whoaru99
Epic Member
Not really. Peak testing is pretty short duration. Of course, it somewhat depends on the load. If your load is a short circuit, well, that may be a different story.
Power Amp Output Questions
- Thread starter Tinnkindling
- Start date
Not really. Peak testing is pretty short duration. Of course, it somewhat depends on the load. If your load is a short circuit, well, that may be a different story.
JoeESP9
ESL's & tubes since 83
There's that continuous power. On this we definitely agree.
Well, he also says "ultimate dynamic ability" - leading me to think in absolutes. I 'think' I understand about short term peak testing.
Thanks
Superampman
AK Member
Wow. I didn't know I had that much power.
whoaru99
Epic Member
Yes, and in the vast majority of cases that continuous power is rated somewhat below clipping, but not 50% below clipping.
whoaru99
Epic Member
Think in absolutes all you want but that's silly in this regard. What's the purpose of testing to the point if you destroy the amp? I can see that, maybe, in development but in terms of practical use effectively useless information because it means destroying the amp to achieve it.
It's kind of like those high current specs you see some amp mfg spec. Who cares if the amp can produce 100A peak current into effectively a short circuit? It means nothing from a practical sense except a number that looks good on paper.
Last edited:
Superampman
AK Member
So I take it you consider headroom specs to be useless? I know that couldn't be true. Would you consider it when listening to classical with 30db or more dynamic range? And I think we better hold on to our hats for what the future holds in this respect.
I'm sure whoaru will answer, but from my perspective I would prefer to get my required headroom from within the amp's continuous output capability (or very close). Headroom specs are not standardized, so it's hard to know how they have been calculated, the level of distortion when pushed into it's "headroom reserve", and for what period of time it can deliver this short term "power boost".
Last edited:
I remembered that you pursued this issue in a past thread in which I was also participating, so I dragged out this exchange between us:
You:
What about dynamic headroom of the amp? 2-3db is fairly typical. My 100wpch Meridian 105s produce 700 watt peaks or 7db headroom each
Me:
I've not found 2-3dB to be typical of the amps I've owned. I know that some manufacturers use unique designs to deliver higher peak power, while others focus on building stiff power supplies with little to no difference between continuous and peak power. There are many pros and cons in this complex topic, and it should be noted that - at least in my understanding - there are no standardized methods for defining peak power and associated distortion (which may increase dramatically; for example, I've seen AVR's with peak power ratings accompanied by 10% THD figures).
Except in the case of manufacturers with strong reputations and a well-respected design specificaly intended to provide dynamic headroom with a well-regulated power supply, I generally assume dynamic headroom as zero (which then provides some amount of insurance, or buffer, to help minimize or avoid clipping).
You:
But my experience dictates the kind of peaks I've been listening to far exceed the rated specs usually given percentage wise. I would think max power given ratings without headroom is a bit misleading and potentially hazardous. After all, music is basically made up of peaks in both directions larger and smaller. A preamp's gain control is typically set so the average amp is driven to full output at around half volume give or take. So if headroom isn't built into the amp, it will usually be over driven which I find totally unacceptable and therefore not the case at all. Not because I happen to find that unacceptable but that the market itself appears to concur.
Me:
Headroom for peaks may be available as a function of design, or more typically as a function of not driving the amp at its full output.
You:
Sure, and it's your latter statement that makes for poor market sense. Which is why I can't accept it as being the case in general. There are limits to everything but an amp with zero headroom is just a poor design imo. Either that or one with over stated specs. For example max power including headroom. In which case it would still max out roughly at half on the pre dial. If it max's out at way less but with similar input sensitivity, then it's inferior imo.
What about dynamic headroom of the amp? 2-3db is fairly typical. My 100wpch Meridian 105s produce 700 watt peaks or 7db headroom each
Me:
I've not found 2-3dB to be typical of the amps I've owned. I know that some manufacturers use unique designs to deliver higher peak power, while others focus on building stiff power supplies with little to no difference between continuous and peak power. There are many pros and cons in this complex topic, and it should be noted that - at least in my understanding - there are no standardized methods for defining peak power and associated distortion (which may increase dramatically; for example, I've seen AVR's with peak power ratings accompanied by 10% THD figures).
Except in the case of manufacturers with strong reputations and a well-respected design specificaly intended to provide dynamic headroom with a well-regulated power supply, I generally assume dynamic headroom as zero (which then provides some amount of insurance, or buffer, to help minimize or avoid clipping).
You:
But my experience dictates the kind of peaks I've been listening to far exceed the rated specs usually given percentage wise. I would think max power given ratings without headroom is a bit misleading and potentially hazardous. After all, music is basically made up of peaks in both directions larger and smaller. A preamp's gain control is typically set so the average amp is driven to full output at around half volume give or take. So if headroom isn't built into the amp, it will usually be over driven which I find totally unacceptable and therefore not the case at all. Not because I happen to find that unacceptable but that the market itself appears to concur.
Me:
Headroom for peaks may be available as a function of design, or more typically as a function of not driving the amp at its full output.
You:
Sure, and it's your latter statement that makes for poor market sense. Which is why I can't accept it as being the case in general. There are limits to everything but an amp with zero headroom is just a poor design imo. Either that or one with over stated specs. For example max power including headroom. In which case it would still max out roughly at half on the pre dial. If it max's out at way less but with similar input sensitivity, then it's inferior imo.
My comments are still a good representation of my thoughts on the subject. Do you feel the same about yours?
I thought that if we are going to go down this path again, it would be helpful to other readers/posters to get a "running start".
Superampman
AK Member
Anything useful would be for the duration of a typical transient. Not sure what you mean by 'how'. 3db headroom is a doubling of power. As far as distortion goes I imagine a lot of the techs can shed some light on it as pretty well all amps are designed with it, are they not?
Superampman
AK Member
Sorry, I left mid-way to eat supper. I'd forgotten about that conversation. Looks like my point of view is indelibly burned into my brain.
Because 'headroom' depends on the duration of what it is provided for, such a characteristic as a spec. can be vague/misleading at best. What I feel comfortable with is the continuous or sine input power rating, and perhaps a peak value of output, although the latter again could be open to conjecture as said.
Pardon if repetition (have not now read through the lot again), but it should be clear that this is mainly a power supply thing. Mostly continuous power drops the rail voltage to its lowest value, which in any proper design should be above 70,7% of the no-load value. That means that if a continuous - peak rating differs by more than 3 dB the design begins to be questionable. From that alone one can conclude that available intermittent peak power would be no more than double the continuous value or 3 db up, which does not represent a significant increase in volume - it is sometimes called 'just audible'.
Any 'headroom' concept is too linked to the type of music to be part of a specification. What is normal, what kind of headroom is required by a specific kind of music? For a classic symphony that might amount to all of 20 dB or more, for modern popular music (with the kind of recorded over-amplitude I read about on CDs) it might not be much. It is a matter of experience/taste, and folks will need to estimate what will typically be required in the case of each person.
Pardon if repetition (have not now read through the lot again), but it should be clear that this is mainly a power supply thing. Mostly continuous power drops the rail voltage to its lowest value, which in any proper design should be above 70,7% of the no-load value. That means that if a continuous - peak rating differs by more than 3 dB the design begins to be questionable. From that alone one can conclude that available intermittent peak power would be no more than double the continuous value or 3 db up, which does not represent a significant increase in volume - it is sometimes called 'just audible'.
Any 'headroom' concept is too linked to the type of music to be part of a specification. What is normal, what kind of headroom is required by a specific kind of music? For a classic symphony that might amount to all of 20 dB or more, for modern popular music (with the kind of recorded over-amplitude I read about on CDs) it might not be much. It is a matter of experience/taste, and folks will need to estimate what will typically be required in the case of each person.
Ostensibly, dynamic headroom would be defined as the short-term, non-clipping power available beyond an amp's maximum continuous power capability. However, a manufacturer might instead choose to calculate it as the power available beyond something less than that maximum continuous power number. Remember - as discussed earlier in this thread - that some manufacturers specify a continuous power rating significantly below the amp's maximum continuous power rating. In this case, I guess that the difference between the two could be referred to as dynamic headroom. However, I would not consider this an appropriate use of the term.
Yes, 3dB is a doubling of power. Not sure of your point.
Unless the manufacturer specifies the period for which the amp is capable of delivering its dynamic headroom, it might well be less than the time required to fully accommodate transient peaks (I believe that 200ms would be the minimum requirement).
Not sure of the meaning of your comment about distortion. It might well be that power available for dynamic headroom is at a significantly higher distortion level than when operating within continuous power spec.
It is my understanding that best in class amps have little/no dynamic headroom. The maximum power that they can deliver before clipping = their maximum continuous power rating.
Again, my view is that generally one should consider dynamic headroom specs with a large grain of salt, and instead size one's amp on its continuous power capabilities. Anything that it can do above that would be a bonus.
Last edited:
fred soop
Super Member
Not necessarily. It could be that the protection circuits don't allow any additional current at 4 ohms.
As for headroom, IHF power, peak power, etc., I've done some simulation and have looked at various parameters. It appears that most "peaks" that can be handled by amplifier design are only good for a few cycles and wouldn't even handle a reasonable musical transient. When selecting power transistors, the DC SOA curve must be used as the short term pulse curves are too short for anything audio.
whoaru99
Epic Member
It might mean nothing other than how Sony chose to rate the amp.
whoaru99
Epic Member
Yeah, more or less useless. From a specs standpoint it may tell you something about the amp, but a high headroom spec isn't anything you can't do, and maybe be better off with just plain old brute force.
If headroom is really important perhaps you could explain how an amp with 25w and 6dB of headroom would be a better performer than a 100w amp that has zero headroom.
Whoaru - to be clear, in your second sentence I believe you are referring to an amp's dynamic headroom function/specification...right?
I think we would all agree that headroom for transient peaks is an absolute requirement. The question is how to achieve it. I'm inferring that superampman would consider an amp with a maximum continuous output of 25 watts and 6 dB of "dynamic headroom" to be the equal of an amp with a capability of 100 watts of continuous output and no dynamic headroom.
As you can see from my earlier comments, I do not agree.
From my standpoint, if running the 100 watt amp at 1 watt of continuous output, I know that I have nearly 20dB of clean headroom available for peaks (and probably around 24dB before clipping distortion becomes audible).
On the other hand, with the 25 watt amp, I really don't know how much headroom I have (since I don't know how its dynamic headroom spec was calculated), I don't know how clean it is (it might have a THD of 10%, as compared to only .5% for the first 25 watts, for example), and I don't know if it can be produced for a period long enough to accommodate the transients for which it is supposedly available.
I think we would all agree that headroom for transient peaks is an absolute requirement. The question is how to achieve it. I'm inferring that superampman would consider an amp with a maximum continuous output of 25 watts and 6 dB of "dynamic headroom" to be the equal of an amp with a capability of 100 watts of continuous output and no dynamic headroom.
As you can see from my earlier comments, I do not agree.
From my standpoint, if running the 100 watt amp at 1 watt of continuous output, I know that I have nearly 20dB of clean headroom available for peaks (and probably around 24dB before clipping distortion becomes audible).
On the other hand, with the 25 watt amp, I really don't know how much headroom I have (since I don't know how its dynamic headroom spec was calculated), I don't know how clean it is (it might have a THD of 10%, as compared to only .5% for the first 25 watts, for example), and I don't know if it can be produced for a period long enough to accommodate the transients for which it is supposedly available.
Last edited:
Superampman
AK Member
Okay. What I'm gathering from all this is that there is no such thing as a free lunch and that the term 'headroom' is at the very least, a spurious indicator of design quality. Otherwise Krell would have revealed the full spec of their amp. Seems they don't like the term but like the concept. They just don't have a word for it but at the same time don't want their lie of omission confused with the word 'headroom'. Is that a fair statement?