• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Overkill amp madness

Graph Feppar

Active Member
Joined
Nov 19, 2018
Messages
195
Likes
154
#1
One very common belief in "high end" audio community is that the more power amplifier has,the better.I saw this being applied to both voltage and current.They say "dynamics is power so more powerfull amp = more dynamics".Bellow I will give reasons why this is not true,that its waste of money at best,degradation of signal,headphone destruction and hearing damage at worst.

1. For me,anything over 85db average SPL or 95db peak is uncomfortably loud with modern music.People rarely reach 100db + even in transients,generally average listening loudness is around 80db.Audiophools,except those with significant hearing damage or old tend to over estimate how much loudness they require.
2. Impedance of headphone have nothing to do with how hard is it to drive,its all about efficiency.
3. Even cheapest portable player and tiny amp + dac combinations have atleast 2 V rms output.
4. High end portable headphone players have over 6 V rms output.
5. Balanced desktop headphone amps like Violectric V281 have 35 V rms output.

6. Relatively high voltage requiring Sennheiser HD800 needs 250 mV for 90db SPL,they have 360ohm impedance and use 0.16 miliwatts to reach 90db.This large Sennheiser flagship is more power hungry than 99% of all headphones.

8. Most powerhungry headphone in the world,the infamous Hifiman HE6 needs 1 volt for 90db,with its 50 ohm impedance,it draws 20 miliamperes for combined power of 20 miliwatts.

9. Amplifiers are trade off between voltage and current capability,for given size,weight,cost and power,the higher the max voltage ( Gain),the lower max current becomes,the higher the output impedance becomes which causes decreased damping (bad) and increased noise.

Furthemore this excessive gain causes the need to strongly attenuate the signal either digitaly in DAC which result in more increased noise or with analog resistive pot which wastes power and if its not exotic high precision matched resistor stepped attenuator like the DACT or the one in Benchmark HPA4,its going to cause channel imbalance.The higher the amp voltage,the stronger the attenuation must be,the larger the imbalance gets.

Ideal amplifier should have just enough voltage to drive headphone to realistic listening level,voltage any higher than that degrades signal.The audiophile mega volt amps have capability to instantly destroy headphones,HD650 will get damaged past 8 volt rms,furthermore 35 volts even through most inneficient headphone in the world,the HE6,will make them play at 120 db SPL which will cause instant permanent hearing damge.

Increasing current capability is better,higher current capable amplifier will produce lower distortion and have better damping factor due to lower output impedance.But there exist other way to lower output impedance and decrease distortion... the feedback.That means big heavy desktop amps are not needed.

To give example,portable player iBasso DX200 have 0.28ohm output impedance.Wallet sized desktop DAC Pro Ject S2 Pre Box which have headphone output with two ESS ES9603 headphone amplifier ICs have 0.1 ohm impedance,I believe same IC is used in certain smartphone like LG V.Chord Mojo,the tiny portable ampDac have 0.2 ohm.

People really need only half a volt and couple miliamperes of current,even 2 volt will drive HE6 to uncomfortably loud levels,2 volts will drive relatively high voltage requiring HD600/650 to 110db which is pain threshold at ear resonance frequency of 3 KHz.Most headphones would reach or approach 120db,at 2 volts.

It is completly irrelevant if amplifier have capability to spit out over 9000 jigga watts of juice,what matters is its THD+N at realistic listening volume which is couple of miliwatts for 99% of people.
 
Last edited:

cjfrbw

Senior Member
Joined
Mar 5, 2018
Messages
310
Likes
278
#2
I would imagine you could use one of those signal time/voltage diagrams to identify peaks and decide what transient wattage was minimally necessary for a given dB level from a particular set of speakers in a particular room at a particular seat.

I don't think you will ever cure male audiophiles from the notion that raw watts from bestial amps are better( just because), any more than you will take away their guns or monster trucks. It's a testosterone thing. Sure, transients can sometimes eat up power, but can you really distinguish that during loud passages with psychoacoustic masking?

It is a valid notion that having lots of overhead means that you are at lower risk of clipping and messing up your drivers. Also, hungry bass signals are less likely cause the higher frequencies to distort. There are drivers that just sound better with lots of current. Complex impedances are also an issue, but you can identify that as a speaker specific issue. I use ribbons which have a flat impedance.
 

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,595
Likes
3,005
Location
Central Fl
#3
I don't think you will ever cure male audiophiles from the notion that raw watts from bestial amps are better( just because), any more than you will take away their guns or monster trucks. It's a testosterone thing.
Spoken like a good little snowflake.
 

cjfrbw

Senior Member
Joined
Mar 5, 2018
Messages
310
Likes
278
#4
I think Amir has a testosterone distortion compensating circuit to adjust his tests.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,018
Likes
3,276
Location
Riverview, FL
#5
For me,anything over 85db average SPL or 95db peak is uncomfortably loud with modern music.
Discounting squashed sources...

You might be a little shy on the peak, add 15 to 20dB to the average, typically...

I have 700 (mostly unused) four ohm watts (claimed) on tap. Might get it for a moment, but the breaker pops with extended over-exuberance, which has happened like once in six years, so not concerned about it)..

Maximum level seen (for a moment) was a very brief calibrated drum solo replay experiment, at 116.9dB peak. Sounded good but also like someone taking a baseball bat to the speakers.

85dB average (C-weighted below) is not uncomfortably loud. Still listening without distress 40 minutes later.

Ambient temperature 76F (windows open, light rain), amplifier heat sinks 115F (left) and 113F (right).

Guitar Bass Drums Trio - Kazumi Watanabe - Tricoroll

1551207588213.png
 
Last edited:

JJB70

Major Contributor
Patreon Donor
Joined
Aug 17, 2018
Messages
1,312
Likes
2,374
Location
Milton Keynes, England
#6
You know when an amplifier has enough power when you need a high voltage feeder and a dedicated generator. Anything less is for wuss's.
 

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,927
Likes
3,391
Location
Monument, CO
#7
For music an AES study ages ago showed the peak-to-average SPL level was about 17 dB, a power factor of fifty. That is, if you average 1 W, then a 50 W amplifier should suffice. Scale accordingly.

More recently I have read 20 to as much as 30 dB (100x to 1000x) peak-to-average in movies. I'd guess the highest peaks are due to explosions and gunshots and stuff that wouldn't sound much different clipped or not so I decided 1000x was overkill (though in practice I have that much most of the time -- see below).

The THX/Dolby/whatever standard is 105 dB peaks per speaker and 85 dB average, i.e. 20 dB (100x power) headroom over average, and that seems to tie in pretty well. My normal average level is lower than 85 dB and I targeted my original system (with less-sensitive speakers than I have now) to handle 105 dB peaks so I should be OK. At 75 dB average I have 30 dB of headroom, and I tend to run closer to 70 dB average'ish.
 

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,595
Likes
3,005
Location
Central Fl
#8
I have 700 (mostly unused) four ohm watts (claimed) on tap. Might get it for a moment, but the breaker pops with extended over-exuberance, which has happened like once in six years, so not concerned about it)..
You need a bigger service!
You got welding cable sized speaker leads and a 15 amp service? :p
I've got a 20 amp service with 7 A/B power amps and 2 Class C sub amps plus a 75" Sony on that line and never popped the breaker while cracking plaster. LOL
 
Joined
Jan 11, 2019
Messages
79
Likes
76
#9
Contrary to the above, I would guess that most audiophiles grossly underestimate the amount of power that they need to properly drive speakers on transient peaks if they listen to either symphonic or uncompressed rock/pop music at "realistic" volume levels. I don't do that all the time, but I do like to, and plan accordingly.

An amplifier with properly implemented clip lights will quickly show that you need large amounts of power to listen to "big" music without clipping. Let's say we want to average, oh, 90dB on a symphonic piece. There will likely be transient peaks in that piece of music which hit +25dB over the mean, for about 115dB peak. Now, assume that the speakers are 90db/1W/1m. Put the listening position out at 2m. The inverse square law is not our friend. 1W will provide all of 84dB at 2M. The second speaker will increase this somewhat, but only if the signals are correlated. So we'll assume we get nothing out of the second speaker since a transient peak will not necessarily be shared equally by both channels. Thus, starting from 84dB, it takes us 100W to get to only 104dB. 107dB will require 200W. 110dB, 400W. You get the idea. Now, there will be some boundary reinforcement or speaker coupling that might provide another, say, 6dB, maximum.

Various measurements have been made of front row symphony seats having 115dB to 125dB peaks. As you can see, either highly efficient speakers or prodigious amounts of power are required, preferably both.

As for that nonsense about Hifiman headphones. Well, I have them. They are awful on 2V, if I want to entertain myself for a bit. They are quite happy to slurp up all 6V my desktop amplifier will give, and could use more. Unfortunately, once you're at that sort of power level already, there isn't all that much more to get.

Nice theories you have, and all, but quite wrong unless your listening choices consist of wimpy chamber music and acoustic guitar.
 

cjfrbw

Senior Member
Joined
Mar 5, 2018
Messages
310
Likes
278
#10
Looking at my own system, with active crossovers and surrounds, there is a bit of power there. Four subwoofs have about 1Kwatt total. Main speakers 80-350 Hz band 140 watt continuous (2db transient overhead)per channel, 350 Hz-7khz band 125 watt continuous (2db transient overhead) per channel, 7khz-infinity band 3 watts per channel (100db efficient tweeters). Surround speakers x 6 have 110 watts apiece nominally through Yamaha RX 9.

However, I can use just a crossover-less panel (88 db efficient??) full range with a 25 watt class A push pull amp (likely 80 watt transient capability into class AB) with no subwoofs or surrounds and run it louder than I can stand with no audible distortion. It won't thump my gut, though, and I like gut thumping.
 
Last edited:

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
2,162
Likes
3,484
Location
The Neverlands
#11
1. For me,anything over 85db average SPL or 95db peak is uncomfortably loud with modern music.People rarely reach 100db + even in transients,generally average listening loudness is around 80db.Audiophools,except those with significant hearing damage or old tend to over estimate how much loudness they require.
My findings are here.

2. Impedance of headphone have nothing to do with how hard is it to drive,its all about efficiency.
Not entirely true. Impedance plays a large role.
A good way to show this is the tube amp Amir measured.
These amps cannot drive low impedance headphones.
A phone cannot drive a 300 Ohm headphone properly (as in sufficient SPL levels).
You can compare the dB/mW levels of a lot of full sized headphones here and may find they usually are not that far apart.

3. Even cheapest portable player and tiny amp + dac combinations have atleast 2 V rms output.
most phones and cheap MP3 players often only reach just over 1V.
The reason for this is the power supply voltage which is around 3.5V this means that even when the output is rail-to-rail you can just reach 1V

4. High end portable headphone players have over 6 V rms output.
5. Balanced desktop headphone amps like Violectric V281 have 35 V rms output.
What is the point here ? there are also amps reaching 3V or less and ones with 10V or 12V output voltages.
The question in most cases is how high the output current is.
An amp that can put out 18V in 600 Ohm may only put out 1.6V at 32 Ohm for instance if current limited to 50mA.

6. Relatively high voltage requiring Sennheiser HD800 needs 250 mV for 90db SPL,they have 360ohm impedance and use 0.16 miliwatts to reach 90db.This large Sennheiser flagship is more power hungry than 99% of all headphones.
Nope it is 98dB/mW which is what most headphones are. Even the HD598 has a similar power rating. Still from an amplifier the HD598 is playing a lot louder than the HD800 (or HD650 which has the same efficiency and impedance) simply because the dB/V ratting differs due to the impedance.


8. Most powerhungry headphone in the world,the infamous Hifiman HE6 needs 1 volt for 90db,with its 50 ohm impedance,it draws 20 miliamperes for combined power of 20 miliwatts.
When you play a DR15 recording at 90dB peak average power levels will be around 75dB. Not sufficient for analytic listening.

9. Amplifiers are trade off between voltage and current capability,for given size,weight,cost and power,the higher the max voltage ( Gain),the lower max current becomes,the higher the output impedance becomes which causes decreased damping (bad) and increased noise.
Max. output voltage is determined by the voltage rails in the amp and the circuit. For tube amps things differ a bit.
One can make an amplifier with a VERY low output impedance yet have a very low current capability.
Max. output current is limited by design (and power supply) not by the output impedance.


Furthemore this excessive gain causes the need to strongly attenuate the signal either digitaly in DAC which result in more increased noise or with analog resistive pot which wastes power and if its not exotic high precision matched resistor stepped attenuator like the DACT or the one in Benchmark HPA4,its going to cause channel imbalance.The higher the amp voltage,the stronger the attenuation must be,the larger the imbalance gets.
An analog pot does not waste power. Excessive gain means usually a bit higher distortion and noise levels and little usable volpot range.
What is true is that analog volpots (the cheaper ones) have a tendency to have channel imbalance at lower volpot settings.
This means the wrong amp is chosen, or the wrong DAC or the wrong headphone.
There are ways around this.

Ideal amplifier should have just enough voltage to drive headphone to realistic listening level,voltage any higher than that degrades signal.
Nonsense.

The audiophile mega volt amps have capability to instantly destroy headphones,HD650 will get damaged past 8 volt rms,furthermore 35 volts even through most inneficient headphone in the world,the HE6,will make them play at 120 db SPL which will cause instant permanent hearing damge.
It will only be destroyed when used with the headphone not being on the head. When the sound becomes so loud you can destroy it you have dialed back.
The 35V amp will not put out that voltage on a 50 Ohm load. It will be lower.

Increasing current capability is better,higher current capable amplifier will produce lower distortion
These things are not related at all.

People really need only half a volt and couple miliamperes of current,even 2 volt will drive HE6 to uncomfortably loud levels,2 volts will drive relatively high voltage requiring HD600/650 to 110db which is pain threshold at ear resonance frequency of 3 KHz.Most headphones would reach or approach 120db,at 2 volts.
Nope in many cases you need more than 0.5V and a couple of mA's.
2V will allow peaks of 110dB for HD650/HD800 but this is accumulated signals. The 3kHz part is usually 35dB down opposite the peak SPL.
So around the 3kHz band peak SPL would be around 75dB. Average would be lower.
You have to realize that the peak SPL an amp can deliver is never a single sinewave but can be an addition of several instruments that happen to have its peaks in the same moment of time. An amp needs to allow for this otherwise it clips.
Also one needs to take the equal loudness curve into account which means a lot of amplitude is taken up by the lows which we don't perceive that loud.

It is completly irrelevant if amplifier have capability to spit out over 9000 jigga watts of juice,what matters is its THD+N at realistic listening volume which is couple of miliwatts for 99% of people.
Indeed THD+N at realistic listening volume is important.
Even more important to have this even at higher output levels.
 
Last edited:

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,927
Likes
3,391
Location
Monument, CO
#12
Nice theories you have, and all, but quite wrong unless your listening choices consist of wimpy chamber music and acoustic guitar.
Actually data from the AES, not my theories. I know how loud an orchestra is, from the back row even, but do not want it that loudly in my home. And when I am in the audience I do not like the front row -- too close to see everyone on stage, and too durn loud for me, plus the balance is rarely as good as back in the middle (where the sound board is usually set, not by coincidence).

But thank you for denigrating some of my favorite music; I don't listen to big testosteronic pieces all the time. If that makes me a wimp so be it. :)

This place seems awfully snarky lately, need to stay away a bit.
 
Last edited:

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,018
Likes
3,276
Location
Riverview, FL
#13
This place seems awfully snarky lately, need to stay away a bit.
Well, just try not to contribute to it (not saying you have).

You got welding cable sized speaker leads and a 15 amp service?
Yup.

It's enough.

Until I run the kitchen overhead lights (320W) and the vacuum cleaner (750W) and the TV (couple of hundred) and the PC (another couple of hundred) AND forget to turn off the Krells at the same time. Then it pops. Yes, it's all on the same breaker. I hate houses.

I have 30 dB of headroom, and I tend to run closer to 70 dB average'ish.
I didn't stop the meter after the 85dB display earlier. Some music, some radio, and now some TV, and the meter says 5 hours 56 minutes and 70.2dB Leq C weighted.

1551228559106.png


Don't believe the peak, I suppose the mic got bumped.
 
Last edited:

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,595
Likes
3,005
Location
Central Fl
#14

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,595
Likes
3,005
Location
Central Fl
#15
Yup.
It's enough.
Until I run the kitchen overhead lights (320W) and the vacuum cleaner (750W) and the TV (couple of hundred) and the PC (another couple of hundred) AND forget to turn off the Krells at the same time. Then it pops. Yes, it's all on the same breaker. I hate houses.
But your the Scrutinizer,
Run over to Lowe's, get some 12g or fatter wire, breakers, boxes, etc; and run a couple new circuits.
Easy Peasy
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,091
Likes
23,278
Location
Seattle Area
#16
Let's dial down the emotions and personal remarks please. There is no award for any winner in these arguments other than respect. Make sure that is earned!
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
7,018
Likes
3,276
Location
Riverview, FL
#17
Run over to Lowe's, get some 12g or fatter wire, breakers, boxes, etc; and run a couple new circuits.
It's all been here for a year, maybe more.

*takes a look

Oh, two years ago, March 15, 2017
 
Last edited:

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,595
Likes
3,005
Location
Central Fl
#18
It's all been here for a year, maybe more.

*takes a look

Oh, two years ago, March 17, 2017
Never do today what you can as well do tomorrow,
something may occur to make you regret your premature action. ;)
 

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
2,255
Likes
3,479
Location
Gold Coast, Queensland, Australia
#19
But thank you for denigrating some of my favorite music; I don't listen to big testosteronic pieces all the time. If that makes me a wimp so be it. :)
I listen to all types of music, chamber music and acoustic guitar included and it doesn't make you remotely wimp-like to do so.

This place seems awfully snarky lately, need to stay away a bit.
Clearly I underestimated the effort Thomas put in to keep the peace. He is greatly missed.[/QUOTE]
 
Thread starter Similar threads Forum Replies Date
W Fun topics, videos, etc. 19
svart-hvitt Amplifiers, Headphone Amps, Phono preamp, etc. 62

Similar threads

Top Bottom