Accelerometers to measure panel vibrations?

[hifijim]

So Remlab, between 550Hz and 650 Hz you have about a 1 dB difference between the two speakers. This means that the maximum null you could produce is -19.3 dB. At this frequency, your test is valid down to -19 dB.

[Remlab]

Jim
Wow, this chart demystifies a lot! Thanks! Now I'm getting a grasp of what JohnK was referring to

[Remlab]

This time, everything is on one graph except for panel vs noise..

Top two traces; Driver matching at 18" measured at exact same spot in room.

Dark green; Single speaker, on axis at 50"

Brown; Panel at 50"

Light green; Nulling baffles on axis, where they meet, with thick blankets over both panels at 50".(picture)

It looks like the blankets are almost totally transparent and not making any difference below 250hz, so the null method looks to be accurate only above that. Does that sound correct? Good cabinet bracing will push it above that point anyway, so it may be a useful tool to analyze the audibility of well braced cabinets, of which the Dunlavy certainly isn’t.

[Pano]

Nice lamp and shade over there in the corner. A recent find, or heirloom?

I was suspecting that the blankets might prove too transparent in the bass. The only practical thing I can think of is sandbags. I have some and could make more, maybe I'll give that a try tomorrow. Thanks for all your hard work!

[Pano]

In the mean time, why not try the same set of measurements but with the mic 19-20" away (50cm)? Or even at 25cm. You might be able to better isolate what the mic is picking up.

[Remlab]

Thank you! I Wanted to see if the Panel resonance could travel that far in the room, but your right, the resolution is not good, that far away. This next graph is my own speaker, but at 20"
Top traces, driver matching.
Green- Driver on axis at 20"
Brown- Panel at 20" using null method.

[Remlab]

Originally Posted by Pano

Nice lamp and shade over there in the corner. A recent find, or heirloom?

I was suspecting that the blankets might prove too transparent in the bass. The only practical thing I can think of is sandbags. I have some and could make more, maybe I'll give that a try tomorrow. Thanks for all your hard work!

The lamp was designed by my father back in the mid 60's. Mid century modern in design. He was a pretty famous sculptor.

[Remlab]

Wow. I did the null test on my speakers with the mic situated in between, and blankets on both. This is what I got. So the bump I thought was coming off my side panel at 500hz is actually coming from a resonance in between the baffles, I'm assuming from the phase plugs interacting with each other and interfering with the null. That's the only thing I can think of. So that's a good thing, because it's not coming off my baffle. It's most likely very silent all the way across the spectrum, Just like the vibration sensor was reading. The good thing about doing lots of measurements, is that a lot of assumptions get overturned. I'm realizing just how important the vibration sensor is in addition to this. They both have limitations, but when used together, you can definitely get a more complete picture.

[Pano]

Ah ha! Good find. I did wonder just how good a null could be achieved. This seems to show that it's better at some frequencies than others.

Thanks for the info on the lamp. Quite the heirloom, then. We had a good collection of MCM lamps before we moved twice. My father was a carver and did very sleek stylized animals in the 60s. Those we still have.