Hi,
here are the panel resonances of the mentioned 3-way. I didn't used a crossover, but the curves are normalized to the nearfield responses to cancel out level differences and the frequency response. The nearfield responses are also attached as a reference. Because of the dip in the woofer's response (interference), the panel resonance curve is only valid up to 900 Hz. The resonances from 1200 to 1600 Hz are higher than in reality. But you can see from the panel response of the tweeter that there is a resonance.
The accelerometer is a B&K type, with upper frequency limit at 10 kHz. It was attached to the side panel.
The speaker is a 140 cm tall 3 way TML speaker with 2 10" woofers, 2 6.5" midranges and a 1" tweeter, all mounted on the front. Contrary to my post above, only one midrange was connected, but both woofers. So you have to subtract approx. 6 dB from the woofer's panel resonances. The TML is 2 times folded, and therefore the side panels are heavily braced. The cabinet is from 16 mm MDF except for the side panels, which are 10 mm MDF layered with an up to 25 mm profiled MDF panel, summed together 35 mm near the front and 20 mm near the back. No damping material attached to any surface. It has an flat area in the mid of the panel, so the accelerometer was easy to attach. I didn't choose the back panel for the measurement, because it is a wall of the TML and the measurements would show many of the internal TML resonances, too (I checked this before).
As you can see, the panel resonances excited by the woofer are stronger with higher frequency (compare the 150 Hz and 800 Hz resonance), the ones of the midrange have the same level (350 Hz and 1.8 kHz). Even the tweeter does excite the modes, but at a lower level (probably because of the plastic frontplate, which reduces the transmission from the tweeter to the cabinet).
Baseballbat
here are the panel resonances of the mentioned 3-way. I didn't used a crossover, but the curves are normalized to the nearfield responses to cancel out level differences and the frequency response. The nearfield responses are also attached as a reference. Because of the dip in the woofer's response (interference), the panel resonance curve is only valid up to 900 Hz. The resonances from 1200 to 1600 Hz are higher than in reality. But you can see from the panel response of the tweeter that there is a resonance.
The accelerometer is a B&K type, with upper frequency limit at 10 kHz. It was attached to the side panel.
The speaker is a 140 cm tall 3 way TML speaker with 2 10" woofers, 2 6.5" midranges and a 1" tweeter, all mounted on the front. Contrary to my post above, only one midrange was connected, but both woofers. So you have to subtract approx. 6 dB from the woofer's panel resonances. The TML is 2 times folded, and therefore the side panels are heavily braced. The cabinet is from 16 mm MDF except for the side panels, which are 10 mm MDF layered with an up to 25 mm profiled MDF panel, summed together 35 mm near the front and 20 mm near the back. No damping material attached to any surface. It has an flat area in the mid of the panel, so the accelerometer was easy to attach. I didn't choose the back panel for the measurement, because it is a wall of the TML and the measurements would show many of the internal TML resonances, too (I checked this before).
As you can see, the panel resonances excited by the woofer are stronger with higher frequency (compare the 150 Hz and 800 Hz resonance), the ones of the midrange have the same level (350 Hz and 1.8 kHz). Even the tweeter does excite the modes, but at a lower level (probably because of the plastic frontplate, which reduces the transmission from the tweeter to the cabinet).
Baseballbat
