"This may be a dumb question, but how would the impedance show the delayed energy?"
As you said, the driver is also a mic, so sound impinging on the diaphragm will create a voltage that modulates the impedance trace.
As you said, the driver is also a mic, so sound impinging on the diaphragm will create a voltage that modulates the impedance trace.
"...the impedance reflects the acoustic effects only to the extent of its efficicency - maybe 5%."
That's about the net efficiency of a CD on your WG's, but that includes the XO/padding; isn't the actual transduction efficiency much higher, like an order of magnitude?
That's about the net efficiency of a CD on your WG's, but that includes the XO/padding; isn't the actual transduction efficiency much higher, like an order of magnitude?
noah katz said:
Yes, its likely to be higher, but not an order of magnitude, because 50% is the theoretical limit that could be achieved and we are nowhere close to that. I say 10% sure, maybe 15%, but more than that would be unlikely.
But the bigger issue is the liklihood that the HOM would not have any effect on the diaphragm.
gedlee said:
gedlee said:
I'm simply more optimistic than you
😀
As for the “average” HOM effect – thats ok for me – if I may put it that way. To quantify the detailed effects on the sound field are a task I'd intentionally split to subsequent “acoustic” measurements.
For a starting point what you think about a “delay pipe” of lets say 2m / 6' and a horn of lets say 0.5m / 1.5' depth?
If we start with a driver of 1” what is the cut off frequency of the delay pipe?
As for the horn we may start with a horn that isn't exactly one – just an other pipe attached of lets say the same diameter as its length giving a 30deg “flare”.
To ensure HOM we even might attach the delay-pipe off axis.
1.Could you give the numbers for the HOM free frequency range of the delay-pipe (cut off frequency of the delay pipe)?
2.Could you give the frequency for the first second … order HOM modes in the horn-pipe?
3.What delay between the mouth reflection of the axisymmetric wave front versus the HOM wave fronts to be expected (difference in time of flight)?
Michael
noah katz said:
have a look at thend's set up.
http://www.diyaudio.com/forums/showthread.php?postid=1843034#post1843034
Michael
Hello,
Modern impedance mesurements are performed using pulse response method.
Noise free impedance mesurements may be obtained using the pulse response recovered by Angelo Farina's method (logsweep + convolution).
As the impedance is presented as a pulse response, the delayed component of the impedance due to reflections, stored energy, etc... may be clearly visible.
As an example give a look to Thend's impedance mesurement for a compression driver mounted on a tube:
http://thend.chez-alice.fr/Audio/Tube_acoustique.html
Now, if a parasitic peak is visible on an impedance curve then it will corresponds to something visible on the response curve and most probably it will be audible.
Best regards from Paris, France
Jean-Michel Le Cléac'h
Modern impedance mesurements are performed using pulse response method.
Noise free impedance mesurements may be obtained using the pulse response recovered by Angelo Farina's method (logsweep + convolution).
As the impedance is presented as a pulse response, the delayed component of the impedance due to reflections, stored energy, etc... may be clearly visible.
As an example give a look to Thend's impedance mesurement for a compression driver mounted on a tube:
http://thend.chez-alice.fr/Audio/Tube_acoustique.html
Now, if a parasitic peak is visible on an impedance curve then it will corresponds to something visible on the response curve and most probably it will be audible.
Best regards from Paris, France
Jean-Michel Le Cléac'h
pjpoes said:
gedlee said:
This makes sense to me. (BUT) The diaphragm is buried behind the phase plug, is in a serious magnetic field and there is high pressure in the throat. If there is a HOM reflection going back I would think it would have almost zero (as in inaudible) effect on the diaphragm.
However - Jmmlc shows it can be measured.
If this can be measured then shouldn't HOM (at least part of it!) also show up in the pulse impedance measurement?
As you have stated before the HOM's are delayed from the original- If we minus out the acoustic gain (?) of the horn (can this be done minus HOM?!) and subtract the response the of the driver "unloaded" from the loaded driver shouldn't we be close to seeing these other effects that would be left in a CSD?
TrueSound said:
The electrical impedance seen at the electrical input contains imformation about the motion of the voice coil. The voice coil motion will be sensitive to the AVERAGE of the forces on the diaphragm (assuming its rigid), but the voice coil will not respond at all to any motions that average to zero across the diaphragm. I was thinking about this and it is true that no HOM will have an AVERAGE effect on the diaphragm because only the lowest order mode has a finite average. Hence it will be impossible to detect HOM through the electrical terminals.
Michael
I still have no idea what your test setup is intended to do,so I can't really help.
Possibly I come back with a simplified simulation of the sound wave splitting into axisymmetric and HOM propagation plus its (partly) reflection at the
Earl, could you possibly provide the numbers I asked for or point to formulas for calculation?
Michael
Earl, could you possibly provide the numbers I asked for or point to formulas for calculation?
Michael
I think what he is asking for is what the formula would be to calculate the frequency of the HOM's in a planewave tube. That or have you calculate it for him. The first question I believe is asking what frequency is the cut off at which no HOM's would exist in this theoretical planewave tube. The next is asking what the frequencies would be for the first higher order mode, second, etc. That way he can look for evidence of HOM's in his impedance measurements.
I don't understand what HOMs in a plane wave tube have to do with HOM in a waveguide. They are not related in any way.
If its the HOM for a plane wave tubes it can be found in any text on acoustics in the section on sound or waves in a duct. It's an important aspect so it should be in most texts. I don't know it by heart and I'd just have to look it up, so I think that Michael can do that without much difficulty. Its the frequency where a wave equals the Bessel functions, so its the root of the Bessel functions, plus another term that depends on the angular modes. Pretty standard stuff.
If its the HOM for a plane wave tubes it can be found in any text on acoustics in the section on sound or waves in a duct. It's an important aspect so it should be in most texts. I don't know it by heart and I'd just have to look it up, so I think that Michael can do that without much difficulty. Its the frequency where a wave equals the Bessel functions, so its the root of the Bessel functions, plus another term that depends on the angular modes. Pretty standard stuff.
Might be it is helpful to show some pix to gain some better understanding of what HOM actually is – and how we possibly come to some quantification – or at least get a more intuitive handle on the issue.
Below I will show 4 pictures that should make clear what HOM looks like in its "pure" form (in ducts / pipes) and a fifth simulation that shows how HOM affects wave front propagation and acoustic impedance in horns.
All sim's were done with CARA.
As CARA actually is based on ray tracing and intended for investigation of room acoustics – it not necessarily yields the most "scientific" results for the topic at hand – but even with the gross simplifications I had to "work around" its limitations, the results are certainly worth for a basic understanding related to HOM.
Except two pictures all sim's show what's happening in the time domain. Meaning we see the wave front – and its reflections - propagating along its way.
I added two pix (#3, #4) showing how – at a certain frequency - the sound field gets deformed in a very specific way – which has given the "high order modes" its name IMO.
1.) wave front propagation in finite pipe with 1% reflection at the mouth:
2.) wave front propagation in infinite pipe:
we see the axisymmetric wave front is followed by the bouncing HOM-wave fronts. If we'd stimulate the pipe by a sinus instead of the Dirac-impulse here - we get a certain overlay pattern – mere interference
3.) stimulation of a infinite pipe with a fixed frequency captured at different phase angles:
4.) sound field (pattern) of a infinite pipe stimulated with the same fixed frequency as above:
if we'd do a cross cut at the first half or so – we would get a picture looking like a standing wave perpendicular to the axis of the pipe.
The more nulls this "standing wave" has - the higher the HOM-mode
The HOM modes are restricted with respect to dimension and frequency.
We also can clearly see that the HOM's are flattening out away from the source.
5.) conical horn with 1% reflection at the mouth:
Certainly not being an expert in horns – my take form the above would be that it does not make a big difference if the reflections of the mouth or the small parts of HOM hit back.
😉
Regarding the acoustic impedance we have to look out for reflections of the wave front hitting the origin (driver) again.
The simulations on the ducts were intentionally done with a source (driver) that does not fit the pipe in order to show that its at this point of severe diffraction that the bouncing HOM wave fronts get created.
Michael
Below I will show 4 pictures that should make clear what HOM looks like in its "pure" form (in ducts / pipes) and a fifth simulation that shows how HOM affects wave front propagation and acoustic impedance in horns.
All sim's were done with CARA.
As CARA actually is based on ray tracing and intended for investigation of room acoustics – it not necessarily yields the most "scientific" results for the topic at hand – but even with the gross simplifications I had to "work around" its limitations, the results are certainly worth for a basic understanding related to HOM.
Except two pictures all sim's show what's happening in the time domain. Meaning we see the wave front – and its reflections - propagating along its way.
I added two pix (#3, #4) showing how – at a certain frequency - the sound field gets deformed in a very specific way – which has given the "high order modes" its name IMO.
1.) wave front propagation in finite pipe with 1% reflection at the mouth:
2.) wave front propagation in infinite pipe:
we see the axisymmetric wave front is followed by the bouncing HOM-wave fronts. If we'd stimulate the pipe by a sinus instead of the Dirac-impulse here - we get a certain overlay pattern – mere interference
3.) stimulation of a infinite pipe with a fixed frequency captured at different phase angles:
4.) sound field (pattern) of a infinite pipe stimulated with the same fixed frequency as above:
if we'd do a cross cut at the first half or so – we would get a picture looking like a standing wave perpendicular to the axis of the pipe.
The more nulls this "standing wave" has - the higher the HOM-mode
The HOM modes are restricted with respect to dimension and frequency.
We also can clearly see that the HOM's are flattening out away from the source.
5.) conical horn with 1% reflection at the mouth:
Certainly not being an expert in horns – my take form the above would be that it does not make a big difference if the reflections of the mouth or the small parts of HOM hit back.
😉
Regarding the acoustic impedance we have to look out for reflections of the wave front hitting the origin (driver) again.
The simulations on the ducts were intentionally done with a source (driver) that does not fit the pipe in order to show that its at this point of severe diffraction that the bouncing HOM wave fronts get created.
Michael
Hello,
Why give too much headache of HOM in a small horns. The biggest HOM will be because of the room! The room HOM will be there whatever you do. Don't you think it would be best to start with the biggest problem first?
Next time someone asks me where do I live, I'll answer: In a HOM generator. That'll give them something to chew about 😀
- Elias
Why give too much headache of HOM in a small horns. The biggest HOM will be because of the room! The room HOM will be there whatever you do. Don't you think it would be best to start with the biggest problem first?
Next time someone asks me where do I live, I'll answer: In a HOM generator. That'll give them something to chew about 😀
- Elias
Michael it appears as if your simulations treat the mouth opening as a solid wall. Wouldn't this be incorrect, the opening of the waveguide or horn is not a solid wall, waves don't hit the opening and bounce back, as it appears in those images. When it was discussed that HOM's might be bouncing back into the diagram, my assumption was that we were talking about reflections far closer to the throat. I'm not sure I can even wrap my head around how diffraction at the waveguide mouth could create reflections back into the diagram that would be of any major consequence.
pjpoes said:
Michael says he's only modeling a 1% (-40dB) reflection back from the mouth so I'm guessing he set the 'wall' to be 99% absorptive in CARA. Seems like a reasonable assumption to me -- maybe not a perfect model but a pretty good way to see what's going on.
Edit: the only thing is the mouth reflection has inverted phase and I don't thing CARA shows that. It's inverted because the high pressure in the horn suddenly gets lower as it reaches the mouth, giving an impulse that goes negative.
catapult said:
Yes, that's one of the work around's in my CARA sim.
In fact – the wave front isn't plane when it reaches the mouth either – look at Jean Michels work in particular – and also it might be that the reflection (from diffraction / scattering) at the mouth isn't uniformly over its entire area but more dense close to the wall, where the second sources are created actually – dunno - haven't seen measurements, simus or "worked through" whole theory about – maybe someone else can comment...
catapult said:
I think CARA actually does as its mere reflection - but even if not – yes - phase inversion wouldn't change the overall picture given - only the details in the pattern of constructive and destructive interference...
pjpoes said:
Look at thend's measurements:
http://thend.chez-alice.fr/Audio/Tube_acoustique.html
Or look at any impedance measurement of any speaker – it always mirrors into which load a driver has to radiate.
The simus shown might give a guide line at which points in time we have to watch for reflections showing up in an impulse response measurement.
Its when the reflected wave fronts hit the driver again.
The finite pipe IMO is the best starting point in this regard, as it can be considered to be the worst possible HOM generating "horn" ever possible.
😉
If we are able to definitely trace down HOM in a finite pipe – we may go ahead with "real" horns and wave guides...
Michael
Here are some impulse just to show how impulse response can be altered.
1. raw driver,
2. large elliptical waveguide
3. small axisymmetric wave guide
4. another small axisymmetric wave guide.
All inpulses have same scale.
Note that first reflection is reduced using a large waveguide which has better directivity control.
Also note what happens before the first reflection.
1. raw driver,
2. large elliptical waveguide
3. small axisymmetric wave guide
4. another small axisymmetric wave guide.
All inpulses have same scale.
An externally hosted image should be here but it was not working when we last tested it.
Note that first reflection is reduced using a large waveguide which has better directivity control.
Also note what happens before the first reflection.