How much amplifier power is required?

[NTK]

... As a follow-up , are there power amps , that can enforce hard limits on o/p power, if say the Volume control on DAC/Preamp , gets accidentally turned all the way ? Is that the same as gain controls on the amp ?

Some pro amplifiers, for example Crown, have limiters. I seriously doubt you'll need that level of protection.
https://www.crownaudio.com/en/site_elements/levelmax-guide-for-i-tech-hd

 

[HifiLove]

I am the volume limiter here.

very funny ;-)

was searching up on limiters.., and found this very informative article to our subject from Harman

https://pro.harman.com/insights/har...of-using-power-amplifiers-that-are-too-small/

 

[RichB]

The Benchmark AHB2 limits based on overcurrent, over voltage, and maximum distortion.
The spartan 6 FPGA is used for protection.

- Rich

 

[HifiLove]

The Benchmark AHB2 limits based on overcurrent, over voltage, and maximum distortion.
The spartan 6 FPGA is used to for protection.

- Rich

$3K and "100 Watts per channel into 8 Ohms"!

 

[DonH56]

Clipping: https://www.whatsbestforum.com/threads/clipping-101.8484/#post-147098

IMD: https://www.whatsbestforum.com/thre...ion-imd-why-dont-we-like-it.7928/#post-135952

@amirm : I did not see these ported over here?

 

[Frank Dernie]

I can see it now. Frank with his head down the horn, symphony level volume, listening for hiss and the cat jumps on the remote...

Luckily no cat and no idea where the remote is!

 

[Frank Dernie]

I looked into it and just for fun , estimated the power requirements for the +20db peak, over reference level(85db), for my listening distance and speaker sensitivity of 91.5dB...and the required power estimate of 70.7w for speaker with 8 Ohm nominal impedence and 141.5w for speaker with 4 ohm minimal impedence

That is where your calculation is simplistic.
Moving coil speakers don't present a simple resistive load.
Not only will the minimum be well below nominal there will be phase shifts too.
This is where the difficulty comes.
There may be 2 amplifiers which are quite happy delivering 100 watts, say, into a 8-ohm powert resistor on a test bench but one of them may clip when the impedance drops and the other not.
It is a fundamental flaw in the testing of amplifiers in the English language press.
As far as I know there is no way from published data, either of the speaker impedance or how a given amplifier may respond to that particular speaker in terms of power output or stability.
Looking for flat tops in the signal on an oscilloscope connected to the speaker terminals will show if clipping is taking place but that is only valid for the amplifier and speaker being measured, it can't be translated to general conclusions.

The main thing is that some speakers are extremely brutal loads. One could argue that good speaker designers take the load into account and some do not.
The ones I know about are Harbeth, which are all easy loads, and Wilson, none of the ones I have seen plotted are.
One would expect Harbeth to perform well with any amplifier and Wilson to need a welding station to power them.

 

[Frank Dernie]

As a follow-up , are there power amps , that can enforce hard limits on o/p power, if say the Volume control on DAC/Preamp , gets accidentally turned all the way ?

To give extra headroom for low output sources (and quiet recordings) most amplifiers will be delivering maximum power with the rated input with the volume control at 1 to 2 o'clock. Maximum rotation of the volume control will be well into clipping with, say, a standard CD player as input.

Some of the more recent amps have a dB scale on the volume showing how close to maximum power you are selecting.

I am familiar with the Devialet models. On the "Expert" amps 0dB on the volume control is full power with rated input (like a digital input) but +15dB is available if a low output analogue input, like a FM tuner, is being used. The maximum output power can also be set in the configuration file to be lower if wanted. The Phantom speakers produce maximum output from a digital signal at a "70" volume setting but go to "100" where there will be limiting to avoid damage.

 

[Unclevanya]

The amp power ratings on speakers follow no standard. That's just marketing and CYASpeak.

Matching an amp:

Gain shouldn't be so high that noise is prevalent at low level listening.

Power and the ability to handle complex loads should be sufficient to avoid clipping or minimize bad behavior during clipping (Original NAD 3020 soft clipping for example)

And let's hope it sounds good - lol.

I have owned many speakers and amps. I've only ever blown one set of Calrad bookshelf speakers. Tiny things with 3 or 4 inch woofers. They were blown when someone turned up the volume on a 45 watt sansui receiver while it was turned off and it was turned back on playing "We will rock you..." The poor things died nearly instantly. Clipping or overpowering I don't recall which - both perhaps, as I recall the woofers and tweeters were fried.

 

[sergeauckland]

To give extra headroom for low output sources (and quiet recordings) most amplifiers will be delivering maximum power with the rated input with the volume control at 1 to 2 o'clock. Maximum rotation of the volume control will be well into clipping with, say, a standard CD player as input.

Some of the more recent amps have a dB scale on the volume showing how close to maximum power you are selecting.

I am familiar with the Devialet models. On the "Expert" amps 0dB on the volume control is full power with rated input (like a digital input) but +15dB is available if a low output analogue input, like a FM tuner, is being used. The maximum output power can also be set in the configuration file to be lower if wanted. The Phantom speakers produce maximum output from a digital signal at a "70" volume setting but go to "100" where there will be limiting to avoid damage.

One of the (many) benefits of active 'speakers is the ability to have different amplifier powers for the different drivers according to their efficiency and power handling and/or limiters to avoid damage. It may be beneficial to have a bigger amplifier for the tweeter than the tweeter can handle on a continuous basis, provided there's a limiter to limit the total energy, and consequently the heating effect to what the tweeter can handle safely. This will allow for high instantaneous, short term power, but extended high levels will be limited, usually with a warning light to so indicate.

This can also be used in the bass, to extend LF whilst managing extension against excursion against distortion and dissipation. It can be low, loud or long, choose any two out of three.

An integrated active 'speaker like Devialet will almost certainly have this. Even my 25+ year old Meridians had it.

S

 

[Biblob]

Can someone explain to me. Is it true that when you go higher up the frequencies the required power lowers (with the same SPL).
So that a 100 hz needs more power than 1 khz, and 1 khz more power than 10 khz?
If so, is there a way to calculate this?

I picked up on this thought that when high-passing in an bi-amped system, the high-passed amp would get less strain.

 

[HansHolland]

Can someone explain to me. Is it true that when you go higher up the frequencies the required power lowers (with the same SPL).
So that a 100 hz needs more power than 1 khz, and 1 khz more power than 10 khz?
If so, is there a way to calculate this?

I picked up on this thought that when high-passing in an bi-amped system, the high-passed amp would get less strain.

No, not true. A loudspeaker with a flat frequency response and a flat impedance has the same efficiency for all frequencies.
Ok, no loudspeaker has a flat frequency response and/or a flat impedance. But, that does not explain your "when you go higher up the frequencies the required power lowers (with the same SPL)".

I think that I know where your thought comes from:
With a lot of music, the power is spread (roughly) equally accros all frequencies like pink noise. Meaning that there is the same amount of power in the range between 20 Hz - 200 Hz and 200 Hz - 2 kHz and 2 kHz - 20 kHz. If a 2-way loudspeaker has the crossover frequency around 3 kHz, then more than 2/3 of the power goes to the woofer and less than 1/3 of the power goes to the tweeter.
Was this it?

 

[Biblob]

No, not true. A loudspeaker with a flat frequency response and a flat impedance has the same efficiency for all frequencies.
Ok, no loudspeaker has a flat frequency response and/or a flat impedance. But, that does not explain your "when you go higher up the frequencies the required power lowers (with the same SPL)".

I think that I know where your thought comes from:
With a lot of music, the power is spread (roughly) equally accros all frequencies like pink noise. Meaning that there is the same amount of power in the range between 20 Hz - 200 Hz and 200 Hz - 2 kHz and 2 kHz - 20 kHz. If a 2-way loudspeaker has the crossover frequency around 3 kHz, then more than 2/3 of the power goes to the woofer and less than 1/3 of the power goes to the tweeter.
Was this it?

Yes thank you.

So if I understand it correctly, it doesn't help an amp if it would only need to perform a portion of the frequency range? So in a multi-amped system, one amp for 20-200 hz, one amp for 200-2khz and another for 2khz-20khz.

 

[HansHolland]

So if I understand it correctly, it doesn't help an amp if it would only need to perform a portion of the frequency range? So in a multi-amped system, one amp for 20-200 hz, one amp for 200-2khz and another for 2khz-20khz.

They all 3 deliver just 1/3 of the power, so they all 3 get less hot. And all 3 can have a smaller power supply (or all 3 get their power from the same big one).

And in the details:
-Less intermodulation distortion
-Less thermal drift because of less power for each amp
-Or each amp can have a single set of power transistors/tubes instead of 3 in parallel (because of less heat)

But is it worth the price?

 

[HansHolland]

A reason why some like using for each loudspeaker unit a seperate power amp is that they can use e.g. a tube amp for the tweeter and a MOSFET amp for the woofer. They do this because they like the tube sound, but their loudspeaker needs some higher damping factor from the amp for the lows.

I don't like this!
Why?
I too like the homogenic sound of a 1-way loudspeaker, but in general the frequency span that a 1-way loudspeaker covers is too small for me. So, for loudspeakers I accept that we "have to" split up the music in at least 2 parts, one for the woofer and the other for the tweeter.
But for amps? Most amps have a broad enough frequency span!

And when the different "sounds" of 2 different amps match the 2 different "sounds" of the woofer and the tweeter better (to make the sound more homogenic), then the design of the loudspeaker is wrong.

 

[Biblob]

They all 3 deliver just 1/3 of the power, so they all 3 get less hot. And all 3 can have a smaller power supply (or all 3 get their power from the same big one).

And in the details:
-Less intermodulation distortion
-Less thermal drift because of less power for each amp
-Or each amp can have a single set of power transistors/tubes instead of 3 in parallel (because of less heat)

But is it worth the price?

So my thinking was in some way correct .

I don't think it's worth it necessarily, but for me it will because I will use digital crossovers.

 

[Unclevanya]

The view of biamping using two stereo amps that are not identical and feeding them the right and left channels of the same frequency is not ideal. If we use identical amps and route the left channel to the left amp, putting high frequency into one of the left amps two channels (call it LH) and low frequency into the other (call it LL) and the same to the right amp so neither amp is called on to handle the low frequencies of both channels simultaneously - this typically is kinder to the power supply as low frequencies tend to peak hard in both channels and demand current at the same time. While wattage required is likely similar (often because tweeters have resistors to match woofer/cabinet efficiencies) the traditional crossover is less reactive and dips less in the high frequency range. By pairing the high and low frequency of one channel rather than sending the same frequency range for both channels your current required to be delivered by the amp at any given peak tends to be lower. This helps in a stereo design where per channel power supply current is shared. In fully dual mono or true mono amp designs it won't matter much if any but I'd still recommend identical amps to ensure similar performance.

 

[Lbstyling]

So.......

What were saying is that my 1100w RMS of Hypex amp into 112Db/1w speakers might be a little overkill for the 6m x 5m room I have them in?

Oh well. Better safe than sorry

 

[A800]

How much amplifier power is required?

A lot.

 

[Frank Dernie]

So.......

What were saying is that my 1100w RMS of Hypex amp into 112Db/1w speakers might be a little overkill for the 6m x 5m room I have them in?

Oh well. Better safe than sorry

Wow, I thought my 109dB/w horns were about as efficient as it got. I have also used them with a 1000wpc amp but not at the moment.