dasdoing
Major Contributor
just throwing this in for now, since I am still trying to understand all this. Sure some of you guys are more qualified to make conclusions:
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
intresting (phase) comparision of laser- and microfone meassurements of a (subwoofer) driver (and more)
- Thread starter dasdoing
- Start date
- Joined
- Feb 23, 2016
- Messages
- 23,432
- Likes
- 43,805
I haven't watched the 30 minute video. Would this not be correct? Remember the thread asking which way a woofers moves. It is backwards to what people think vs the air pressure waves. I'd think it would result in exactly this out of phase condition with such a test as in the video using a pressure microphone.
www.audiosciencereview.com
Which way is up? (Which way does a loudspeaker driver move?)
In a recent video by PS Audio/Paul, he mentions the sentiment that many probably share with him, that a loudspeaker driver works by compressing air as it moves outwards. This illustrates perfectly that you can work with loudspeakers for decades without ever understanding how a loudspeaker...
www.audiosciencereview.com
Last edited:
thanks for sharing. I guess @René - Acculution.com will find it interesting
The acoustic pressure is generated by the acceleration of the air "particles" driven by the diaphragm, not displacement. When displacement is sinusoidal, differentiate it twice (to get the acceleration) and you end up with a sinusoid of reversed polarity.
René's thread on this subject:
www.audiosciencereview.com
René's thread on this subject:
Which way is up? (Which way does a loudspeaker driver move?)
In a recent video by PS Audio/Paul, he mentions the sentiment that many probably share with him, that a loudspeaker driver works by compressing air as it moves outwards. This illustrates perfectly that you can work with loudspeakers for decades without ever understanding how a loudspeaker...
www.audiosciencereview.com
- Joined
- May 1, 2021
- Messages
- 461
- Likes
- 1,405
What a nice measurement setup. It is too bad that so many people insist on trying to understand theory via measurements, instead trying to understand the theory first, and then use measurements to validate and investigate. Modelling first, measurements later. You can clearly see that he finds it difficult to conclude anything, and it becomes somewhat hand-wavy. I have work with engineers that did this also, and I simply asked them to only show the measurements, and no analysis, because managers were eager to jump on any conclusion, and it would be very difficult to contain a lot of bad information circling the company.
Remember first: To talk about phase, we need have steady-state conditions. To even talk about frequency(!), we need to look at steady-state.
He seems to be confused about what he is finding, but a lot of it is very much to be expected. My article here (https://audioxpress.com/article/simulation-techniques-misconceptions-in-the-audio-industry) would probably have helped him.
- Polarity choice -
So, he wants to define a correct polarity, which is fine. He does something that I don't like and that is doing it via a click or impulse. This is problematic, since the immediate response will be determined by whether the transfer function related to this physical aspect is minimum-phase or not. The transfer function in question is indeed minimum-phase, but he does not necessarily know that. What he instead should do, is to look at steady-state conditions. For DC, he should look at whether the driver stays all the way in, or all the way out, after the displacement has settled. The common choice of polarity for a single driver, is 'all the way out for positive voltage'. But, what is then often forgotten or not understood is that that will lead to negative(!) displacement for positive voltage above the resonance frequency of the driver, which is the relevant frequency range for the audio.
- Phase vs delay -
He tries, like many others, to understand phase via delay, instead of understanding phase via how it falls out of complex numbers and the phasor aspects related to steady-state conditions. There are several delays that can be derived from phase, but only one phase. So when talking about delay, you should state very clearly, which delay you are talking about. In his case, what he is showing is phase delay, which is an apparent delay.
- Microphone -
It is a good idea to check the phase of the voltage output to the phase of the pressure for the microphone. Some preamplifiers flip the polarity, and there are also complications coming from the different types of microphones that exist, as some are better suited for near-field than others and similar considerations. I would again do it steady-state. For small drivers, putting a microphone in front of it, can modify the pressure, but in his case that would not be an issue.
- Voltage vs current -
He measures this, and a lumped element model will also show that the magnitude peaks at the fundamental resonance. The magnitude and phase aspects will be coming from a combination of electromagnetics, structural mechanics, and acoustics characteristics. This is important when relating voltage to force, since the force tracks with current, but displacement is typically of more interest than force.
- Voltage vs displacement -
As already mentioned, for low frequencies, the displacement will be in-phase with voltage. He also sees that for 1 Hz signal. He then seems confused about the fact that this changes as he increases the frequency, all the way to displacement being in anti-phase with voltage, and talks about maybe the stiffness being an issue. What he sees, is what he should see. The mass-spring system will lead to a 180 degree phase shift across the resonance frequency, so he captures this very nicely, and we can see that the battery test performed will directly lead us towards having negative(!) displacement for positive voltage in the frequency region of interest.
- Displacement vs pressure -
Here, we need to be really careful. What I have stated in articles, is that if we are forced to say anything about which way a driver moves compared to the resulting pressure, our best bet is to say that negative displacement goes along with positive pressure phase-wise (or rather positive acceleration tracks with pressure both magnitude-wise and phase-wise, but most people will think of movement in terms of displacement, hence my sticking to mentioning displacement). This is, however, not a general thing that has to hold. All that has to hold steady-state is that the pressure divided by the volume velocity has to give us the acoustic impedance seen from a surface in question, so an acoustic version of Ohm's law in electrical engineering(*). It just so happens that when loudspeakers are being measured, it is done under anechoic conditions, and for the typical driver size and relevant pass-band frequencies (and looking at the pressure at some distance away from the driver), the impedance is quite mass-like, although not perfectly so. So, for the elevator pitch, you would say that acceleration will drive the pressure, full well knowing that this would not be the case for a sealed subwoofer at low frequencies in a leakage-free room, since here the room would act as a compliance at low frequencies, and so positive(!) displacement would track with positive pressure. And so, one complication with his setup, is that his room will influence the displacement vs pressure aspects, and so it would have been better to first start with perhaps a smaller driver and preferably somewhat anechoic conditions.
Another issue related to the above, and to using a transient signal for drawing conclusions, is that he looks at the resulting pressure from a voltage 'blip' transient signal, and so the room is not seen, since this is not steady-state conditions, where the room pressure has had time to settle. This is then more of an anechoic response, and he also sees relationships between signals changing when comparing the immediate response and the steady-state response, which again is a result of both this room-aspect, and the driver response itself.
- Voltage vs pressure -
This combines the two points above, and so both the transducer and the room come into play. What we can say is that the choice of 'all the way out for positive voltage at DC', leads to 'all they way in(!) for positive voltage well above resonance', and this fits with the pressure being somewhat in phase with voltage in measurements, since the inwards displacement leads to positive pressure. So positive voltage approximates positive pressure in the relevant pass-band. So, I think there are two reasons for people saying that positive displacement leading to positive pressure: 1) At low enough frequencies, or even DC, we can SEE the driver move outwards, so we intuit that this a general thing, forgetting the second-order mass-spring aspect leading to a 180 degree phase shift, and 2) always measuring pressure and relating back to voltage, since you completely skip the structural mechanical displacement aspect and that there is a "180 degree shift" coming from voltage vs displacement, and then another coming shift from 'displacement vs pressure into mass-like impedance'. Without realizing it, we have falsely concluded how the driver moves, without actually having that aspect in our measurement. But again, his setup is complicated by the fact that he measures pressure in a room.
- Conclusion -
I wish he would do another video going much more step-wise through the above. This seems to be very a confusing topics even for seasoned loudspeaker engineers, and it really is not that difficult. You just need to separate the effects out, and show for example that for the surface displacement vs pressure the transducer principle does not matter (but acoustic environment, driver size, and frequency does), whereas for the voltage to displacement or to pressure, the transducer itself further complicates the results. A lot of effects are here combined, and the room aspect should probably have been taken much more into consideration in order to predict anything about how the pressure relates to the other variables.
(*) A further complication is that voltage and current are scalar variables, but volume velocity is a (normal) vector quantity involving a surface with a spatial aspect to consider, and the pressure and velocity in general vary across this surface, especially towards higher frequencies, while the associated acoustic impedance stays well-defined nonetheless. This relates to why I mention that the pressure of interest is away from the surface, as for a mass-like load, the acceleration will be in-phase with the pressure up to higher frequencies than when looking at the pressure right in the middle of a radiating surface. The alternative is specific acoustic impedance, which is pressure divided by particle velocity in a particular direction, and this holds in a differential (point) sense.
Remember first: To talk about phase, we need have steady-state conditions. To even talk about frequency(!), we need to look at steady-state.
He seems to be confused about what he is finding, but a lot of it is very much to be expected. My article here (https://audioxpress.com/article/simulation-techniques-misconceptions-in-the-audio-industry) would probably have helped him.
- Polarity choice -
So, he wants to define a correct polarity, which is fine. He does something that I don't like and that is doing it via a click or impulse. This is problematic, since the immediate response will be determined by whether the transfer function related to this physical aspect is minimum-phase or not. The transfer function in question is indeed minimum-phase, but he does not necessarily know that. What he instead should do, is to look at steady-state conditions. For DC, he should look at whether the driver stays all the way in, or all the way out, after the displacement has settled. The common choice of polarity for a single driver, is 'all the way out for positive voltage'. But, what is then often forgotten or not understood is that that will lead to negative(!) displacement for positive voltage above the resonance frequency of the driver, which is the relevant frequency range for the audio.
- Phase vs delay -
He tries, like many others, to understand phase via delay, instead of understanding phase via how it falls out of complex numbers and the phasor aspects related to steady-state conditions. There are several delays that can be derived from phase, but only one phase. So when talking about delay, you should state very clearly, which delay you are talking about. In his case, what he is showing is phase delay, which is an apparent delay.
- Microphone -
It is a good idea to check the phase of the voltage output to the phase of the pressure for the microphone. Some preamplifiers flip the polarity, and there are also complications coming from the different types of microphones that exist, as some are better suited for near-field than others and similar considerations. I would again do it steady-state. For small drivers, putting a microphone in front of it, can modify the pressure, but in his case that would not be an issue.
- Voltage vs current -
He measures this, and a lumped element model will also show that the magnitude peaks at the fundamental resonance. The magnitude and phase aspects will be coming from a combination of electromagnetics, structural mechanics, and acoustics characteristics. This is important when relating voltage to force, since the force tracks with current, but displacement is typically of more interest than force.
- Voltage vs displacement -
As already mentioned, for low frequencies, the displacement will be in-phase with voltage. He also sees that for 1 Hz signal. He then seems confused about the fact that this changes as he increases the frequency, all the way to displacement being in anti-phase with voltage, and talks about maybe the stiffness being an issue. What he sees, is what he should see. The mass-spring system will lead to a 180 degree phase shift across the resonance frequency, so he captures this very nicely, and we can see that the battery test performed will directly lead us towards having negative(!) displacement for positive voltage in the frequency region of interest.
- Displacement vs pressure -
Here, we need to be really careful. What I have stated in articles, is that if we are forced to say anything about which way a driver moves compared to the resulting pressure, our best bet is to say that negative displacement goes along with positive pressure phase-wise (or rather positive acceleration tracks with pressure both magnitude-wise and phase-wise, but most people will think of movement in terms of displacement, hence my sticking to mentioning displacement). This is, however, not a general thing that has to hold. All that has to hold steady-state is that the pressure divided by the volume velocity has to give us the acoustic impedance seen from a surface in question, so an acoustic version of Ohm's law in electrical engineering(*). It just so happens that when loudspeakers are being measured, it is done under anechoic conditions, and for the typical driver size and relevant pass-band frequencies (and looking at the pressure at some distance away from the driver), the impedance is quite mass-like, although not perfectly so. So, for the elevator pitch, you would say that acceleration will drive the pressure, full well knowing that this would not be the case for a sealed subwoofer at low frequencies in a leakage-free room, since here the room would act as a compliance at low frequencies, and so positive(!) displacement would track with positive pressure. And so, one complication with his setup, is that his room will influence the displacement vs pressure aspects, and so it would have been better to first start with perhaps a smaller driver and preferably somewhat anechoic conditions.
Another issue related to the above, and to using a transient signal for drawing conclusions, is that he looks at the resulting pressure from a voltage 'blip' transient signal, and so the room is not seen, since this is not steady-state conditions, where the room pressure has had time to settle. This is then more of an anechoic response, and he also sees relationships between signals changing when comparing the immediate response and the steady-state response, which again is a result of both this room-aspect, and the driver response itself.
- Voltage vs pressure -
This combines the two points above, and so both the transducer and the room come into play. What we can say is that the choice of 'all the way out for positive voltage at DC', leads to 'all they way in(!) for positive voltage well above resonance', and this fits with the pressure being somewhat in phase with voltage in measurements, since the inwards displacement leads to positive pressure. So positive voltage approximates positive pressure in the relevant pass-band. So, I think there are two reasons for people saying that positive displacement leading to positive pressure: 1) At low enough frequencies, or even DC, we can SEE the driver move outwards, so we intuit that this a general thing, forgetting the second-order mass-spring aspect leading to a 180 degree phase shift, and 2) always measuring pressure and relating back to voltage, since you completely skip the structural mechanical displacement aspect and that there is a "180 degree shift" coming from voltage vs displacement, and then another coming shift from 'displacement vs pressure into mass-like impedance'. Without realizing it, we have falsely concluded how the driver moves, without actually having that aspect in our measurement. But again, his setup is complicated by the fact that he measures pressure in a room.
- Conclusion -
I wish he would do another video going much more step-wise through the above. This seems to be very a confusing topics even for seasoned loudspeaker engineers, and it really is not that difficult. You just need to separate the effects out, and show for example that for the surface displacement vs pressure the transducer principle does not matter (but acoustic environment, driver size, and frequency does), whereas for the voltage to displacement or to pressure, the transducer itself further complicates the results. A lot of effects are here combined, and the room aspect should probably have been taken much more into consideration in order to predict anything about how the pressure relates to the other variables.
(*) A further complication is that voltage and current are scalar variables, but volume velocity is a (normal) vector quantity involving a surface with a spatial aspect to consider, and the pressure and velocity in general vary across this surface, especially towards higher frequencies, while the associated acoustic impedance stays well-defined nonetheless. This relates to why I mention that the pressure of interest is away from the surface, as for a mass-like load, the acceleration will be in-phase with the pressure up to higher frequencies than when looking at the pressure right in the middle of a radiating surface. The alternative is specific acoustic impedance, which is pressure divided by particle velocity in a particular direction, and this holds in a differential (point) sense.
audiofooled
Addicted to Fun and Learning
- Joined
- Apr 1, 2021
- Messages
- 617
- Likes
- 667
I agree. With further complications that we have drivers of different sizes and that we put them in boxes and rooms, understanding what is it that we measure usually gets even more complex.
What happens at higher frequencies and different measurement positions quickly reminds us that we also have directivity:
Loudspeaker size vs frequency and what we see at the end of the video may also be interesting thing to consider.
Similar threads
