You may be interested to hear that some recent measurements and investigation by wout31 led to the realization that the wavelet based spectrogram methods we discussed back in Post#201 actually have enough time resolution to show the internal reflections from the dustcovers you had mentioned back in Post#154. That was an unexpected discovery for me.
See measurements in Posts #15 - 20 here: https://www.diyaudio.com/community/threads/esl-high-frequency-chaotic-modes.382598/post-6939879
Hope this is on topic. I'm wondering what effect on resonance an out of phase panel would have sandwiched between two other 'in phase' panels? Could this be an effective dampening method when all 3 panels are bonded isobarically and all are driven? My particular focus here is on Acoustat panels.
Hi,
my initial thought was no, it'd just reduce output level.
On second thought it was .... hmmmh, if You use just two panels and drive one panel via a phase shifting equalizer that shifts phase only around the resonance, but works in phase for the remaining freq range.
At resonance it might cancel the resonance peak just enough and at freqs up to ka~4 (iIrc) the output adds up if the two membranes are just close enough to each other.
Remains the Q where to get, or how to derive the drive signal for the second panel from .... and would it be worth the extra effort.
jauu
Calvin
my initial thought was no, it'd just reduce output level.
On second thought it was .... hmmmh, if You use just two panels and drive one panel via a phase shifting equalizer that shifts phase only around the resonance, but works in phase for the remaining freq range.
At resonance it might cancel the resonance peak just enough and at freqs up to ka~4 (iIrc) the output adds up if the two membranes are just close enough to each other.
Remains the Q where to get, or how to derive the drive signal for the second panel from .... and would it be worth the extra effort.
jauu
Calvin
It won't give any effect. That is, it will give, but the opposite effect, there will be no bass at all.
In order to suppress the main resonance of the panel, it can be divided into several sections with different physical values, I did this on my first huge two-lane panels, their photos are here in my topic.
There, as a bass driver, I used a double bass membrane with an isobaric method and three stators. The total size of the bass panel was 40cm/2000cm. They were quite effective, at least twice as effective as a single membrane.
But you should not be afraid, as it seems to me, of the main resonance of bass panels of such a large size, because the resonance can go beyond the working frequency range of the musical material. Or if the resonance is, for example, 25 or 30 Hz, then you can simply cut it out with an equalizer .
Personally, I have been actively using the equalizer lately and have ceased to consider it an unnecessary thing
.It's not an issue with me. I have a great set up I've described here: www.diyaudio.com/community/threads/isobaric-esl-alternative.364330/
I read your post about felt as a damping material.
I am forced to write here my complete disagreement with this method of damping. The fact is that the electrostatic is a very weak motor, unlike the electromagnet of the speaker.
What is good for an electromagnet is very, very bad for an electrostatic. Any damping material in the bass segment completely kills signal energy. A large membrane is already unable to move the mass of air attached to it, so the bass of an electrostatic is very sluggish, and if the membrane is also dampened with some material such as felt or foam ... That is why electrostatics do not make acoustic design in closed cases.
Thanks for this Steve. Wavelet analysis certainly shows dustcover reflections clearly. Maybe time for dis beach bum to lern 2 reed, rite & kunt to unnerstan Don Davis's stuff. But for those versed in the art, it's fairly easy to pick this up from KEFplots (KEF CDS aka waterfalls) & PAFplots ... or even just straight Frequency Response.
In fact, you can probably see these on a raw Impulse Response. Laurie Fincham, co-inventor of KEFplots, claimed that after a bit of experience with CDS, he could recognise important features just from the IR.
If you've got the original IRs, you might want to see if you can do this too
In theory this actually works surprisingly well. I experimented a bit with the concept a few years ago when working on a double diaphragm project to improve sensitivity. In the end the complication of an extra amplifier, tuning the filtering, and another step-up transformer just didn't seem worth it when properly mounted acoustic resistance mesh could accomplish essentially the same thing.
I looked for my measurements but can't seem to locate at the moment. So here are a few modeling plots to think about.
First consider a theoretical situation where the diaphragms of ESL1 and ESL are coincident rather than one spaced behind the other.
- Left Column of plots: If both ESL1 and ESL2 are driven with the same signal you would get the expected 6dB increase and resonance would still have Q=10.
- Middle Column of plots: Insert a 2nd order AP filter @ 40Hz into the ESL2 drive signal and you damp the resonance as desired, but acquire a notch cancellation right at 40Hz.
- Right Column of plots: If you now reduce the ESL2 drive signal by about -1.3dB the notch disappears and you are left with text book LF roll off at the cost of about -0.5dB loss in sensitivity.
Now consider some Acoustat panels. At best, the diaphragm of ESL2 will be no closer than 35mm behind the diaphragm of ESL1.
- Left Column of plots: If both ESL1 and ESL2 are again driven with the same signal you still get the same LF behavior, but now the time delay of ESL2 results in comb filtering.
- Middle Column of plots: Using the same AP filter and -1.3dB signal reduction as above, LF is nearly text book with a small hitch, but the comb filtering remains
- Right Column of plots: If a 2nd order AP filter @ 4.9kHz is added to the ESL1 drive signal the LF returns to text book and the first notch of the comb filtering is removed.
If stator construction allowed getting the diaphragms closer together, this approach could remove all anomalies in the audio band.
But, I personally don't feel it is worth the extra cost or effort.
If we are talking about a large membrane and its main resonance, then this resonance is perfectly damped by the attached air mass, because this air mass will be heavier than the membrane mass.
But it's another matter if we build membranes by the isobaric method, then each subsequent membrane doubles its own resonance.
For example, the main resonance of a single membrane is 20Hz., Then the double membrane in an isobaric way will be already 40Hz.
But it's another matter if we build membranes by the isobaric method, then each subsequent membrane doubles its own resonance.
For example, the main resonance of a single membrane is 20Hz., Then the double membrane in an isobaric way will be already 40Hz.
Last edited:
It's not damped because the membrane makes that air mass flutter which is in turn what amplifies the resonance. So there needs to be a mechanical or in the case of Bolserst's example, electrical brake. The set up I have is an improvement in that sensitivity is increased. The felt provides good damping, probably more than needed but I have a pair of subwoofers to augment. It's well balanced as is. However I think I will remove the felt in favor of the mesh that's been referenced in a few threads. I'm waiting for Bolserst to chime in here.
Hi,
with a tightly stretched membrane you´ll get a resonance which is far from beeing perfectly damped, as typical Qs of >2 show.
You could omit with the resonance if the membrane is not stretched or just stretched a little bit.
You could achieve this by implementing a mesh of very soft foam, or an array of very soft supporting spots for the membrane.
In theory tis method should improve several parameters such as as efficiency, SPL, flashover level etc. and it makes mounting the membrane easier as You don´t need a stretching frame.
In praxis though its very difficult to find a suitable foam or supporting material, that doesn´t reduce efficiency too much.
jauu
Calvin
with a tightly stretched membrane you´ll get a resonance which is far from beeing perfectly damped, as typical Qs of >2 show.
You could omit with the resonance if the membrane is not stretched or just stretched a little bit.
You could achieve this by implementing a mesh of very soft foam, or an array of very soft supporting spots for the membrane.
In theory tis method should improve several parameters such as as efficiency, SPL, flashover level etc. and it makes mounting the membrane easier as You don´t need a stretching frame.
In praxis though its very difficult to find a suitable foam or supporting material, that doesn´t reduce efficiency too much.
jauu
Calvin
At one time I experimented a lot with different damping materials. My personal conclusion is that, for example, the same foam or other material at first glance will seem to be absorbing, but only at first glance, but in fact, any material that is located in a certain proximity to the membrane will be absorbing for certain frequencies, and for certain other frequencies reflective and create unnecessary interference for those frequencies in the audio range. And again, depending on what size the membrane is, with what tension it is and how thick the film itself is? All this directly depends on where in the sound range we will have the main resonance of the membrane body. If we can play with these damping materials in the medium and high frequencies without an audible drop in efficiency, then at low frequencies such damping materials act deadly on the bass register, that is, more than half of the pressure at low frequencies simply disappears, self-destructs. It is in this part that I am against damping (any). As for medium and high frequencies, then the matter is much more interesting, for example, you have a dip in a certain frequency range, in the 4-6 kHz region, by selecting the necessary material, including damping at first glance, you can place it at the required distance from the membrane and by reflection from this material to even out the unevenness of the amplitude-frequency characteristic. So this topic is very fun, but not for low frequencies 
.
.
The double membrane has half compliance. The moving mass is near identical (dominated by air mass), ergo Fs will be cca. 28 Hz.
Do you mean a separate second membrane or two membranes hermetically sealed in one bag with an attached mass of air between the two membranes?
So multiple membranes sealed isobarically renders moot the issue of resonance?
Hermetically sealed, of course, as near as possible. The mass of closed air (between membranes) is much lower (at frequency of basal resonance), than total moving air mass.
I don't understand your question exactly. Did you mean basal resonance point at dipole speaker?