Dynamic Headroom: Where is the limit?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
We all know that sustained output power is directly proportional to an amps size but is it possible to build amps with smaller output power (ie:10W instead of 100W) and make up the difference with peak headroom? How much peak headroom is possible and what is the best way to get it? To give some commercial examples of what I am talking about, take the typical NAD amps. They describe +6dB (2 to the second power above the rated power) of "headroom" in their literature. How do they achieve this? What would need to be done to an 8W Zen-X to make +21dB (2 to the 7th power above rated power) of dynamic headroom on it? Bigger, higher power rated amps sound better because they have the power in reserve when large dynamic range is required but can we achieve the same with a smaller output stage? After all, who wants to go through the trouble of matching large mosfet sets if we can get the result with the power supply alone?

Mr. Pass, is there a practical limit to "dynamic headroom" on a small output stage and if so, what is it?
 
Truth In Advertising.

Hello Nania,
The NAD spec of having a 6 dB headroom is the marketing and sales department's way of saying that the amp has a lousy power transformer that sags badly, and low power rated output transistors.
Big and solid amplifiers quote 0 dB headroom.
Having a droopy PSU causes more problems than it solves and is bad engineering.
Class A amps are by definition not suitable.

Regards, Eric.
 
nania

When I use the thinking "dynamic headroom"
I usually think in recordings.
A recording can be compressed or dynamic.
A dynamic recording doesn't cut the tops, peaks
of recording.
That means that a dynamic recording have to have
lower average dB output.
The standard for CD output is set to +-2.8 volt peak.
(for other sources it can be different)
Can as a consekvens require some more voltage gain,
not to be to quiet when you listen.

Also I associate to the working-space of the voltage-gain-stage
in my amplifiers.
good working space=supply voltage,
provides smaller procentual changes to handle
for the transistor.
If I use that term you refer to correctly,
that is a matter of language.

But to be sure that we are "on talking terms"
I explain what I mean.
 
Ex-Moderator
Joined 2002
quote:
--------------------------------------------------------------------------------
The NAD spec of having a 6 dB headroom is the marketing and sales department's way of saying that the amp has a lousy power transformer that sags badly, and low power rated output transistors.
--------------------------------------------------------------------------------


Eric is basically right, although he has a "glass half empty" way of putting it. The NAD transformers and transistors are fine for delivering the continuous rated power. You have the added bonus of being able to operate in the "red zone" for a short time which give you the headroom to keep away from that rude transistor clipping for a few extra db. It has probably saved a tweeter or two somewhere.


quote:
--------------------------------------------------------------------------------
Big and solid amplifiers quote 0 dB headroom.
--------------------------------------------------------------------------------


Right again, but you pay for the "big and solid". The whole idea behind "dynamic headroom" is economy. "Get the power of a big amp at the price of a medium size amp". It's not very fair to compare the performance of a multi-thousand buck amp with that of a sub-thousand buck amp. Although the NAD's (at least the 80's versions) will hold their own against more then a couple of the high dollar amps around even today.


quote:
--------------------------------------------------------------------------------
Having a droopy PSU causes more problems than it solves and is bad engineering.
--------------------------------------------------------------------------------


Now here I have to kindly disagree. The NAD amps were "engineered" as a low to medium priced amp that could attract some hi-end support. They have done just that. The NAD's are an excellent example of good solid engineering.

We all love to have the super hi-end, pull out all stops amp that double as a welder, and you can buff up with moving it around to just the right spot. But there aren't many that can afford it. My wife was happy with her NAD for years. At least until she heard my Pro Bryston 4B. Now its her Bryston. Oh well, I went an got a 4BST.

Rodd Yamashita
 
Ex-Moderator
Joined 2002
fcel,

My wife is into the music. Although, she is almost the perfect equipment tester. She hears with her ears, not with her brain. I had 3 amps that I had an opinion on, and wanted to qualify what I thought them. They were an AR D52, the Pro Bryston 4B, and an NAD 2600. She listened to each off and on for a few days, and she would tell me what she thought of each amp. So when we were done I put the NAD in her system, and the Bryston and AR went in my system. In the mean time, I pick up a Bryston 4BST to drive the mid bass in my tri-amp system. Thats when the brow-beating started. To make a long story, she got the Pro Bryston, and I got the NAD for my subs. It takes the SAF to a new level.

Rodd Yamashita
 
roddyama said:
quote:
--------------------------------------------------------------------------------
The NAD spec of having a 6 dB headroom is the marketing and sales department's way of saying that the amp has a lousy power transformer that sags badly, and low power rated output transistors.
--------------------------------------------------------------------------------


Eric is basically right, although he has a "glass half empty" way of putting it. The NAD transformers and transistors are fine for delivering the continuous rated power. You have the added bonus of being able to operate in the "red zone" for a short time which give you the headroom to keep away from that rude transistor clipping for a few extra db. It has probably saved a tweeter or two somewhere.


quote:
--------------------------------------------------------------------------------
Big and solid amplifiers quote 0 dB headroom.
--------------------------------------------------------------------------------


Right again, but you pay for the "big and solid". The whole idea behind "dynamic headroom" is economy. "Get the power of a big amp at the price of a medium size amp". It's not very fair to compare the performance of a multi-thousand buck amp with that of a sub-thousand buck amp. Although the NAD's (at least the 80's versions) will hold their own against more then a couple of the high dollar amps around even today.


quote:
--------------------------------------------------------------------------------
Having a droopy PSU causes more problems than it solves and is bad engineering.
--------------------------------------------------------------------------------


Now here I have to kindly disagree. The NAD amps were "engineered" as a low to medium priced amp that could attract some hi-end support. They have done just that. The NAD's are an excellent example of good solid engineering.

We all love to have the super hi-end, pull out all stops amp that double as a welder, and you can buff up with moving it around to just the right spot. But there aren't many that can afford it. My wife was happy with her NAD for years. At least until she heard my Pro Bryston 4B. Now its her Bryston. Oh well, I went an got a 4BST.

Rodd Yamashita

I must agree with Eric on all 3 points ... a well engineered product does not have rails that collapse as badly as is the case with these and most other Class AB amps. Under-rated power supplies give rise to higher distortion amongst a number of other things.. Just because it is adequate for management and PR does not mean it is good engineering practice... The same goes for pcb mount RCA connectors ... they save a few pennies but sacrifice the structural integrity of the connection.

Also, to address the initial question .. Dynamic Headroom is not applicable to single ended class a amps because the output stage must always carry the full bias current .. The exception to this is if you replace the upper bias resistors with Aleph current sources. You'll gain higher efficiency so for a given maximum output you wont have to dissipate quite as much and you'll get greater voltage swing but as said already this is still Zero Dynamic Headroom... Of course it wont sound the same but that's what you get. After that if you still want more you would have to replace the lower resistors with a normal current source. But at this stage you really dont have much left of the original soz.
 
I Hardwire

Whenever it is possible.
Reason,
half the number of soldier-joints.
Reduces number of "wires".
Gives Shorter signal travellin'.
Air is a good separator isolator.
A Headphone amp I´ve designed built, 2 stage, 12V supply,
Have 15-20 mm signal travel.

Saves me the work of makin'
what you call PCB.
Not long ago I wondered what that was??...
And why you used this thing?

I am no monkee
but monkees see....

gro
 
So it seems "dynamic headroom" is a fictitious term and all that is necessary to accommodate the dynamism of some music is to have the ability to reproduce the required level at the speaker. If a speaker can produce 80dB with 2.83V at 1 meter, it will have enough to do the job with the amplifier at 0.5W providing the amp can deliver the signal in phase and with the required immediacy. Do we all concur on this?
 
I never use speaking elements

under 90 dB
You have to apply to much energy.

look for 92-96dB speakers

And I use never over 10W class A
As it is active I have no 1-3 dB losses in
those garbage they damp speakers
with L C

If you have bad hearing, most audiophiles have?,
you need more energy
to listen
And yet they hear better then us?
Can't figure out how?
;) ;) ;) ;)
 
AX tech editor
Joined 2002
Paid Member
Dynamic headroom etc

Hi Guys,

I have a problem with some of the above statements that a sagging PSU has no place in good engineering, or words to that effect. Tha NAD is engineered for a certain sustained power. It's open in the specs. Yes, it may have a weak xformer that gives out higher voltages when loaded lightly, and thus gives higher short-term power output called high dynamic range.

I think this makes a lot of sense. Try measuring the average output level of your super no-holds-barred amp in an average listening session. I would be surprised if it topped 5 Watts. So, really, that 200W monster is bad engineering! That doesn't mean that I don't like these monsters, but for different reasons, not because they are well engineered for the task, because they are not.

I like Nelson's amps, they are beautiful in their elegant simplicity, almost a work of art. But they are not "well engineered" in the original sense of a product well adapted to what it is supposed to do. There, the NAD would win hands down.

Cheers, Jan Didden
 
audiofreak wrote
Just because it is adequate for management and PR does not mean it is good engineering practice... The same goes for pcb mount RCA connectors ... they save a few pennies but sacrifice the structural integrity of the connection.

There is a diffrence between a well engineered product and good engineering practice:eek:
Thats what makes the price diffrence between mid-fi and hi-fi:p

If i asked you to go and design a mid-fi surround amp to be mass produced and you came back with a design with about 30 RCA connectors all connected to the PCB by wires i would slap you about the head with it (including those overly expensive transformers), bwahaha:p

About the dynamic headroom thing, IMHO i don't think this term can be applied to an amp, sounds like specmanship.
 
Interesting thread; power suppies and their sonic influences are not well understood despite years of work in SS and tube design. Truth is, the electrical environment of all active elements in audio is completely dictated by the power supply.

I have found that the power supply is probably at least one half of the sonic influence on an amplifier. This applies to SS and tubes. The other half is the amplifying stage itself; topology, operating point, layout and passives.

The degree and profile of the voltage sag is a huge factor in the sound. If you regulate, and the choice of series or shunt is very important, then the sonic characteristics change completely. If you do not regulate, then the nature of the inductance, capacitance and resistance of the power supply is the primary determinant of sound.

This all applies at even tiny, low levels, and as much to preamplifiers as to power amplifiers. I should add that with proper regulation, and working at line level, it is possible to make a tube stage sound almost indistinguishable from Solid State. This phenomenon, in my view (and nothing is more dangerous than one man's opinion, egads, it might be quite wrong!!) makes mincemeat of the perennial tube/SS debate.

Even using the same operating point, it is possible to manipulate the sound to your liking by tinkering with the power supply. The effect is so pronounced that it makes you scratch your head about whether the choice of tube is really all that important. This is, I believe, the reason a lot of very good amplifiers use obscure tubes, such as the 6CK4, the 6BM8, the 6AS7, the SV811, and the PL509. It is essentially the power supply which makes them deliver the goods, providing the operating point is judiciously chosen. 'Voicing' the amplifier consists in both tinkering with the tube operating point but also getting the power supply dimensioning just right. This takes a lot of time, and is completely empirical. PSpice is useful however for identifying the range extremes, and should never be dismissed.

Generally, solid state is just as susceptible to this sort of tuning. Recently, during development of a preamplifier pulling about 15mA from each 15V rail, we discovered that the sound was hugely influenced by the choice of series or shunt regulation, and finally settled on an emitter follower to supply the rail, using a simple, neanderthal zener as the voltage reference. This gave the best balance of tonality, imaging and engagement and easily beat a TL431 TI shunt element, a straight zener, and a resistive feed. Clearly neither voltage accuracy nor control is pivotal to best sound, something which surprised us.

Power supply, and the nature of the 'sag' Janneman refers to, is crucial to the sonics, and not well understood.

Cheers,

Hugh

www.printedelectronics.com
 
POWER MUST BE delivered

So I simply desides what peak voltage
I would want over my special speaker (, what power is that?
i usually count for half speaker nominal impedance!)
which is going to be adopted to output of the program material
I want to use.
For Cds with low average I put more voltage gain in
to be safe.
To get the trafo dimensioned
I take class A values for peak output over
mentioned load value (1/2*nominal)

As there is some recalculation of AC to DC
there are koefficients I use.
This is to find in basic PSU knowledge.
The type of rectification selects for different koefficients.

Then I get my peak energy output translated to TRAFO spec.
(see TRUTH in advertising)
to be safe from manufatures false spec of trafo
and Mains variations, sag
I have to add some 20-50%, always take next higher spec trafo
Then I buy TRAFO

My way to design confirm the basic message of AKSA
who also DESIGNS and sells good amplifiers
If they are better than my own,
that should be a matter of taste.

What AKSA says of the sonic/non-sonics of power supply
is best explained as

if power have signal
it is the mixture of 2 signals
that are delivered to speakers

a photographer DO NOT use anything
but paper without graphics,
usually totally white,
to make a picture out of his negatives
I have done that, even one hair is annoying
when you magnify it

regards AKSA
gro
 
Ex-Moderator
Joined 2002
"Dynamic Heedroom" >>> "Dynamic Range"

So it seems "dynamic headroom" is a fictitious term and all that is necessary to accommodate the dynamism of some music is to have the ability to reproduce the required level at the speaker. If a speaker can produce 80dB with 2.83V at 1 meter, it will have enough to do the job with the amplifier at 0.5W providing the amp can deliver the signal in phase and with the required immediacy. Do we all concur on this?
"Dynamic Headroom" I wouldn't say "fictitious", although it is a marketing term, and there are those manufacturers that will use it to "hide" short comings in their products. But there are also many manufacturers that have tried to "engineer" a reasonably priced amp that can provide some extra power for short periods. NAD, Apt, Hitachi, Technics, Soundcraftsmen, and others have all used various amplifier designs, with various levels of success, to provide more power at a reasonable price.

"Dynamic Range" This is the real issue. The selection of an amp is not the first item on the list of "How do I reproduce the dynamics in music". You must first have the speakers that can reproduce the dynamics and all the other characteristic necessary for hi-fidelity. The efficiency of those speakers will dictate how much power you need to reproduce the full dynamics of the music. Then it's a mater of how much do you want to spend.

Rodd Yamashita
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.