Is a bunch of amp headroom necessary for nearfield?

[Justin8888]

Im trying to wrap my head around the need (or maybe lack of need) for a bunch of power for a nearfield/desktop setup. Im looking into running two Schiit gjallarhorns f in mono and have not settled on a speaker. The price on these amps is attractive, they are compact, American made, well priced, seemingly well built with a strong reputation behind them. And placing them near each speaker is ideal for my layout. Redundancy is nice as well if one goes down.

But this concept of needing headroom or, as I noted below, "a beefy amplifier" is confusing to me.

https://www.audiosciencereview.com/...-debut-reference-dbr-62-speaker-review.12232/

However, note the area I have circled. Impedance is high at about 8 ohms but the phase angle is quite acute at nearly 50 degrees. This means the speaker will ask for current when the output voltage is very low. So you better have a beefy amplifier to drive this speaker.

In this case with the dbr62, theres a sharp drop to 5ohms. These are 86dB speakers. Assuming I am listening at the max volume I would ever consider, lets say average 90dB (rare, but sometimes I get there). Im using less than 2 watt of power which means if these suddenly require a spike in power due to an impedance drop, they can spike to around 10x the continuous and before they hit the ~2A cutoff on the ghorn.

Anyway, this is one example. I have not yet settled on speakers, but I would like to know how I should be thinking about how much power I need to accommodate large, and particularly, low swings in impedance to maintain consistent performance in nearfield. Regardless of speaker, if the continuous wattage is low, like 2W, how much headroom to a really need, even for "hard to drive" speakers ie speakers that drop even lower to like 2ohm (looking at Q1 meta and ML 15i, and some others). Do I really need 125-200w per channel for this application? What constitutes a "beefy amplifier" other than power output?

 

[wwenze]

That phrase is kind of misphrased, but it's not easy to explain the full picture anyway

The amplitude of the current follows the amplitude of the impedance. E.g. 8V into 8ohm resistor results in 1A of current. Similarly, 8V into a capacitor that is 8ohm impedance will also result in 1A of current. I should state clearly that voltage from source results in current through load, not the other way round.

The difference between a resistor and capacitor is that, with a resistor, the current is in phase with the voltage. With a capacitor, the voltage at the capacitor lags the current by 90 degrees.

Now, the problem this presents to real-world amplifiers is that, the amount of heating across the output transistor is easily modeled as I * V across the transistor, where V is the voltage between power supply voltage and output voltage. And class D gets high efficiency by only operating in the "high V zero current" and "zero V high current" states. Meanwhile class AB operates like... normal...? for a lack of better word, where the I and V just goes up and down with the waveform. Still, V is usually lower when I is higher.

Notice that in the case of capacitance or inductance, the current is highest when voltage is zero. Meaning all the energy is being dissipated by the output transistor since it takes all the voltage while there is current. In fact, a capacitor or inductor has a power factor of zero, because no energy is dissipated by the inductor or capacitor. So any electrical energy that gets pushed into the capacitor or inductor gets thrown back into the amp as the voltage changes over the cycle. Some amps (i.e. class D) are good at eating this back power, other amps not so well.

In short, while the VA (volt-amp) or apparent power calculation remains straightfoward, any power not dissipated by the speaker from the resistive component is sent back to the amp and needs to be gotten rid of.

On the other hand, this issue can get overblown, humans are always scared of the unknown. In practice, music is a combination of frequencies. Some frequencies have voltage lagging current, some frequencies have current lagging voltage. Combine all the frequency components together and the currents mix and cancel out. Such peaks are only a problem when you decide to play a single frequency at that peak.

Also, we know whatever worry we have about this topic is unscientific, because if it were scientific, people would be telling you how much extra power amp you will need. And we don't know, not even a rule of thumb, whenever this topic comes up. Which is strange considering audio people are a group that even have rule of thumb for values like RMS/program/peak power and crest factors and cooling and arguing each other to death over what is the correct value, yet we never ever see people arguing over how much extra power amp we need to deal with impedance dips and phase angles. Thus there is only one conclusion: Nobody knows this topic enough to even reach the hill of the Dunning-Kruger curve, because nobody has even encountered the effect in real life.

 

[Justin8888]

Appreciate the reply. I’m trying to think about this in practical terms. I hear people say things like “this speaker really came alive when I hooked it up to more power” is there any truth to this if you’re not hitting the amps current cutoff? Are some amps better at handling big swings in power even if it’s all below the current limit? If so, are there specs on the amp that can be used to compare this? I’m having a hard time on paper justifying the need to move to something like ncore monos when I shouldn’t actually need 80% of the power in my application. But if there’s a real measurable use case ie my speakers will be lacking on the low end (despite not hitting current cutoff, clipping) I would love to know this and be able to compare appropriately

 

[Joe Smith]

I use one original Gjallarhorn currently in my nearfield system, paired with a Saga 2 preamp. Practically speaking, with two Gjallarhorns, I think you will have plenty of power even for those ELACs (I currently am using the 2022 ELAC Debut B6.2 speakers in this rig). Unless you plan to play at very high volumes nearfield (not a great idea for your hearing) you should be fine.

Given my room sizes, and speaker efficiencies (most of what I own is ~87 dB efficient) I have considered the entry-level Hypex amps from Buckeye, but it just feels like overkill.

 

[wwenze]

Debut has a similar impedance curve

And mine is running on a TA2024 in the living room.

 

[DVDdoug]

Regardless of speaker, if the continuous wattage is low, like 2W, how much headroom to a really need, even for "hard to drive" speakers ie speakers that drop even lower to like 2ohm (looking at Q1 meta and ML 15i, and some others).

I would be "safer" to get an amplifier rated down to 2-Ohms.

The good news is that you only occasionally and briefly play full-power "notes" at that worst-case frequency. Constant test tones at the frequency of lowest impedance could be risky. (Constant full-power tones can be dangerous to the speakers too!)

Impedance (and resistance) are both measured in Ohms and they are "the resistance to current flow". (Lower impedance results in proportionally more current.) Ohm's Law says: Current = Voltage/Resistance.
Power (Wattage) = Voltage x Current. From Ohm's Law you can also derive Power = Voltage squared / Resistance (often handy) and Power = Current squared x Resistance.
As above, with a non-resistive impedance you get you get a phase difference between the voltage and current. That results in less actual power to the load (speaker) but it will still stress the amplifier.

Amplifiers are considered "constant voltage" devices, which means the voltage is independent of load. If an amplifier is playing putting-out a 10V test tone with no speaker connected, and then you connect a speaker, it will still put-out 10V but now current will flow through the speaker. Unless the impedance is too low and the amplifier can't supply the "calculated" current...



If you "try" to get more current than the amplifier is designed for, a few things can happen:

- It might Clip (distort), which means the voltage and current are both clipped.

- The amplifier might overheat and go into protection mode, or it might blow a fuse or permanently "burn up" and die!

Do I really need 125-200w per channel for this application?

It depends on how loud you listen and the sensitivity of your speakers, etc. I you aren't "pushing" the amplifier to 125 watts you don't need the "headroom".

Headroom is a funny thing - If you aren't using it you don't need it. If you are using it, it's no longer headroom!

What constitutes a "beefy amplifier" other than power output?

That's probably what they mean. But sometimes it might just mean heavy, or perhaps electrically rugged so it can put-out extra current with low impedance loads, or maybe that it's more likely to survive when over-stressed, etc.

Appreciate the reply. I’m trying to think about this in practical terms. I hear people say things like “this speaker really came alive when I hooked it up to more power” is there any truth to this if you’re not hitting the amps current cutoff?

No. If aren't clipping you don't need more power (or more current). If you aren't using the power you aren't using the power.

If so, are there specs on the amp that can be used to compare this?

There is no standard spec/measurement but some amps do have temporary reserve power (headroom) above their continuous rating/output. Maybe the amp can put-out 100W continuously but when playing normal program material (let's say 20-30 average), it may be able to put-out more than 100W for a fraction of a second for a loud drum hit, etc.

 

[Digital_Thor]

Headroom is a funny thing - If you aren't using it you don't need it. If you are using it, it's no longer headroom!

Also depends on how well a given amplifier handle clipping. I've heard some smaller class D that sounds really bad as soon as they hit the wall - so to speak.
But at you also mention, when does it happen, and how to know when enough is enough?
I just went full throttle, and have tons of power in reserve, since it's not that expensive and it's easier to have to much than to little, when doing DIY and being creative.
But in the case of the thread starter, I believe that many amplifiers today can make him happy, if the room, setup and EQ is done well.
Biggest pitfall I often see, is that neither output power, class, size of either speaker or amplifier, truly define whether it's easy to drive or capable of difficult loads.

I would look into what max spl you wish for, then see if either EQ or a subwoofer can make your speaker capable to your preference at the given listening distance. Then see if that speaker has any tricky loads, like low impedance or twisty phase issues in the passive filter. Then you can narrow down which amplifier can live up to this challenge.
Everything within your scope of abilities and budget.
I have the KEF R3, and as I understand it, they are not an easy load... But some small class D amplifiers don't care about this
.. so read carefully into the details when choosing

 

[Mnyb]

How long is a string ? I say practical experiment/experience in your own setup will tell ?

Speakers sensitivity ( current and future ) room size and distance and what you consider typical listening loudness ( only you can decide )
And how much headroom will your EQ eat up if you boost certain areas.

I would guess that a normal 100wpc at 8ohm amp would be fine for most nearfield application not 5w set or 2500w PA amps .

Buy a reasonable amplifier that may survive several speakers in the future, as you are nearfield I would pay attention to noise performance after sufficient power is satisfied .

Otherwise I tend to spec my amps at the speakers max rating , so I can utilise my speakers fully in whatever setting is needed in current and future use. Earlier in life I moved to different apartments so rom size was not a determining limit .

For nearfield it would bee uneconomical to spec 250/500 w mono blocks .

Maybe overkill but what about a stereo purify amp the lowest power spec that’s currently on sale not mono you save one power supply.
Maybe to much power but they are very quiet and if you buy one with selectable gain you can have it in low gain setting for even more quiet , you might need to crank the volume high for normal use , but that would be perfect nearfield ?

 

[Digital_Thor]

I'm quite sensitive to high pitch coil wine from any switch mode design/construction - so do consider whether that annoys you, before planting one half a meter from you, on the desk. I'm not saying they all do this.. but some do... like two NC502MP based amps I once borrowed. It is not much... but at close distance... it was sadly there, which some can hear... and some can't. But try it... then you know, by visiting some who have one, and then put your ear to it
At 3 meter in a normal living room listening position, that is of course rarely a problem.... here it is often more big tranformer hum, that can be tiresome, during silent moments.

Maybe you would be a happy camper with either Q3 Meta or Concerto Meta with the Wiim amp pro/ultra - using the EQ that spinorama/Amirm/Erin specify for the specific speakers - and then later, you can consider plugging the ports and add a little subwoofer, to have a little "omphf" at lower volumes - because Fletcher muncher is a real thing.

 

[ZolaIII]

How much power you need/want is more to driver sensitivity than anything else. Let's say in home, HiFi this range is between 9 dB (from low 80's to beginning of 90's per W at 8 Ohms) disregarding extremes. +9 dB is 8x the power. Smaller the speakers they will also be less efficient to same riggednes and stiffness of a cone material. Bigger the speaker it will be able to hold in linear excursion and to higher SPL but at the cost of directivity (over 5.5~7") and at time domain over the certain point (10"). With increase in the distance you lose to the free field - 6 dB with every distance dubbing. In the room is less and not all rooms are same but let's say it's about - 5.5 dB. So far mid field you need 4x more and for far field 16x more. Ideally you do want amplifier noise levels (SINAD) are above speakers sensitivity so that they stay silent when not reproducing anything (and still more important to near field and to distance doubling according to actual losse).
But you won't push mains, instead you will relax them with sub's and if high enough (high pass) up to their impedance/phase plots meet it will be much easier to any amplifier to drive them. Still those that drop very low regarding impedance and more so in areas that are hard to stress out should be avoided putting both stress and heat upon them self and whatsoever is driving them and there for decreasing the lifespan. Schiity is at best most part assembled in US. You have more than several tested small and affordable class D amps with ironed out things (like load dependency, noise increase with F...) which will give enough power to average 85 dB speakers in near field and more so with sub/sub's.
Psy how we hear is really as good in our capacity we can to full octave spectrum to how sound spreads physically so we increase amplitude to compete for time domain. That's ELC to no ELC reference calibration point of 85/88 dB SPL mono/stereo focused white noise. For better ELC integration for let's say under 70 dB program SPL (0 dB to material DR) you need two sub's as extension of mains each to it's own as knew for ELC is 105 Hz and because you want to sustain it's fundamental propagations on subwoofer's driver as boost in low and sub bass becomes more significant. There for 120 Hz crossovers. In that case bottom limit will be determined by level of environmental noise and decay times of the room even under the level of normal come speach (60 dB SPL). For usual front desired listening levels of mid to low 70's even single chenel sub/sub's LFE to 80 Hz crossover will do just fine.

 

[Ropeburn]

Given my room sizes, and speaker efficiencies (most of what I own is ~87 dB efficient) I have considered the entry-level Hypex amps from Buckeye, but it just feels like overkill.

It factually is overkill - say, 2x250W - but only for that specific use case. Do you only sit there? Do you ever stand up and walk around, and turn the music up, just having a blast? What about the future - maybe different speakers, bigger room, using the amp for a big living room stereo?

There's countless scenarios where plenty of headroom comes in handy. These amps are still cheap enough. Personally I'd always want options and have it be powerful enough for all occasions so I simply never have to worry about anything, whatever happens. Mainly running just a few Watt while it's designed for heaps is also good for heat and longevity, compared to a tiny weak amp you actually use most of the power of.

As long as money and size isn't an issue - the nCore series basically solved both problems, let alone even cheaper class D - I'd always go for what's mostly overkill but comes in handy at least occasionally. I found around 2x200W is a very good sweet spot between price and performance, you almost never need more in a typical home scenario.