its rated at that 2 or 3% specification .. the picture are rattling because the amplifier is being extremely over driven which means the distortion numbers are out of control. I'm not at all saying that these low watt amps can not be enjoyed but I really get sick of reading that they are capable of playing loud cleanly...it is just not true. Plenty of positives to talk about with these amps but playing loud is just not one of them.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...