Thanks Revb.
Not disgruntled, just a bit frustrated. You are right about people reading their own biases into it, but I had not expected it.Yes, I'm glad that Tom is here, he has a lot of good advice. Some of the things that have come up deserve their own threads.
Ah I've been assuming that as the original sine wave was recorded at -12dB as an average level that the output on the amplifier, when measured in an rms voltage was also representative of the rms figure. Hence if you multiplied it by 4 you'd end up with the equivalent requirement for an rms wattage. I may have added to some slight confusion in some of my posts.
Going through the maths though its obvious as the peaks of the -12dB tone actually hit -9dB and if multiplied by 4 we end up with +3 which is impossible.
The -12dB tone might be recorded as an rms value and the DMM might read an rms voltage, and indeed to figure out how much peak voltage you need to reproduce the -9dB peaks you need to multiply it by 1.41. Now you're up at a peak level = -9dB, so to raise that up to a 0dB peak you actually need to multiply 1.41 up by 2.82 = 4 volts. That is 4 volts to reproduce the peaks of a sine wave with peaks hitting 0dB.
Naturally if one starts out with an rms figure of 1 volt for the -12dB, you multiply it up by 4 to arrive at 4 volts, which as explained above = the peak value of a maximum amplitude sine wave. I had assumed that this was the rms value and hence you can convert it directly into an 8 ohm watt figure. You cannot, to do that you have to divide it by 1.41 first.
In other words, if I've just figured this out correctly
o) the people who are in category 1 and measuring less then 2 volts are actually using no more then 4 watts, and those in category 2 are using between 4 and 25 watts rms, as measured into a resistive 8 ohm load.If that's correct then it appears that I inadvertently added to some of the confusion.
Hi mr push pull
The reason for the null test is it would show you if your amplifier has run out of voltage swing, is large enough for what your listening to.
Usually clip indicators and normally level indicators only show the level as averaged over a longer period of time compared to what is happening cycle by cycle like I am talking about.
Thus, as one might find with an oscilloscope, an amplifier may be instantaneously clipping while showing no signs of that on a level meter or clip indicator or be audible as a familiar flaw (an example of it I cited earlier).
I suggested this test because well before clipping makes a sound that you recognize, it is compressing the dynamics of the signal. If preserving the dynamics is important or desirable, having "enough" is also desirable.
With that null test, if the amp input and output are the same, there is nothing left to hear and no problem. AS soon as an amplifier misbehaves and cannot follow the input signal (like instantaneous clipping) a large difference signal is produced.
The presence of that signal (your amp running out of swing or distorting etc) tells you that for what you were trying to do, your amp is not big enough.
The null test would work with any recording at any level.
5th, sorry you were having trouble with it, the program has been out for a good while.
In the way old days, one could do this with a pair of mic or signal transformers, the input drives one, an attenuator made of a 10k resistor and 2k pot could be used to make it adjustable connected to the output, the out puts are put in serried but out of phase to cancel.
You put in a sine wave or other signal at a low level and adjusted the pot to get the minimum level, the null.
Then you could amplify the difference and do the same thing, instantly “hear” when it isn’t following the input signal.
This old way works real time but I don’t imagine many folks have a pair of mic or signal transformers lying around and I figured software might be easier.
The cool part about the old way is it doesn’t care about ground convention and allows you to see or rather hear real time when you have run out of gas well before you would be able to hear it from your speakers and it works with any signal at any level with any program material..
So, if one considers the instant an amplifier is no longer able to do what it’s told is an indicator it’s not powerful enough or otherwise unable then this seems like a good bet and doesn't require an oscilloscope.
Best,
Tom
The reason for the null test is it would show you if your amplifier has run out of voltage swing, is large enough for what your listening to.
Usually clip indicators and normally level indicators only show the level as averaged over a longer period of time compared to what is happening cycle by cycle like I am talking about.
Thus, as one might find with an oscilloscope, an amplifier may be instantaneously clipping while showing no signs of that on a level meter or clip indicator or be audible as a familiar flaw (an example of it I cited earlier).
I suggested this test because well before clipping makes a sound that you recognize, it is compressing the dynamics of the signal. If preserving the dynamics is important or desirable, having "enough" is also desirable.
With that null test, if the amp input and output are the same, there is nothing left to hear and no problem. AS soon as an amplifier misbehaves and cannot follow the input signal (like instantaneous clipping) a large difference signal is produced.
The presence of that signal (your amp running out of swing or distorting etc) tells you that for what you were trying to do, your amp is not big enough.
The null test would work with any recording at any level.
5th, sorry you were having trouble with it, the program has been out for a good while.
In the way old days, one could do this with a pair of mic or signal transformers, the input drives one, an attenuator made of a 10k resistor and 2k pot could be used to make it adjustable connected to the output, the out puts are put in serried but out of phase to cancel.
You put in a sine wave or other signal at a low level and adjusted the pot to get the minimum level, the null.
Then you could amplify the difference and do the same thing, instantly “hear” when it isn’t following the input signal.
This old way works real time but I don’t imagine many folks have a pair of mic or signal transformers lying around and I figured software might be easier.
The cool part about the old way is it doesn’t care about ground convention and allows you to see or rather hear real time when you have run out of gas well before you would be able to hear it from your speakers and it works with any signal at any level with any program material..
So, if one considers the instant an amplifier is no longer able to do what it’s told is an indicator it’s not powerful enough or otherwise unable then this seems like a good bet and doesn't require an oscilloscope.
Best,
Tom
Hi Tom,
if only determining if clipping occurs is what you're after, simply using the sound board as a scope may be even simpler. one can even record and do post-analysis. not to mention real-time analysis.
an advantage of recording real music through the sound card instead of using sines is that one can capture behavior with a real, reactive load with the added bonus of not risking to burn a driver with high power sines. I guess I'm among the very few that sent some 100W of pure sine into a $1000 woofer (it is alive and well, thank you).
if only determining if clipping occurs is what you're after, simply using the sound board as a scope may be even simpler. one can even record and do post-analysis. not to mention real-time analysis.
an advantage of recording real music through the sound card instead of using sines is that one can capture behavior with a real, reactive load with the added bonus of not risking to burn a driver with high power sines. I guess I'm among the very few that sent some 100W of pure sine into a $1000 woofer (it is alive and well, thank you).
Correct! The test tone is 12dB RMS below FULL SCALE. That makes it 9dB below the highest possible sine wave value. A little confusing, I know. But I believe that full scale should always be the reference.In other words, if I've just figured this out correctly
My measured value.
Thanks Pano for an educational and informative thread.
My measured value is.
1.9 V -12dB at 120Hz.
Loudspeaker sensitivity of 92dB (2.83V/1m).
Room 4x5m.
If I understand correctly. this means??
My max value is 4x1.9 = 7.6 V
My RMS W value of 8ohm is 1.9 Vx2, 83 = 5.38 V -- SQR 5.38 / 8 = 3.6 W.
If my speakers has its lowest impedance 3.5 ohms at 120Hz. My maximum power peak is aprox. 16.5 W (SQR 7,6/3,5).?????
Thanks Pelle
Thanks Pano for an educational and informative thread.
My measured value is.
1.9 V -12dB at 120Hz.
Loudspeaker sensitivity of 92dB (2.83V/1m).
Room 4x5m.
If I understand correctly. this means??
My max value is 4x1.9 = 7.6 V
My RMS W value of 8ohm is 1.9 Vx2, 83 = 5.38 V -- SQR 5.38 / 8 = 3.6 W.
If my speakers has its lowest impedance 3.5 ohms at 120Hz. My maximum power peak is aprox. 16.5 W (SQR 7,6/3,5).?????
Thanks Pelle
Hi Push pull
A problem with a sound card based oscilloscope and the file comparator software that uses a sound card that I had suggested earlier is like I also said, that what you’re looking for can contain or predominantly be information above what the sound card can capture.
For audio work, one wants an oscilloscope with at least a decade greater BW than what you’re looking at.
For example, a sound card can capture a 20KHz sine wave, but if your amp was clipped to a full square wave at 20KHz, all the 16/44 sound card would see / capture is a sine 20KHz wave.
Now, most of the audio is well below that and so a sound card based method would be of value (and is why I suggested it).
I mention the old way too because 5th had problems with the software but the idea is you compare the input to output, the amp gain is only scaling, it does not change the wave shape and THAT is what you’re looking for (the output NOT matching the input, such as at instantaneous clipping).
I mention the transformer based approach because that is real time and an ancient way to get the difference between an input and output but mention it last because if people don’t have test equipment lying around, they probably don’t have a pair of signal transformers lying around.
Unfortunately, to examine features in signals that happen very quickly and are over just as fast, takes more than ones ears or a simple meter.
Best,
Tom
A problem with a sound card based oscilloscope and the file comparator software that uses a sound card that I had suggested earlier is like I also said, that what you’re looking for can contain or predominantly be information above what the sound card can capture.
For audio work, one wants an oscilloscope with at least a decade greater BW than what you’re looking at.
For example, a sound card can capture a 20KHz sine wave, but if your amp was clipped to a full square wave at 20KHz, all the 16/44 sound card would see / capture is a sine 20KHz wave.
Now, most of the audio is well below that and so a sound card based method would be of value (and is why I suggested it).
I mention the old way too because 5th had problems with the software but the idea is you compare the input to output, the amp gain is only scaling, it does not change the wave shape and THAT is what you’re looking for (the output NOT matching the input, such as at instantaneous clipping).
I mention the transformer based approach because that is real time and an ancient way to get the difference between an input and output but mention it last because if people don’t have test equipment lying around, they probably don’t have a pair of signal transformers lying around.
Unfortunately, to examine features in signals that happen very quickly and are over just as fast, takes more than ones ears or a simple meter.
Best,
Tom
So you need 25 watts, or 50 watts, what's 3 dB amongst friendsGoing through the maths though its obvious as the peaks of the -12dB tone actually hit -9dB and if multiplied by 4 we end up with +3 which is impossible.
In other words, if I've just figured this out correctly
If that's correct then it appears that I inadvertently added to some of the confusion.
.
no it doesn't. because at that power and that sensitivity even average levels are very high. I'd say that his average is louder than my peaks which is very loud IMO.
A 90 dB one watt one meter speaker can do 113 dB at one meter with 200 watts. Add a second, and in room response goes up about 3 dB, 116 dB maximum. Move to 2 meters, peaks are down to 110 dB.
A snare drum peak level can sound about twice that loud (120 dB) at that distance, assuming no power compression, the speakers would require 2000 watts to achieve that level.
For those that like levels approaching loud live music without distortion, 200 watt peaks in to a pair of 90 dB speakers is not unusual.
Seems odd you have NEVER been near a stage with rock music, or a horn section, or a disco, where peak listening levels are frequently above what a 90 dB speaker with 200 watts can achieve
.anything with notes in it. he has the most eclectic tastes I can think of. he said that the maximum level was with opera.
Describe your setup , you maybe experiencing dynamic compression and not know it ... I'm sure your friend is not using 200 watts RMS and I'm sure his avg din is not over 88 db, @ 88 db he will need all of that 200 watter to not have dynamic compression on opera..
I have found some listen at 3 m on the Avg... most 3- 4M ......
You obviously don't listen to very dynamic music or recordings, 84 din can rise up to 103 db ..
This makes sense, one of the hardest recordings to reproduce and sound a live is opera or any kind of stage production. I cannot listen to my Sarafina recording unless i have the big amps out, if not dynamic compression kills the experience and then when you have the avg down, makes the whole thing useless, no reality !!!!
If you listen to a lot of pop and rock type music , it will take a lot less to sound loud, they are usually recorded hot, with limited dynamics and mucho noise...
Last edited:
256watts for 114db @ 1 m, 5 db for the additional pr and channel , extrapolate listening distance at 4 M works out to 107db for 512 watts total both channels...
Means his 200 watter will not hit dynamic peaks of 107 db without distortion and compression. If it's opera, it's easy to conceive he is listening at an avg din of 84-86 db leaving only 10-12 db of headroom before clipping..
no way. it's just that I don't ever listen to opera, listening position is a bit closer and I live in a block of flats.