How Much Current Capacity do we Need from our Amplifiers?

[terryforsythe]

Therefore clearly these 60 A peak output current claims (if these claims were actually made) cannot have any real world relevance and are enitrely bogus.

I think the person claiming the 60A output capability misunderstood the magazine article to which he provided the link. The article states "[t]he output transistors of the NAD 2200 amplifier are high-powered, fast-switching devices capable of delivering some 60 amperes of peak current for brief periods." He just assumed the entire amplifier had that capacity, which does not appear to be the case for a normal speaker load. In that regard, the article further states:

"... If that is not enough, the amplifier can also be operated in a bridged (mono) mode, in which it is rated to deliver up to 400 watts of continuous output into 8 ohms—or, in terms of dynamic power, 1,200 watts into 8 ohms and 1,600 watts into 4 ohms! It is also said to have a wide “dynamic power envelope,” which means that it can maintain these high levels for longer than the standard 20-ms bursts."

Using the 1,600W/4 ohms example, the peak current is 20A for short bursts.

 

[peniku8]

Don't mind me chilling here with my 200A peak current amp.

 

[wwenze]

The problem with engineering is that the entire truth cannot be condensed properly into a single word or few words.

Transistors with 60A output is probably true.

Take a look at this PSU too, rated for 100A at 12V

ASRock SL-1200P 1200 W Review

Compact, Platinum-rated, future-proofed and packing a clever safety feature for your GPU, this PSU goes up against the best power supplies on the market. We tear down the ASRock Steel Legend 1200W to see if it lives up to its name.

www.techpowerup.com

The 8 PWC012N04ES MOSFETs are each rated for 220A tho.

It's never simple, and more transistor rated current is a common practice. NAD 3020 has 2N3055 & MJ2955 rated for 15A continuous collector current. It pumps out 72W at 2 ohms. I²R says 450W, so clearly this isn't the right formula. Even with (I² / 2)*R for sine RMS power that's still 225W.

 

[peniku8]

Using the 1,600W/4 ohms example, the peak current is 20A for short bursts.

The peak current of the sinusoidal signal in your example will be 20A*sqrt(2)=28A
The 20A you calculated is the rms current.

 

[Ropeburn]

The problem with engineering is that the entire truth cannot be condensed properly into a single word or few words.

Transistors with 60A output is probably true.

Take a look at this PSU too, rated for 100A at 12V

ASRock SL-1200P 1200 W Review

Compact, Platinum-rated, future-proofed and packing a clever safety feature for your GPU, this PSU goes up against the best power supplies on the market. We tear down the ASRock Steel Legend 1200W to see if it lives up to its name.

www.techpowerup.com

The 8 PWC012N04ES MOSFETs are each rated for 220A tho.
View attachment 540196

It's never simple, and more transistor rated current is a common practice. NAD 3020 has 2N3055 & MJ2955 rated for 15A continuous collector current. It pumps out 72W at 2 ohms. I²R says 450W, so clearly this isn't the right formula. Even with (I² / 2)*R for sine RMS power that's still 225W.

Or it kinda is the right formula because you'll never want to have your parts actually reaching their maximum rating. It just isn't good for longevity so you're going to overbuild it a little - or a little more. Always a good thing if it's within budget. These chonky AB amps last as long as they do for that reason.

 

[terryforsythe]

The peak current of the sinusoidal signal in your example will be 20A*sqrt(2)=28A
The 20A you calculated is the rms current.

"Peak" was the incorrect term for me to use. Maybe "dynamic" current would have been better. But, yes, I calculated it as RMS.

 

[EJ3]

I think that I will help muck up things by (using an amp dyno) showing what amount of watts a NAD 2200 can do in 8, 4 and 2 ohms. so that the calculations can have an accurate example of an old NAD 2200 to calculate from:

 

[solderdude]

The fun part of power envelope is that an amp can (for short duration peaks only) deliver a lot more power than what it can do continuous.
It works because it basically has 2 internal voltage rails (+/- 62 and +/-95V) but the higher one is only 'used' above a certain output voltage and can only deliver power during a short moment as that collapses quickly only to recharge when the peak is gone again.
Basically a class-G design but not with a fixed level high-voltage rail as the higher rail voltage is only available for short moments.
The output stage of the 2200 does have emitter resistors but no current limiter, instead when over current is detected the output relay protection is triggered (which is more sluggish than a limiter but at least it disconnects the load. For very short duration (before the output is switched off) the max output current is determined by the power supply.
The earlier power envelope models had a tendency to blow their output devices when driven to high for longer periods (thermal runaway due to the absence of emitter resistors).

As music consists of peaks this type of class G works quite well.

Carver used class H in their PM1.5 which also had 2 power rails but +/- 50V and +/-125V but had a modulated 125V rail and had an actual current limiter and input limiter (compressor).

Now we have class D which switches between power rails which takes advantage of the max. output current of an output device when 'on' and max voltage rating while 'off'.
In class A, AB, G and H one has to respect the output Power Derating Curves (Safe Operating Area) of output devices, which is also temp dependent, so one never can utilize the 'switched max current' rating of such devices.
This is because when the output voltage is not reaching the voltage rail yet (which it basically never does otherwise the amp would be clipping) but there is a lot of current flowing the output devices have to dissipate a lot of power (heat), hence the big cooling fins and power transformers.

 

[pjug]

But the question of how much current a normal speaker will draw remains and is a valid one, especially when the signal is not a simple sine wave. Let's see if we can answer that.
.....
View attachment 539929

Thanks for doing this and it is very interesting. But just doing the quick estimate I wouldn't be using the dip at ~5KHz, I would use the broad dip around 200Hz where it goes to about 7 ohms or a little more. So then that isn't too far from the peak V/I ratio of 7.8 value that you calculate.

 

[Jazigo]

Could we maybe organize a practical test with an ammeter with a sampling rates of 1 MS/s? That should be the definitive answer for all parties.

 

[Miercoles2022]

PV = nRT

 

[NTK]

PV = nRT

Useful equation when the magic smoke is released

 

[Ropeburn]

Useful equation when the magic smoke is released

Reminds me of PMPO calculations which are very easy. Long ago when that became a thing with plastic PC speakers and such, dad and I figured it out quickly: PMPO = inrush current times mains voltage.

 

[SSS]

Could we maybe organize a practical test with an ammeter with a sampling rates of 1 MS/s? That should be the definitive answer for all parties.

Could be easily done with a digital oscilloscope but not necessary. Maximum amperes can be calculated from the schematic of the loudspeaker crossover values. For this case think all caps are not loaded. So they will draw max current at turn on say with the front of a rectangle pulse. For this set the value of all capacitors in the circuit with zero ohms (a short). Then the resistance the amplifier will see can be calculated using the resistor values of the network. Because in a regular crossover there is no capacitor directly across at the speaker inputs there is no short. If it were then max peak amps are amplifier rail voltage divided by the output internal resistance.

 

[john61ct]

When "max peak amps" is used in this context

is there safety / linearity included?

Or is that more a voltage thing? Or SPL?

I have no interest in driving my LS50s past the point of clean SQ output, much less near any danger-of-damage zone.

However, I do want to push up close to maximum SPL within that context

 

[SSS]

Don't test your loudspeakers for max SPL with sine or squarewave or similar. They are not built for this. I think M-sound is probably the best method but needs some equipment. Without special gear one could use a SPL meter with M-sound and slowly increase volume until the SPL does not follow linearly to the amplifiers output. But still may be dangerous for the speakers.

 

[popej]

In the case of the Stereophile simulated load, the minimum impedance occurs at ~4 kHz, a frequency range which the M-Noise (and most music/audio signals) does not contain a lot of energy. The peak voltage to peak current ratio depends heavily on the spectral energy distribution of the signal, and the speaker impedance to frequency characteristics. If a speaker has low impedance for a substantial portion of the low frequencies and the signal has a lot of energy in the low frequencies, this ratio could potentially be significantly lower and be closer to the minimum speaker impedance, meaning a higher demand for current from the amplifier.

That is a great method to estimate needed current. Thanks!

Your speaker model is a bit uncommon. I think usually the lowest impedance is at much lower frequency, probably somewhere around 100-400Hz. You will probably get similar current with sinus sweep and with M-noise. Which actually would be quite convenient result for any quick estimation.

 

[john61ct]

Don't test your loudspeakers for max SPL with sine or squarewave or similar. They are not built for this. I think M-sound is probably the best

So I'm guessing you mean "M-Noise", aka AES75 https://m-noise.org/procedure
??

...

and how does one tell when "SPL does not follow linearly to the amplifiers output" ??

Ears + brain judgment ? Or can be done objectively in REW?

...

For an SPL meter, besides REW + XLR measurement mic

I plan to use UMIK-2 + phone app
AudioTool or Room Acoustics Meter

Have UMIK-2 for measurements - ISO reco's Android app to record SPL accurately

Looking for app recommendations to measure speakers' SPL as accurately as possible from Android using a UMIK-2 Apparently for this use case, integrating the calibration file may not be so important? Would prefer an app that will allow logging timestamped results, ideally with user notes. NOT...

www.audiosciencereview.com

 

[NTK]

When "max peak amps" is used in this context

is there safety / linearity included?

Or is that more a voltage thing? Or SPL?

I have no interest in driving my LS50s past the point of clean SQ output, much less near any danger-of-damage zone.

However, I do want to push up close to maximum SPL within that context

Speakers respond to voltage. The speaker sensitivity gives you the relationship between amplifier output voltage and its SPL output. For example, the original LS50 was about 83-84 dBSPL @ 2.83 V for 1 m free space. Ten times the voltage should increase output SPL by 20 dB (assuming that the speaker is still behaving linearly). Current is just what it is based on Ohm's law and the speaker impedance.

I also want to apologize for not providing some info on the "pre-requisites" to fully understand the math in the opening post. You'll probably need to have taken classes equivalent to senior level undergraduate control systems courses to have been exposed to these methods. My purpose was to show we can predict quite well the current required to drive a speaker, and it is not infinite. Marketing speak such as an amplifier can output many tens of amps of current has no real world relevance, and worse if it was only the output device spec and not the finished amplifier spec.

 

[Dialectic]

Wait, you mean amplifiers don't have to "double down" all the way to 0.02 ohms to be proper, manly amplifiers?