I currently don’t use crossover between DML and tweeters either.
At high frequencies, there is also a capacitor
But I adjusted the EQ on Full frequencies
I also have tests that I can accept without using EQ.
But I prefer to use a little EQ to adjust
Didn't pay attention this morning
I found that @Andre Bellwood cannot be opened either.
Is there anyone else who can upload it?
Because I can't open it either...XD
Thank you very much for sharing
Hello Audiofrenzy,
I hope you will accept to answer few questions... if not showing the details of those panels
- Which material is the panel?
- What is the cut off frequency (or capacitor value) of the tweeter?
First, I won't push the DML to cone comparison very far. A cone speaker is a mass loaded device. This is a first important limitation in its bandwidth so in its cone diameter. In addition the diameter of the cone as a direct consequence on the directivity.
In DML, the voice coil is loaded by the mechanical impedance of the membrane. The HF limitation is no more linked to the diaphragm mass. The mechanism in HF limitation might be different.
About the link between the HF and the dimensions of the panel, I have the feeling through what I have experimented, it might be more complex.
I have 2 quite large plywood panels for which I don't have the feeling neither the evidence with measurements of a lack of HF.
At the opposite, while experimenting with polypropylene, the lack of HF is here with small dimensions piece of this material.
I am wondering about material in between, with an HF absorption such as this effect of lack of HF becomes visible. EPS? Unfortunately, I haven't managed to find high density enough to experiment with.
Christian
Christian,
Yes, that's what I understood their idea to be, that is, that the transition frequency should be as low as possible, most ideally below the audible range (or at least below the frequency range of application). That's actually quite consistent with what I have been trying to achieve with high aspect ratio panels combined with strong perimeter damping.
I too would be interested in a method to measure the transition frequency. A scanning laser vibrometer would help but I don't have access to one! I've been giving a little thought to what measurement(s) I could make using the tools I have (FEM, tap testing, close-micing, impedance testing) that might provide an elegant alternative to the laser vibrometer but I don't have the answer yet. If you have any ideas let me know.
Eric
Eric,
I read the paper...
The explanation about the Modal Overlap, the comparison to the Schroeder frequency makes sense to me.
The paper seems too limited in the relation between the membrane material and the named transition frequency. At least with not enough evidence (in my opinion of a person with limited knowledge in such topic) Only results for acrylic are shown; what about others? It was shared here that HF, in some materials (plywood?) can go far no? OK the only sensor we have is a mic...
About the post, you are looking for... difficult. I have in mind exchanges with Steve about the different modes the DML works but I have not specific enough keywords to find it...
Christian
Christian
deanznz,
I read that too and thought it a bit curious. I am frankly a bit confused about what they are trying to say there. They single out "rectangular panels with constrained edges" but as far as I can tell, panels of any shape and unconstrained edges also display sharp peaks in their frequency response. And often sharper peaks, as there is less opportunity for introducing damping with free edges.
Later in the same section they add to my confusion when they cite Heron's patent as an example of a "highly compliant panel with free edges". Heron's patent indeed describes a panel with free edges, but the panel itself is a 40 mm thick (!) panel with aluminum honeycomb core and skins, and hence is about as far from "highly compliant" as I can imagine!
From Heron's patent:
That's probably the least compliant panel I've ever heard of used for a "DML"!
All that said, I do see how there could be merit to using truly compliant panels, as such would shift all the natural frequencies down, and thereby increase the density of natural frequencies in the audible range.
Eric
An idea I have (I don't remember if I already shared it) is the use of a cartridge (this old thing for vinyl disc). What I don't know is the transfer function. This will sense only one point not giving a full membrane view in one shot.
Christian,
Yes, I agree, validation across other materials and/or other levels of damping would have made the work more convincing.
No worries about not finding the old posts. I was just curious, but probably there's no real value in it.
Eric
Christian,
Can you explain more what you are suggesting? I'm not following.
Concerning the use of a mic as a alternative to a vibrometer, as you may suspect, I think it has great potential value. I'm pretty sure one could create similar graphics to the vibrometer results by driving the panel using an exciter at a fixed location and frequency (as they did in the paper) and measuring the SPL using a close mic (say 2 mm) at locations across the entire surface of the panel, and using SPL as a proxy for displacement amplitude. Of course, it would take forever, so it would not be practical.
But is there a more elegant way? That through judicious selection of the driving point and measuring points, and using white or pink noise instead of a single frequency, that the transition frequency could be estimated? That's how I'm leaning now...
Eric
Hi Eric, would the chladni plate test be useful ?
Do watch it for 6-7 minutes if you can.
The video link should jump to the correct part.
Last edited:
In theory, yes, I think it would. If you ramped the frequency, at some point you would stop seeing patterns, and that should be the transition frequency.
Last time I tried a Chladni plate demo I got some cool figures but burned up an exciter, as you may recall.
https://www.diyaudio.com/community/...s-as-a-full-range-speaker.272576/post-7198955
Hello Eric,
The idea is to use a phono cartridge. Have a look here for some illustrations : Equivalent circuit of the dynamic system of a phono cartridge. A phono cartridge transforms the speed of the stylus in volt. The difficulty is to have an idea of this transfer function according to the frequency.
With a mic, there is no reason to proceed with a single frequency in my opinion. My favorite test is the IR with a logsweep. Depending of the sweep duration, the frequency resolution is good. Treatment like smoothing are possible afterwards. It is more difficult with a noise where the frequency resolution is defined at the measure by the RTA or spectrum analyser. setup.
Christian
I can't help but think..... was there a time when we could see sound
from:
https://goreguitars.com.au/design/#Modal_Tuning
from:
https://goreguitars.com.au/design/#Modal_Tuning
Hmm, that would be cool.I can't help but think..... was there a time when we could see sound
I must admit that I'm initially a bit skeptical of this claim:
I would really like to hear the explanation of that. The only thing I can think of offhand is the possibility that at each of the different frequencies with the monopole shape on the face, the sides and back are doing something different.
Eric
Not just strings either. Pretty much all instruments produce a series of harmonics (i.e overtones that are integer multiples of the fundamental). Even percussion instruments produce a series of overtones, though they are typically not harmonic overtones.
Eric
Ah, I see. The idea is to use a phono stylus/cartridge as a displacement sensor. Makes sense.
Concerning the single frequency test, I mentioned it only as the simplest (least elegant) example of a "brute force" method to try and qualitatively replicate the laser vibrometer result.
I like the frequency sweep test best for most applications too. The "noise" method is nice for some special applications though. For example, in the test I did some time ago when I continuously moved the mic across the surface of the panel to look for comb filtering. In that case the "noise" method was much quicker.
I'm not sure yet which approach would be best for investigating the modal/statistical crossover frequency, however. Perhaps neither.
Eric
Yes, the noise + RTA is excellent to find the exciter best place. I did it in my last panel (polypro version) built. The place I choose is totaly different the one I thought.
A sweep keep the time information, not the noise test. I just read a paper about the auditory system saying we are able of accurate evaluation in frequency and in time. What an FFT can't do. One day or later, the time aspect will appear.
I don't have it in mind for now but as you mentioned, the impendance curve should remain in the game.