Watt for speaker and Watt for power consumption, Class D and AB

[antcollinet]

Nah, there's something called CLAMP MULTIMETER, you just put the cable inside the circle and it will read the current.
No touching involved.

View attachment 160683

Unlikely to have any accuracy at the levels you are measuring though, unless you get one with a 1A and 5A range (or in that area)

 

[kchap]

Just measuring realtime voltage sent to speaker by amplifier using normal volume we're comfortable with. You shld try and share your Aiyima A07 with whatever adapter you're using so we can compare our numbers. The same variable we all have would be Aiyima A07 if you have it. I know Raindog123 has it from his posts, he even bought two iirc.

BTW I have a kwh meter and it shows that it only uses 0.06A with Aiyima A07+48v5A and the watt is 6.4Watt (3.2w/ch?) , this thing is for power consumption though, but I can probably assume using it as watt coming out from my speakers. I tried using hairdryer and it jumped to 196watt, so this kwh meter is not wrong.

I have given the wrong impression. I do not own an Aiyima A07. If you were reading 0.06A at 120V, that's 7.2W. There would be some losses and maybe a bit of phase shift. 3.2W per channel sounds reasonable. Remember that is the average. Your meter will average it's measurements. There could be lots of peaks at 10 or 20W but the might only last 100ms or so.

 

[pjug]

Measuring with a multimeter isn't going to tell you what the peaks are, which is what is most important. Also, the meter probably is only responding to low frequency with any kind of accuracy. You might be able to very roughly estimate peaks from the average, use something like 20dB difference (or 10X voltage) between peaks and average. But this will be different with different music. If you are measuring something like 2V RMS playing as loudly as you want, then the 48V/5A supply should be perfectly fine.

If you really want to find the voltage peaks, you need to use an oscilloscope. But this is a bridged amp so be careful with that. You would need to measure A-B with an oscilloscope in that case, and make sure the ground clip on the probe does not connect with the amp/speaker connections.

I would say have fun with the meter and learn what you can, but at the same time don't worry about whether the amp and power supply give enough power. It sounds like you are good with what you have.

 

[antcollinet]

Measuring with a multimeter isn't going to tell you what the peaks are, which is what is most important. Also, the meter probably is only responding to low frequency with any kind of accuracy. You might be able to very roughly estimate peaks from the average, use something like 20dB difference (or 10X voltage) between peaks and average. But this will be different with different music. If you are measuring something like 2V RMS playing as loudly as you want, then the 48V/5A supply should be perfectly fine.

If you really want to find the voltage peaks, you need to use an oscilloscope. But this is a bridged amp so be careful with that. You would need to measure A-B with an oscilloscope in that case, and make sure the ground clip on the probe does not connect with the amp/speaker connections.

I would say have fun with the meter and learn what you can, but at the same time don't worry about whether the amp and power supply give enough power. It sounds like you are good with what you have.

He is sending sine waves. RMS values are fine.

 

[pjug]

He is sending sine waves. RMS values are fine.

Oh, I see. I don't know if you can get a good feeling of max comfortable level this way, but otherwise I'd still say no worries with the 48V/5A supply.

Sorry if this is what OP is already doing, but the best way I know to use a multimeter to do this is with the Pano method, described by Archimago here:

MUSINGS: How much amplifier power do you really "need"? Why not test for yourself?

A blog for audiophiles about more objective topics. Measurements of audio gear. Reasonable, realistic, no snakeoil assessment of sound, and equipment.

archimago.blogspot.com

 

[jst]

But this will be different with different music. If you are measuring something like 2V RMS playing as loudly as you want, then the 48V/5A supply should be perfectly fine.

I would say have fun with the meter and learn what you can, but at the same time don't worry about whether the amp and power supply give enough power. It sounds like you are good with what you have.

Exactly what I thought, I should be alright even with 32v 5A.

At least all this dicussion will save me money from buying Meanwell LRS-350-48 (48v 7.3A).

Unlikely to have any accuracy at the levels you are measuring though, unless you get one with a 1A and 5A range (or in that area)

The point is, it doesn't involve short circuiting anything like you thought it would, and not even gonna happen even with probes.

 

[antcollinet]

Exactly what I thought, I should be alright even with 32v 5A.

At least all this dicussion will save me money from buying Meanwell LRS-350-48 (48v 7.3A).



The point is, it doesn't involve short circuiting anything like you thought it would, and not even gonna happen even with probes.

I assumed you were using the multimeter for current measurement.

You are correct - clamp meter used like that is perfectly safe.

 

[jst]

Remember that is the average. Your meter will average it's measurements. There could be lots of peaks at 10 or 20W but the might only last 100ms or so.

This kwh meter has Hi and Lo too, and you're probably right, the Hi says 68watt even though I didn't change volume during listening session, probably because of any of the random source of audio I played, but it did capture a 68 watt for Highest watt.. I was using CA Azur that's average at 25-27w. Probably would happen too if I use Aiyima that's averaged at 6-7w with peaks you're guessing, this is surprisingly low power, dont know whether it's good or bad, but it is what it is and as long as it sounds good then let physics do what they need lol.

 

[kchap]

This kwh meter has Hi and Lo too, and you're probably right, the Hi says 68watt even though I didn't change volume during listening session, probably because of any of the random source of audio I played, but it did capture a 68 watt for Highest watt.. I was using CA Azur that's average at 25-27w. Probably would happen too if I use Aiyima that's averaged at 6-7w, this is surprisingly low power, dont know whether it's good or bad, but it is what it is and as long as it sounds good then let physics do what they need lol.

The low power is a good example of the efficiency of class D. Most of energy is dissipated in the speakers, where you want it to go. According to the manual, the Cambridge is doing better than expected. It could draw up to 35W with no signal connected.
Watt for speaker and Watt for power consumption, Class D and AB # 3

 

[DonH56]

There are a lot of "interesting" numbers thrown around in this thread... Not worth trying to correct at this point.

Most people listen at power levels far below what they expect. If you have 87 dB/W/m speakers 1 m in front of you and put 1 W into them, then 87 dB is very, very loud. 1 W into 8 ohms is about 3 Vrms, and 2 Vrms into 4 ohms.

You should also do sine-wave testing in very short bursts, just long enough to read the meter, and let the speakers cool for a minute or two between tests although at these levels and frequencies the speaker can probably survive all day.

Also be aware that most inexpensive meters have limited frequency range (though 50 Hz should be fine) and do not measure true RMS values (though probably good enough with a sine wave).

 

[rdenney]

I assumed you were using the multimeter for current measurement.

You are correct - clamp meter used like that is perfectly safe.

But a clamp meter should only go over one speaker wire, not the pair of speaker wires. The current going out and returning cancel each other if you clamp both together.

Rick "but bandwidth will be optimized for 50-60 Hz for those kinds of meters" Denney

 

[jst]

Most people listen at power levels far below what they expect. If you have 87 dB/W/m speakers 1 m in front of you and put 1 W into them, then 87 dB is very, very loud. 1 W into 8 ohms is about 3 Vrms, and 2 Vrms into 4 ohms.

Agree. I remember this vid.

 

[Raindog123]

@jst So, as I've promised. Here are my quick measurements to compare with yours (Aiyima A07 @ "9 o'clock" into Wharfedale Diamond 8.4):

1) Speaker DC resistance =3.6 Ohm (spec's "nominal impedance" is 6 Ohm);

2) 1kHz (-10dB test tone): 0.72V and 25.7mA --> ~0.02W (and 28.1Ohm calculated resistive impedance);

3) 100Hz test tone: 0.70V and 50.6mA --> ~0.035W (35mW) (and 14Ohm impedance).

4) Actual music (an Accept heavy metal track): peaking to ~0.7V and 0.4A = ~0.3W (again, this is on a slow multimeter).

 

[Raindog123]

@jst Now...

1) As many here point out - and possibly eye-opening for some - this is a [rough] measurement of the actual electrical power "used/consumed" by a speaker to produce not-overly-loud sound (80-90dB) at a short distance (~1m) from a consumer speaker (~90dB sensitivity): from a few tens of mW for test tones to under 0.5W for "real" music.

2) However, (1) this "consumed" power does not equate to the amount of power needed to be produced/output by an amplifier. This is due to a highly reactive (non-resistive) component of the speaker load - reflected by a high V-I phase shift - up to 40° at some frequencies, eg, see here. If there is a phase between voltage and current, only a portion of the signal gets used by the speaker and the rest is "reflected" back to the amplifier, eg, see a good discussion here. So, to provide the amount of power from (1) to a speaker, an amplifier has to offer the "power factor" [= 1/cos(ϕ)] more power!

3) In a transistor amplifier (eg Class AB), this returned power gets dissipated into heat. In our case of a Class D amplifier, this power can cause all kind of issues (like raising rail voltage, see "bus pumping" in the article above). [With a proper design - the BTL bridge mode and large capacitors to absorb returned voltage/power - this can be mitigated. My understanding, Aiyima does it, to a degree.] Still, to provide the needed "acoustic" power to the speaker, the amplifier has to generate the "power factor" more power. [And this is not even accounting for the "dynamics” and “transients" - instantaneous high-power peaks - in the music itself.

Bottom line, as what you really after is just an estimate of your power supply [DC] current I, I would recommend measuring it in your existing setup. (Knowing the power supply VDC, will also allow you to calculate your consumed DC power, if you are interested. And remember, it feeds both channels.) My bet, you'll find out that for your listening rig, this current is very low (<<1A)! So, any power supply on the market will do its job fine!

 

[Raindog123]

Lastly, here are the two Impedance and Phase plots measured by @amirm for two Wharfedale's. While unfortunately neither your nor my model, they still illustrate well (a) dependence of the impedance and phase on frequency and (b) strong variance with the speaker model/design:

 

[jst]

@jst So, as I've promised. Here are my quick measurements to compare with yours (Aiyima A07 @ "9 o'clock" into Wharfedale Diamond 8.4):

1) Speaker DC resistance =3.6 Ohm (spec's "nominal impedance" is 6 Ohm);

2) 1kHz (-10dB test tone): 0.72V and 25.7mA --> ~0.02W (and 28.1Ohm calculated resistive impedance);

3) 100Hz test tone: 0.70V and 50.6mA --> ~0.035W (35mW) (and 14Ohm impedance).

4) Actual music (a heavy metal Accept song): peaking to ~0.7V and 0.4A = ~0.3W (again, this is on a slow multimeter).

1. I just found out that impedance and resistance can be different, impedance is AC while resistance can be both, that's why when measuring using multimeter it's different from impedance spec. (or cmiiw)

So my multimeter isn't faulty, it's just that at that distance and volume, that kind of "power" is enough to produce loudness, and don't know why Aiyima power output is kinda low.

I wonder how loud amirm did his measurements to produce 44w, 61w, 77w.

Here's my power consumption using kwh meter tool and aiyima a07 with 32v5a.
Average watt (this is 6.4w using 48v5a, so adapter does matter)


Highest watt (probably occassionally)


Lowest watt (probably when just powered on or idle)


Current


For comparison , testing a hairdryer at home it used 196 watt written as average. This Aiyima power consumption is so low, and it's comparable with the low power on the speaker even though it's quite loud.

 

[pjug]

I wonder how loud amirm did his measurements to produce 44w, 61w, 77w.

Fortunately for Amir's ears, those measurements are into a resistor.

 

[jst]

@jst Now...

1) As many here point out - and possibly eye-opening for some - this is the [rough] measurement of the actual electrical power "used/consumed" by a speaker to produce not-overly-loud sound (80-90dB) at a short distance (~1m) from a consumer speaker (~90dB sensitivity): from a few tens of mW for test tones to under 0.5W for music.

2) However, (1) this "consumed" power does not equates to the amount of power needed to be produced/output by an amplifier. This is due to a highly reactive (non-resistive) nature of the speaker load - reflected by a high V-I phase - up to 40° at some frequencies, eg, here. If there is a phase between voltage and current, only a portion of the signal gets used by the speaker and the rest is "reflected" back to the amplifier, eg, see a good discussion here. So, to provide the amount of power from (1) to a speaker, an amplifier has to produce "power factor" = 1/cos(ϕ) more power!

3) In a transistor amplifier (eg Class AB), this returned power gets dissipated into heat. In our case of a Class D amplifier, this power can cause all kind of issues (like raising rail voltage, see "bus pumping" in the article above). [With a proper design - the BTL bridge mode and large capacitors to absorb returned voltage/power. My understanding, Aiyima does it, to a degree.] Still, to provide the needed "acoustic" power to the speaker, the amplifier has to generate "power factor" more power. And this is not even accounting for the "dynamics" - instantaneous high-power peaks - in the music itself.

Bottom line, as what you really after is just an estimate of your power supply [DC] current I, I would recommend measuring it in your existing setup. (Knowing the power supply VDC, will also allow you to calculate your consumed DC power, if you are interested. And remember, it feeds both channels.) My bet, you'll find out that for your listening rig, this current is very low (<<1A)! So, any power supply on the market will do its job fine!

Yea I'm surprised, everything I thought I knew about amplifier, speaker and watt suddenly are different from reality lol. (not that I know a lot, it's just my perception from what I thought I knew).

I think looking at my kwh meter is enough, it only uses like 3.8watt using 32v5a, and 6.4watt using 48v5a. There's heat, there's efficiency, etc, divided by 2 per ch, so the watt that goes into a speaker is just low enough to know.

Makes me wonder how loud amirm used when doing measurement since his result about watt was quite high. (44w,61w,77w etc)

 

[jst]

Fortunately for Amir's ears, those measurements are into a resistor.

You mean, not actually using speaker out ? just for the sake of measurement.

 

[DonH56]

Impedance comprises real (resistive) and imaginary (reactive) components. Z = R + jX where R is resistance, X is reactance (capacitance or inductance), and j = sqrt(-1) -- mathematicians use i instead of j, but engineers use i for current so we use j.

At DC the impedance of an ideal inductor is 0 ohms and an ideal capacitor's impedance is infinite. AT infinte frequency, the inductor's impedance is infinite and capacitor's impedance is 0. In the real world there are parasitic (non-ideal) components so you don't quite get 0 or infinity.

I have not looked at this but here is Wikipedia's take: https://en.wikipedia.org/wiki/Electrical_impedance

HTH - Don