How Much Current Capacity do we Need from our Amplifiers?

[Ropeburn]

At this point, and regarding John's valid safety worries, it'd be much better to get robust outdoor speakers, active with limiting and protection and all. Then the kids can do whatever, and the only risk is to their ears. Then set up an electronic volume limit on the source or something to protect them. Will be nicely portable too, compared to a regular home stereo.

When we're talking about outdoors and long distances, all the LS50s' strengths like balanced response and pinpoint imaging go out of the window anyway.

 

[john61ct]

You can see from the above that if you plan to produce 96dB SPL, at 1m from the speakers, and keep below 1% distortion, then you would already want the crossover to be higher than 600 Hz.

Y'all keep implying the opposite of my approach!

1% seems higher than what I want, and I doubt the HPF will be much over 200Hz.

96dB is nowhere near my target, in fact there IS NO number SPL I am targeting.

I am NOT seeking "high SPL" in any given space, at any specific distance.

I am accepting a total SPL for the system overall, based on the limited "max safe output" of the OG LS50.

So I will not bother spending money on the other dozen amps & LF boxen, other than matching THAT as my target.

 

[-Matt-]

At what distance will you listen?

And what SPL do you want to reach at that distance?

These are the key questions you must answer in order to work out the power requirement.

If you want a fairly quiet 66dB SPL when listening 32m away, then your speakers will have to put out 96dB if you measure just 1m from them.

In that configuration they would have massively higher than 1% distortion if your HPF is at 200Hz.

 

[john61ct]

much better to get robust outdoor speakers

Third time now, the OG LS50s are a given, it is only THEIR limits I am seeking to work within.

> Whether these speakers are "appropriate" for a given space is not the issue.

The system will be actively DSP'd

> with limiting and protection and all

likely all other enclosures being single-driver.

The kids are not a major factor, nor the outdoors aspect, maybe 3-5% of usage.

 

[-Matt-]

Third time now, the OG LS50s are a given, it is only THEIR limits I am seeking to work within.

If you want to keep below 0.5% distortion, with a 200Hz HPF, (look at the 86dB distortion graph).

You can have:
86dB @ 1m
80dB @ 2m
74dB @ 4m
68dB @ 8m

 

[john61ct]

At what distance will you listen?

And what SPL do you want to reach at that distance?

I realise for most use cases these are important factors. In this case however, they are irrelevant.

The safe max SPL of THOSE speakers is what determines that of the whole system.

For critical listening I will get closer. For family film watching we will be further away - but the Center stack and LF boxen will be contributing relatively more.

Dance party mode, we will just have to accept the system limitations.

Background musice, same same...

If you want to keep below 0.5% distortion, with a 200Hz HPF, (look at the 86dB distortion graph).

You can have:
86dB @ 1m
80dB @ 2m
74dB @ 4m
68dB @ 8m

Thanks for that, good to know

 

[peng]

Worst I have ever seen advertised is the 50 W 8 ohm rated Denon PMA-SX1 limited 2Ch integrated amp, 240 A instantaneous. For that amp to push 240 A, even instantaneous peak, the load would have to dip to fractional ohms.

 

[LSPhil]

I keep reading about these “60‑amp peak current” amplifiers, and honestly… that belongs more in a welding workshop than in a living room.
Let’s stay grounded in real loudspeaker physics for a moment.
Take something as simple as a Monacor 3.9 mH inductor — its saturation current is 7.9 A.
That’s it. Not 20, not 40, and definitely not 60.
So unless someone is testing amplifiers with a pure resistor just to justify the price tag, those huge current numbers are mostly marketing aerobics.
And loudspeaker drivers aren’t any more magical.
A typical 12-inch chassis with a 77 mm voice coil defines the maximum current long before any “60 A peak” fantasy becomes relevant.
Sure, you can claim “1000 W”, but that’s music power, not a continuous sine wave.
In the real world, I have yet to see a voice coil that can survive a true 1000 W thermal load.
To build a coil that actually handles that continuously, you’d need so much copper that it starts looking less like a speaker part and more like a commodity investment.
At that point you’re not buying a driver — you’re buying a small futures contract in Cu.
So yes, 60 A peak current sounds impressive on paper…
But in practice, the loudspeaker will tap out long before the amplifier ever gets the chance.
Just a friendly reminder that physics doesn’t read marketing brochures.

 

[MaxwellsEq]

I am accepting a total SPL for the system overall, based on the limited "max safe output" of the OG LS50

OK what is the total SPL for the system overall? Perhaps we can make some calculations based on that?

 

[john61ct]

OK what is the total SPL for the system overall? Perhaps we can make some calculations based on that?

LOL, no idea, that is not even a design goal.

The fixed parameter is the "safe max OG LS50" output.

The MBM couplers, center stack, surround pair, and multiple subs will all be level-matched to blend well with that.

And then turned way down, 95% of the time.

With a top notch variable automatic Loudness contour usually active...

...

And, as I outlined in the post above about different "listening modes", the total SPL maximum will vary by use case.

Finally, when deployed for use in for a tiny space, e.g car camping, might leave nearly all the other boxen back home.

Other cases maybe half of them.

All of them ? going to need a bigger boat...

 

[marin.weigel]

Third time now, the OG LS50s are a given, it is only THEIR limits I am seeking to work within.

We all need to realize, that no one can honestly answer your request satisfactorily by doing theoretical calculations, looking at specs, measurement graphs, and what not.
You really need to whip out the 100 squid for a UMIC1, install REW, download and ingest @Keith_W 's pdf to learn what you're about to have to do for yourself.
Best to take your measurements outside to measure distortion and compression with minimal environmental contributions, while applying your intended high pass filters and then being able to listen to the mess with some music excerpts from a hearing-safe distance.
Good luck with your endeavour ...

 

[Sokel]

Except there are very few amplifiers that actually double when impedance is halved. Even huge Krells and Levinsons didn't, instead they underrated them at 8 ohms which would show up when actually tested. The only amps I recall actually doubling were the Sunfire amps, but they were hampered by other issues (i.e. low sn ratios and moderate distortion).

Tiny Ice's (module fits in one's palm) double just fine with the right PSU and cooling:

(link)

 

[john61ct]

Good luck with your endeavour ...

Thanks, I never expected much, aside from information, what I've learned far exceeded that.

Of course no substitute for hands-on, I have bought a UMIK-2 and getting a working PC put together is near the top of the list, along with moving so far collected components into the main room to get some space.

Kids will have to just get all their stuff moved to their bedrooms, or (ideally) take the hint, time to get their own place before dad moves out.

 

[milosz]

Rating an amplifier's maximum current at some ridiculous figure (e.g., 60 amps!) isn't useful at all. The amplifier's ability to drive reactive loads is what these specs are trying to indicate, albeit poorly. There are better ways to express this ability, and Audio Science Review has made an attempt to characterize this performance by showing power levels into loads of varying reactance. Impedance ( hence reactance ) is a vector quantity, so it's not as simple as saying "XXX watts at ZZZ ohms" which treats impedance as a scalar.

 

[Salt]

Rating an amplifier's maximum current at some ridiculous figure (e.g., 60 amps!) isn't useful at all. The amplifier's ability to drive reactive loads is what these specs are trying to indicate, albeit poorly. There are better ways to express this ability, and Audio Science Review has made an attempt to characterize this performance by showing power levels into loads of varying reactance. Impedance ( hence reactance ) is a vector quantity, so it's not as simple as saying "XXX watts at ZZZ ohms" which treats impedance as a scalar.

^This^

 

[Morinor]

Your numbers for Hypex/Purifi are flawed because at 2 Ohm, power is usually current limited by the PSU while at 4 and 8 it isn't.

Hypex NCx500 Class D Amplifier Review

This is a review and detailed measurements of the Hypex NCx500 NCOREx reference design amplifier. It was sent by the company and kindly assembled by our forum expert, @Rick Sykora. As you see the sample is a "monoblock" configuration (one channel) driven by Hypex SMPS1200 switching power...

www.audiosciencereview.com

View attachment 540328
As you see, 700W at 2 Ohm doesn't mean 350 at 4, rather you get 670.

You didn't got my point. If you wanted for any reason to use that 670 watts, you simply can't if your speaker can have its impedance dropped nearer to 2 ohms. It will start distorting as it is getting current limited. So unless you know that the speaker it is a relative easy one getting limited above 4 ohms, those watts remain just numbers. Especially if you know nothing about the speaker impedance curve, you can know that you amplifier above can easily manage playing around 175 watt at 8ohms or 350 watt at 4 ohms. There are amps that advertise a good 8 ohm watt performance, a less than 50% more at 4 ohms and 2ohm that performance doesn't go far. This means that you should be careful on speaker pairing if you know that you need so much watt/current.

 

[EJ3]

In my opinion amplifiers should do their job relative with the load to be more stable. This means double power capabilities at 4 ohm and quadruple at 2ohm. Only AB amplifiers lower than 1000-1500€ can be justified to not be such capable. Otherwise the opposite way of calculation takes place to evaluate an amp that will not sweat. Meaning divide the capability of 2ohm by 4 and 2 in order to find how much power it can handle at 8ohms and 4 ohms. Above those prices there are few AB amplifiers that can truly follow this rule. Cambridge Edge A (or W) are one of those examples. Some Rotel around the 3000-5000€ can do that and I think NAD 40n is also capable. Topping B200 although very clear at 4ohms is very limiting reaching lower impedances. Class D from Purifi or Hypex are the better options for not caring about current. 500 watts at 2ohm mean that the amplifier can drive any 8 ohm nominal speaker with 125 clear watts or any 4ohm nominal speaker with 250 watts. Purifi 1ET9040 mostly at 1200 watts at 2ohms means equivalently 300 watts for 8 ohm and 600 watts for 4 ohm.
One comment about the example used on the original post, we would wished to had such easy speaker. On most frequencies it doesn't reach below 6ohm and at high frequencies although dropping there is usually less need for power/current there. Only few expensive speakers can do that and most of them are two way. Usually there can be drops at around 3-3.5 ohm and even 2.5ohm at low frequencies that are more demanding.

In "theory" they should double the watts for each halving of the Ohms: "under perfect conditions". Which almost never happens in the real world. So it is not that it normally happens in reality.
There are many nuances involved. But, like shotguns and hand grenades, coming very close: counts.
Here is another (interesting to me [because I own one {as well as 6 NAD 2200's and 2 NAD 2100's}) amp dyno test for a PROTON D1200 which seems to show some good responses to 4 ohm and 2 ohm loads VS the 8 ohm load:

 

[john61ct]

So, which DC-powered power-efficient chip amp with a better SQ to price ratio than 3e

enough power + current to destroy a pair of LS50s (more the better, may also use to power subs)

and balanced signal inputs

do y'all recommend, for USA purchase, without tariff risks?

TPA3255 based is fine, but open to other options.

Finished box no raw boards, but built in bass management / DSP would be nice, 4-channel even better.

Fosi CA30 is the one to beat, but Amir not yet tested...

@harkpabst ideas?

 

[wwenze]

EDPR is 1.7 Ohm, right. Now Ohm's law: Current = squareroot (power / resistance). Makes sqrt(128/1.7) = 4.28 Ampere.

EDPR does not increase the current that the transistor needs to handle. It does not increase the amount of power the amp needs to push out either.

It does, however, give us a prediction of the required amount of cooling needed at the transistor in the form of equivalent resistive load. Not a perfect prediction either, since we know good class D amps don't care about that number.

 

[Ropeburn]

EDPR does not increase the current that the transistor needs to handle. It does not increase the amount of power the amp needs to push out either.

It does, however, give us a prediction of the required amount of cooling needed at the transistor in the form of equivalent resistive load. Not a perfect prediction either, since we know good class D amps don't care about that number.

Yeah I'm aware. It's an absolute worst case estimation - for safety since that was a general concern.

Worst case scenarios have the benefit of removing all doubt and providing good rules of thumb. Here it's telling us: no worries.