Any ideas as to what would happen with the radiation pattern, and on-axis sensitivity, of a driver that happens to be on a baffle completely covered with a thick layer of felt? I'm talking enough dense, real wool felt to absorb >90% (or for the sake of argument, 100%) of the sound energy in the frequency band of interest.
If the size of the baffle dictated that the radiation normally be into half-space, what would this absorption do? Would the driver start behaving as some sort of mix of half-space and full-space? Would it still radiate into half-space, with no effect on on-axis sensitivity, with the only effect being some reduced baffle diffraction?
If the size of the baffle dictated that the radiation normally be into half-space, what would this absorption do? Would the driver start behaving as some sort of mix of half-space and full-space? Would it still radiate into half-space, with no effect on on-axis sensitivity, with the only effect being some reduced baffle diffraction?
As long as that felt is mounted on a solid baffle, the baffle step is still a function of how wide the baffle is and it’s shape (round, rectangle, trapezoid, flower, etc). The felt only affects the higher midrange and treble (above 1kHz, maybe as low as 700Hz) by smoothing out any little surface irregularities on the baffle that might cause ripples in the tweeter response. If you mounted the felt on a frame and it was porous to the backside, that would be a leaky baffle and you would get even more baffle step loss as it would appear to be a smaller effective size. But for purposes of calculating the baffle step correction for typical 8in to 12in wide baffles, I don’t think the felt changes anything below circa 1kHz where the baffle step begins. I have covered baffles with melamine foam - a very good absorber. Probably better than felt and haven’t noticed much impact on the BSC.
Last edited:
In your thought experiment wouldn't the sound waves that one would consider a drivers "beaming" be the only ones mostly audible? The rest would be mostly absorbed by the felt baffle.
This is quite interesting because 4"-6" of fiberglass/rockwool is able to absorb a large amount of sound (upwards of 80%) down to 500 Hz or even less.
But taking the example of a tweeter, if it's flush-mounted to a solid baffle that is completely covered in a perfect absorber, it would have the same on-axis sensitivity as a tweeter on the same size baffle that had no absorption at all?
I'm quite curious as to the mechanics of how the diaphragm "sees" the baffle width even when its waves are absorbed before reaching the edges.
4in to 6in of fiberglass covering a baffle is different than half an inch of felt which is sort of how I was interpreting your felt scenario. If the driver cone is setback in a 6in deep hole from the fiberglass, sure it’s going be shielded. I was assuming driver cone is flush with the felt or whatever material is covering the baffle.
OllBoll took measurements of a felt baffle in an open baffle design. You can use that as a comparison to your question.
Dipole wool carpet baffle experiment
Dipole wool carpet baffle experiment
You will find it very hard to achieve this. The absorption amount will be very much frequency related. Thick wool on the baffle will have no affect at the frequencies involved with baffle step.
dave
It's missing the context from the OP and the follow-up posts. But the question isn't relevant to very low frequencies anyway.
Sorry, that is the bit confusing.
The thickness of the fiberglass would not be 4"-6", maybe more like 1" to 2", if we were using fiberglass. But the "thickness" as perceived by the sound waves would be the distance that the sound wave travels across the baffle, ergo, the "thickness" of insulation should be equivalent to the width of the insulation on the baffle. In other words, if the baffle is completely covered in absorptive material, the thickness of the insulation would be 1/2 the width of the baffle, if the drivers were aligned centrally and symmetrically on the baffle. Does that make sense?
The baffle reflects the part of the spherical wave formally propagating towards the back redirecting it forward and then it recombines with the other half of the wave. At low frequencies the baffle is not efficient at reflecting the sound wave and therefore there is no increase in SPL. At slightly higher frequencies the baffle becomes more efficient for reflecting waves and there the baffle step happens. For even higher frequencies there will be sometimes an alignment causing a lot of reflections and sometimes not so many. This will increase the frequency response anywhere between 0 dB (no diffraction from the whole baffle towards a certain direction) and 9,5 dB (full diffraction from all edges of the baffle) depending on the specific alignment. On average the increase will be 6 dB since there will be usually diffractions from some parts of the baffle but not from others. Since diffractions are only a redistribution of energy there will be a similar behaviour off-axis but with shifted frequencies for the maxima and minima.
Now, if you put an ideally absorbing felt over the whole frequency range on the baffle non of this would happen. The baffle step would be gone and there will be no longer a 6 dB gain at higher frequencies. Also there are no more deviations in the frequency response due to diffractions, neither on-axis nor off-axis. However, depending on the size of the chassis they still may show a certain degree of beaming which may also mess up the off-axis response. (http://www.acousticfrontiers.com/wp-content/uploads/2015/02/driver-beaming.png)
If you have a more realistic felt which only works somewhere above the baffle step. You will start at low frequencies with a 0 dB increase, then the baffle step will happen pushing you to about a 6 dB increase and above that the effect of the felt will kick in pushing the frequency response back down to a 0 dB increase.
Now, if you put an ideally absorbing felt over the whole frequency range on the baffle non of this would happen. The baffle step would be gone and there will be no longer a 6 dB gain at higher frequencies. Also there are no more deviations in the frequency response due to diffractions, neither on-axis nor off-axis. However, depending on the size of the chassis they still may show a certain degree of beaming which may also mess up the off-axis response. (http://www.acousticfrontiers.com/wp-content/uploads/2015/02/driver-beaming.png)
If you have a more realistic felt which only works somewhere above the baffle step. You will start at low frequencies with a 0 dB increase, then the baffle step will happen pushing you to about a 6 dB increase and above that the effect of the felt will kick in pushing the frequency response back down to a 0 dB increase.
It is a complex arrangement that varies with frequency. One way to begin imagining it would be to assume the baffle isn't there, and so diffraction occurs where previously it didn't.
I'd be hesitant to refer to anything as a 'loss', diffraction has consequences (this thread is about those). It infers that there can be a 'correction', but you cannot correct diffraction.
That makes sense to me. Do you have any empirical testing that correlates with this?
I have heard that covering a baffle with thick felt (meaning 1/2 inch) results in a smoothing of diffraction very similar to rounded baffle edges. If true, this implies that one could use felt covering on a hard-edge baffle and get the same benefit as rounding or deep chamfering. It seems that Vandersteen and Wilson do this.
Anybody ever try this? or try to measure it? I am curious, but I am not in a position to test it out.
Anybody ever try this? or try to measure it? I am curious, but I am not in a position to test it out.
The driver would act as if there's no baffle there.
Consider this: an open window has 100% absorption. Sound going in that direction never comes back.
Similarly, the very absorbent baffle is absorbing 100% of sound going in those directions. ie, the baffle becomes a window.
The baffle step frequency will increase, corresponding to the driver's faceplate or basket diameter.
Baffle diffraction will be eliminated entirely.
It's an interesting design idea. I'd suggest it'd lose up to 6dB of potential output compared to a design with a baffle, but using decent drivers with plenty of power handling can make that up just fine. The elimination of all baffle diffraction would be a very interesting thing to listen to.
If you get to the prototyping stage, I'd be interested to know how it goes.
Chris
If you hung a tweeter by string in free air, wouldn't there be significant diffraction at its edges?
Probably, which is why I specified baffle diffraction.
Since the path lengths are short (ie, only HF effects), some felt over the faceplate would probably eliminate that, too. Having zero secondary sources would be an interesting thing to listen to.
Chris
Since the path lengths are short (ie, only HF effects), some felt over the faceplate would probably eliminate that, too. Having zero secondary sources would be an interesting thing to listen to.
Chris
