How Much Current Capacity do we Need from our Amplifiers?

[CDMC]

I've seen this mentioned all too often, but everyone needs to realize, that there's a price to pay when generously over-provisioning amp power, which is much lowered S/N values when usually listening at a comparatively tiny fraction of full scale output from pre/power amps.

Crippled by this shortcoming, there's these audiophiles that then sweat about their DACs specs, upsampling algorithms, modulators and what not - but it's all swamped by comparatively high noise, and I'm not even going to discuss actual audibility of such measures under optimum conditions.

The ticket to best sound is to get the system chain's gain structure optimized to wring the maximum transparency from it during one's most usual listening scenario.

So what do you aim for? With a single listening level of say average 80db and a 83db efficient speaker with a 20db peak, you go from .5 watt to 50 watt. That is relatively loud, but high on the distortion curve for most amps. Slide down to moderate 70db listening, now your average power is .05 watt. What amp has its best specs between .05 watt and 20 watts? It seems to me, using an amp where the s/n ratio never gets worse than 85db, like a hypex or purifi makes it a non issue, still well below audibility.

 

[CDMC]

For some people $750 is a huge investment for just two channels. Better to be able to afford better speakers if a 100W power amp is enough

Yes, but I would also argue people on tight budgets who want higher spl levels should be conscious of their speaker sensitivity and difficulty to drive. Don’t go buying old Kappa 9s if you are on a tight budget!

 

[wwenze]

I've seen this mentioned all too often, but everyone needs to realize, that there's a price to pay when generously over-provisioning amp power, which is much lowered S/N values when usually listening at a comparatively tiny fraction of full scale output from pre/power amps.

Crippled by this shortcoming, there's these audiophiles that then sweat about their DACs specs, upsampling algorithms, modulators and what not - but it's all swamped by comparatively high noise, and I'm not even going to discuss actual audibility of such measures under optimum conditions.

The ticket to best sound is to get the system chain's gain structure optimized to wring the maximum transparency from it during one's most usual listening scenario.

Amir tests at 5W so let's use that as a starting point.

With a 83dB/W sensitivity, that is 90dB SPL.

If the amplifier has noise that is -90dB below the signal at this point, the noise level would be at 0dB SPL.

Not a difficult target to hit nowadays. The cheaper / poorer performing amps give you -70 to -80dB? But the good ones recommended here clear that easily, including the >100W amps.

 

[popej]

I was at the 2006 demo, which was a lot of fun. One track by Rickie Lee Jones showed average power on the playback system of 1-2 Watts, but reached 250 Watts on a snare drum hi

Then it is 21-24dB peak to average.
I have never been able to find a music track with that kind of dynamic. I mean, if this really happens, then should be easy to provide reference: artist, song. And we could verify on PC.
It could be different for live performance, but it is not, what we use our amps for.

 

[Sokel]

Then it is 21-24dB peak to average.
I have never been able to find a music track with that kind of dynamic. I mean, if this really happens, then should be easy to provide reference: artist, song. And we could verify on PC.
It could be different for live performance, but it is not, what we use our amps for.

Here's one for example (Ron Tutt) :

Many more at classical, 20 dB is just about the norm there.

 

[popej]

I have looked at DR measurements for albums:

Album list - Dynamic Range DB

There are 32 albums with DR above 20. Very rare, but they exist.

 

[DanielT]

I don't know if this question has been addressed in the thread: If an amplifier provides twice as much power in 4 Ohm vs 8 Ohm, is that a rule of thumb to consider a good design for an amplifier, it should have the current capacity to handle that?

Speaking of the opposite, Hitachi HMA-6500. I bought it because of the good price, mostly because I like the VU meters.

The amp works flawlessly but it struggles when I turn it up really high with a pair of 90 dB sensitive, according to spec 6 Ohm, speakers. I have the opportunity to press the sound pedal to the metal now. This is clearly visible because the lamps for the VU meters are powered by the amp's power supply so at higher volumes they start to blink, or the brightness varies in step with the music. When the bass is pumped up they light up less. The lamps go in rhythm when power for the speaker is needed. They almost function as clipping indicators, you could say.

That 50 watt Hitachi power amp, nice but feels a bit weak. 50 watts in 8 Ohm, 53 watts in 4 Ohm, peak current capability 5.3 amps.


My Cyrus 2 amp with PSX Power Supply feels much more capable of giving the 90 dB speakers a match. Actually (subjectively), a really big difference in what it can perform. Purely objectively more power vs HMA-6500 in low Ohms, (which the speakers can dip into) so it might not be so strange).
70 watt in 8 Ohm, 125 watt in 4 Ohm, current capability 60 amps peak to peak
(I don't know what the continuous current level is, more than the Hitachi anyway.)

(Reservoir capacity 70,000 uF...doubtful if AI got it right, seems way too much)
(Screenshot: Gemini AI )

 

[Ropeburn]

View attachment 540526

I haven't forgotten.

Ähähähähä I see you're learning

 

[Geert]

I don't know if this question has been addressed in the thread: If an amplifier provides twice as much power in 4 Ohm vs 8 Ohm, is that a rule of thumb to consider a good design for an amplifier, it should have the current capacity to handle that?

Lets poke a hole in another myth "doubling down" is not really needed it does not affect the amps performance within its limits .

Some less informed seems to think an amp that does not double down loses dynamics or compress dynamically or is nonlinear within its useful envelope . There basically a current limit a voltage limit and a thermal limit untill you "hit" a limitation the amp is linear and works fine .

You can run around like crazy in your room untill you hit a wall the non doubling down amp moves some of the walls a bit closer . But if your amp is powerful enough anyways you wont suffer any consequence. IMO is the usual marketing scheme when you take a kernel of truth and stretch it beyond its applicability .

 

[wwenze]

I don't know if this question has been addressed in the thread: If an amplifier provides twice as much power in 4 Ohm vs 8 Ohm, is that a rule of thumb to consider a good design for an amplifier, it should have the current capacity to handle that?

1) It means nothing

or

2) It means the amplifier has low power supply voltage, which if a fanboy starts crying he can choose to go back to point (1)

Optimizing for best price/performance means you want exactly the most amount of current and the most amount of voltage you can get for that load impedance at that $$$.

This is why for centuries amp designers have optimized their amp to have maximum POWER with a speaker load of somewhere between 4 ohm and 8 ohm, such that you don't suddenly lose half the output power by going either direction.

As for burst/peak/whatever current (which, I need to first mention, does not imply an explicit support for lower impedance load), that is up to the designer, as it is often tied to the distortion numbers during normal operation. When you reduce the load impedance, distortion goes up. Don't take my word for it, look at all the 4ohm vs 8ohm measurements on this website.

 

[DanielT]

Makes sense.

Thanks for the helpful answers, @Geert @wwenze @Mnyb

 

[Ropeburn]

I've seen this mentioned all too often, but everyone needs to realize, that there's a price to pay when generously over-provisioning amp power, which is much lowered S/N values when usually listening at a comparatively tiny fraction of full scale output from pre/power amps.

Crippled by this shortcoming, there's these audiophiles that then sweat about their DACs specs, upsampling algorithms, modulators and what not - but it's all swamped by comparatively high noise, and I'm not even going to discuss actual audibility of such measures under optimum conditions.

The ticket to best sound is to get the system chain's gain structure optimized to wring the maximum transparency from it during one's most usual listening scenario.

That's such a non problem with modern amps. The noise floor is fixed, as long as it's low enough to not be audible it's... always inaudible, no matter how silently or loudly you play. 500W Hypex something something does this. There is no relevant noise from those amplifiers.

There is no "much higher noise" and certainly no "crippling" even with the most powerful versions. I'm afraid your points once used to be true long ago and are now hopelessly outdated.

 

[Ropeburn]

For some people $750 is a huge investment for just two channels. Better to be able to afford better speakers if a 100W power amp is enough

It's certainly an investment but it's also absolute endgame for home stereo, both power and quality. Considering that it's dirt cheap.

 

[pjug]

I have looked at DR measurements for albums:

Album list - Dynamic Range DB

There are 32 albums with DR above 20. Very rare, but they exist.

The ones at the top of the list are insane. Tracks with crest factors of 45dB! Thanks a lot; now I know I need a more powerful amp.

 

[popej]

I don't know if this question has been addressed in the thread: If an amplifier provides twice as much power in 4 Ohm vs 8 Ohm, is that a rule of thumb to consider a good design for an amplifier, it should have the current capacity to handle that?

Basically, it is not capacity to handle current, but capacity to deliver enough voltage to drive high current.

There coul be a few reasons for design, that double power:
Amp is using stabilized power supply. IMHO the only reasonable solution is to use SMPS supply, which provide stabilisation without losing efficiency.

Amp is overdesigned for testing. While it can be well engineered and powerful, it will be expensive.

Manufacturer lowers specification for 8 Ohm.

 

[john61ct]

The ability to deliver a lot of current to handle short term peaks

is this part of "transients" performance?

is what maintains the high voltage despite lotsa Watts power demanded.

Voltage dropping during peak demands is a symptom of not enough current being available.

If impedance dropping dramatically is the cause, LS50s rated nominal 8Ω but in fact touching 3.2Ω

then yes, amps that (close to) double power as impedance is halved

especially those rated to handle 2Ω or lower

are indeed those with a higher chance of performing well with those "difficult to drive" speakers.

I've gone back to the LS50s review thread, and yes rated for 100W, but in fact for short peak cycles will draw 200-250+ Watts if the amp can supply them.

So in measuring their performance with a chip amp comparing to my GFA-555 can normal equipment + say REW document such differences?

 

[EJ3]

So what do you aim for?

I aim for being able to listen when I am out in my yard. And I am almost never in the room that the stereo is in, when I am listening. So, a different use case.

 

[Spkrdctr]

I aim for being able to listen when I am out in my yard. And I am almost never in the room that the stereo is in, when I am listening. So, a different use case.

Holy Cow!! You took the "My wife can hear the difference in the kitchen" and stepped up to I listen to my music "in the yard"! Now that means speakers that can put out some serious spl and an amp than cane handle it all without massive clipping. You are what I like to call a 1 percenter. That rare individual who has an extreme use case. I have to tip my hat to you, I never thought of having a system designed for in home use but using for yard listening. I learn something here every day. Your system must have some serious balls as it would be referred to by audiophiles!! I guess the obvious question is, what do the neighbors say?

 

[Ropeburn]

I guess the obvious question is, what do the neighbors say?

 

[EJ3]

View attachment 540636

Why would I have a yard so small that I would bother the neighbors? That defeats the purpose of having a yard.
Not an issue here, though:

The last home on the electrical grid, nearest neighbor: 1/4 mile. Second home (40 miles away) back yard: