What is a good way to measure peak watts I'm using during music?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi Guys

Since you know the rated sensitivity of your speaker at 89db/1W at 1m, then your listening level of 80db is almost 10db less, so about 100mW.

On a system I had a long time ago, I measured the SPL in the middle of the room, about where I would be for most listening. A set the power to each speaker to 1W approximately using voltage as a guide, so really 2V8 average, then 2W roughly into the room. SPL was 105db and louder than I'd want for continuous listening. Needless to say, no one was home when I did this test!

Personally I find 90db to be a comfortable loudness for loud listening, but tend to mostly listen at 70-80db. This means the amount of power needed is very low, as my current system has 90db/1W at 1m rating.

Dynamic range is a tricky issue. Most pop music exhibits practically zero dynamic range - it is either full-tilt or off. Certainly most modern metal is like this. When you watch the scope while listening to music it is surprising or not to see that the average level stays about the same even as the music seems to swell or open up as instruments come and go. This is mostly a matter of modern production values, where level is maintained by compression and the coming and going of voices and instruments is merely a textural change. Older recordings exhibit real dynamic range that tests your system.

The idea that you can reproduce the dynamic range of a live performance in your living room is ridiculous. That you would even want to is also ridiculous. Just as larger picture screens are intended only to allow more people to sit in front of them while all sitting farther away as the screen size is increased, large concert halls are designed for many people to be present for a musical performance. By rights, the area closest to the stage should be left empty so everyone hears about the same loudness and image of sound. To fill this space requires extra power, either as more musicians per section of an acoustic band, or by a louder PA for electronic bands. The fact an orchestra can achieve 120db of SPL does not make the dynamics useful nor does it make that loudness capability useful. For most listeners this is a painful loudness. Reprdoucing aural torture is not something most people fancy.

Dynamic peaks of 120db are harmful enough and they cause your hearing to close up to protect itself. This makes hearing the quietest passages more difficult. So, if the band is toying with "complete" dynamic range as Ritchie Blackmore did in the Rainbow On Stage live album, the quietest parts will be difficult to hear because your hearing is already adjusted to the louder average. Even though that album was released on vinyl with its limited dynamic range, the range present was very dramatic on a human scale - more than most people are aware of since most people listen to much more compressed music most of the time.

The idea of having sound in excess of 120db should be criminal. The sound-offs of car audio systems where they deafen young girls with 140dB sound waves is pretty much in the same class of entertainment as blood sports and public hangings. Anyone who wants to have a system with 120dB dynamic range can have it but they won't be able to listen to it for very long without going deaf.

Most music has less than 25db dynamic range. The lower limit for playback is room noise. The upper limit is where your own hearing starts to protect itself and this should not be the normal arbiter for how you set the loudness in your room.

Keep loudness levels sane. YOU have the power to retain your hearing for a lifetime of music enjoyment.

Have fun
Kevin O'Connor
londonpower.com
 
This reply completely ignores current clipping.
As digital audio data are limited to a maximal value called Full Scale Deviation, it is easy for a competent person to limit the signal in a reproduction system. It only requires to set the volume beyond the point where the signal can clipped in the forward path (the power amp).

A multimeter with +/-peak voltage memory is very useful tool to evaluate precisely what happens., I own four of such multimeters, I just sold one Multimètre METRIX MX57EX Bricolage Vienne - leboncoin.fr

Many power supplies or amps have a feature to react to (calculated) instantaneous peak powers, however they usually incorporate a bass pass filter to make the reaction by limiting current not to happen too fast. Once again, we see time appearing when considering the effect of electrical power.
 
Place a small, low tolorance (acurate) sensing resistor in series with the speaker and measure the load across the resistor with the scope. Use Ohms Law to figure how much current is flowing through the speaker and at what phase. Also since the angle of the Z of the speaker is not the same for all frequencies, the voltage waveform across the resistor will vary acordingly to be different than that of the output voltage at the output node. Reactive power has to be considered.
 
The idea that you can reproduce the dynamic range of a live performance in your living room is ridiculous.
I have listened to a live performance inside a room. I sure the majority have also listened likewise.

What's so ridiculous that we set ourselves a target to try to reproduce that live performance with our equipment.
It's been done once (live), all we want to do is repeat it.
 
Place a small, low tolorance (acurate) sensing resistor in series with the speaker and measure the load across the resistor with the scope. Use Ohms Law to figure how much current is flowing through the speaker and at what phase. Also since the angle of the Z of the speaker is not the same for all frequencies, the voltage waveform across the resistor will vary acordingly to be different than that of the output voltage at the output node. Reactive power has to be considered.

I wonder at the meaning of instantaneous power values obtained. I think they become of interest only if integrated.

I've already mesured some peak voltages and currents applied to loudspeakers, instantaneous voltage drops of amp power supplies and voice coil heatings. All very interesting results, I may publish some one day.
 
Place a small, low tolorance (acurate) sensing resistor in series with the speaker and measure the load across the resistor with the scope. Use Ohms Law to figure how much current is flowing through the speaker and at what phase. Also since the angle of the Z of the speaker is not the same for all frequencies, the voltage waveform across the resistor will vary acordingly to be different than that of the output voltage at the output node. Reactive power has to be considered.

I'd use the current sensing resistor (maybe even the resistance of the speaker wire would work here, assuming you can determine its value accurately) and an appropriately scaled voltage divider input to a four quadrant analog multiplier like the AD633. The output would be a voltage (or current) that is proportional to the power being delivered. Use some sort of peak detector (fast attack, slow release) to make useful measurements. Integrating the output should show what the average looks like too.
 
Hi Guys

I'll never understand the human propensity to take a simple issue and complicate it. I suppose it comes down to everyone trying to demonstrate their worth to the village.

If you look at the name of this thread, the question is simple: how does one measure peak power they are actually using with music?

The answer has been stated pretty plainly a few times.

The ideal is to have a scope to look at the transients of the music. Peak voltage is what matters as transients are short peaks. Handily the scope shows instantaneous voltage. Simple ohms law tells you resistive power for the peak. If you want to make an educated guess at reactive power, you can use rules of thumb based on how reactive typical speakers are, you can find out how reactive your speaker system is, you can use a worst-case guess. None of it matters all that much at the power levels in question.

If you only have a slow-responding meter, then you have to use rules of thumb to guess at the peak using the average value the meter shows. Again you are calculating resistive power, but again, who cares? The ratio of peak to average for music falls within a very narrow window so whether it is reactive or resistive power calculated or measured makes no difference.

The _used_ power is in the range of 100mW to 1W as previous posts suggest. Peaks will be 4x this value for pop music but could be 8x for classical. The poster never said what kind of music he listens to or even if it dynamic. The implication is that there is 'some' dynamics since the peak power is in question. Because the power level is so low, even an amp designed only for 8R with no current reserve for reactive dips will have ample current for this output level provided the amp is rated for 10W+.

Production amps designed to drive difficult loads typically have a minimum of 2x up to 5x peak current reserve. So, a 100W-8R amp has to put out 5Ap for the basic load, but might actually provide 10-50Ap. This accommodates reactive loads greater than 60-degrees, which few loads ever approach.

At 1W into 8R, current is only about 500mA peak. If the load was severely reactive and needed 2Ap, then an 8R amp rated for 16W+ would drive it well provided its protection circuit is correctly designed.

The original Q was about transient power - peak power - not average power.

When one designs an amp - particularly a solid-state amp - you use peak power, current and voltage to set basic values like power supply voltage and current capability. Average and RMS values come into play for selecting the PT and the filter caps. The rest falls into place after that.

Have fun
Kevin O'Connor
londonpower.com
 
People usually complicate because they have the need to demonstrate their skills

about calculation...it is not a matter of answer the one made the question..they are answering their own needs to show themselves under the spotlight...human things.

I am glad i am an extra terrestrian from a red planet:(

regards,

Carlos
 

Attachments

  • September, 18th, 2011.jpg
    September, 18th, 2011.jpg
    72.3 KB · Views: 64
Last edited:
Here's how. all you have to do is have a 4GS/s scope with 8Mb sample memory, an isolated voltage probe with 100MHz bandwidth (in case its a bridged load - you can probably just use an ordinary voltage probe) and a DC-50MHz current probe.

Measure the voltage V(t) and current i(t) thusly, then use scope maths to multiply the two channels together giving P(t). Use scope maths to calculate the mean power, and to find peak power put the scope on infinite persistence and run for however long.

the scope is pretty easy - it looks like all the tek-like DSOs now have 1GS/s sampling and plenty of sample memory (I have LeCroy scopes). The real problem is the current probes - make sure you are sitting down before checking out the price. Dont be surprised if a decent current probe costs twice as much as your scope (my ones did).

But there is a low-cost alternative - use a LEM DCCT. you have to feed a wire through the hole, and make up a small interface board to supply them with power, but you can get decent LEM DCCTs for $50 or so new, and cheaper on ebay.

Or the cheapest approach is to use a current sensing resistor - downside is SNR, you need Rcs << Rload, and this can result in a fairly small signal. whereas the LEM DCCT will give +/-12V FS.

attached is an instantaneous & average power measurement of the output of a 100W PFC smps. this topology has quite a bit of 100Hz ripple in the output, so I had to measure instantaneous power Pout(t) and then average it over an integer number of AC line cycles - a pair of DMMs gave garbage results.
 

Attachments

  • smps power measurement.jpg
    smps power measurement.jpg
    248.1 KB · Views: 58
Hi Guys

I'll never understand the human propensity to take a simple issue and complicate it. I suppose it comes down to everyone trying to demonstrate their worth to the village.

If you look at the name of this thread, the question is simple: how does one measure peak power they are actually using with music?

The answer has been stated pretty plainly a few times.

The ideal is to have a scope to look at the transients of the music. Peak voltage is what matters as transients are short peaks. Handily the scope shows instantaneous voltage. Simple ohms law tells you resistive power for the peak. If you want to make an educated guess at reactive power, you can use rules of thumb based on how reactive typical speakers are, you can find out how reactive your speaker system is, you can use a worst-case guess. None of it matters all that much at the power levels in question.

If you only have a slow-responding meter, then you have to use rules of thumb to guess at the peak using the average value the meter shows. Again you are calculating resistive power, but again, who cares? The ratio of peak to average for music falls within a very narrow window so whether it is reactive or resistive power calculated or measured makes no difference.

The _used_ power is in the range of 100mW to 1W as previous posts suggest. Peaks will be 4x this value for pop music but could be 8x for classical. The poster never said what kind of music he listens to or even if it dynamic. The implication is that there is 'some' dynamics since the peak power is in question. Because the power level is so low, even an amp designed only for 8R with no current reserve for reactive dips will have ample current for this output level provided the amp is rated for 10W+.

Production amps designed to drive difficult loads typically have a minimum of 2x up to 5x peak current reserve. So, a 100W-8R amp has to put out 5Ap for the basic load, but might actually provide 10-50Ap. This accommodates reactive loads greater than 60-degrees, which few loads ever approach.

At 1W into 8R, current is only about 500mA peak. If the load was severely reactive and needed 2Ap, then an 8R amp rated for 16W+ would drive it well provided its protection circuit is correctly designed.

The original Q was about transient power - peak power - not average power.

When one designs an amp - particularly a solid-state amp - you use peak power, current and voltage to set basic values like power supply voltage and current capability. Average and RMS values come into play for selecting the PT and the filter caps. The rest falls into place after that.

Have fun
Kevin O'Connor
londonpower.com

+1
...as post #15 :)
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.