I'll test noise as well.
I'll be using a microphone - my aim is to measure the actual sound radiated by the material under somewhat realistic conditions.
It is really only the *difference* between test panels which will matter anyway. Looking forward to seeing the results
Hello Geotone... thanks for chiming in...
Please allow me to suggest that to truly discern between materials both room (microphone) and panel face (accelerometer) testing would be valid in the examination. If possible, you should consider doing both.
Also looking forward to your results.
Please allow me to suggest that to truly discern between materials both room (microphone) and panel face (accelerometer) testing would be valid in the examination. If possible, you should consider doing both.
Also looking forward to your results.
Last edited:
It is really only the *difference* between test panels which will matter anyway. Looking forward to seeing the results
Exactly! It will allow me to do comparative curves.
Douglas, since I already have a loudspeaker measurement setup, I'll be using that. I'll follow the same rigorous procedure as when measuring speakers. Branching out into yet another technique is one project too many right now! Besides, I don't really understand what is to be gained from those accelometer readings - if the DUTs transmit significantly different levels of sound, the mic will capture it. If they don't, the effect can be neglected, which I would actually welcome as a finding!
I get the point of using an a'meter when you have a non-isolated speaker, but isn't it generally a bit of a crutch, since it's difficult to gauge the actual SPL in relation to the speaker itself? I seem to recall Earl having pointed this out somewhere.
Anyhow, it will be a few months before I have anything to show - moving the workshop to a new location right now. And even once I pour: concrete needs to cure for a several weeks!
Last edited:
I suspect the reason is that a microphone may be located on a node for a given set of frequencies which may coincide with a resonant frequency of a test panel.
Yes, you could pick a node on the surface too. You can run the test in many places to achieve a meaningful average. I can't think of a reason right now to use both and acc'mtr and a mic AND multiple locations.
I have built massive (3" thick) turntable plinths (for a Garrard 401) from concrete. They rang like a bell. There is very poor damping of concrete structures. The resonances of the concrete box would swamp any resonance of the panel. Just tap a length of 12" diameter concrete plumbing pipe. Maybe if the concrete box is buried up to the baffle in the ground or in sand, you would get a dead test enclosure.
A pipe and a single panel supported on its corners have very different geometries for resonances from a braced box*. But yes, I shall test that before as well, and am thinking about a sandwich construction for the test box too!
At some point obviously this becomes a quixotic endeavour.
*Hence also the idea of using the same construction for the DUT as in an actual box, instead of a single panel clamped down somehow, without much relation to how it's actually used.
Last edited:
The accelerometer is a non-audio measurement of movement... a vibration detector. It would let you monitor the actual vibration of the panel, which might give hints to it's influence on radiated sound ... for example, if the panel vibrates at 100hz and you have a bump or dip in the 100hz sound field, you have a good indication of why. Without the second measurement, your guess is as good as mine.
Plus, I would be very interested in vibration readings from the panels themselves, even without SPL measurements. That would give valuable clues to which viscous material works best.
None the less ... if it's troublesome, I can understand why you would not do it. No worries. It was just a suggestion.
Thanks Douglas, I know what an accelerometer is ;-) I just have strong doubts about its usefulness for audio, since the correlation of measurements to audibility is mostly conjecture.
My measurements will be comparative in nature: MDF vs ply vs whatever sandwich etc. So it's about getting an overall picture of what works best and where (in the spectrum).
My measurements will be comparative in nature: MDF vs ply vs whatever sandwich etc. So it's about getting an overall picture of what works best and where (in the spectrum).
Last edited:
How about:
- Dig a deep, baffle-sized hole.
- Dump a big pile of fiberglass or similar in it.
- Place a speaker face-down on corner supports, so that it fires into the hole while the box's other 5 sides are exposed.
- Use a dense felt gasket to seal the perimeter.
- Measure actual acoustic output with a microphone.
I‘ll leave that to Patrick Bateman with his yard in the San Diego sunshine I live in an apartment in central Vienna, few opportunities for holes in the ground. And also lots of rain.
A 60kg concrete box will do its job just fine. The ringing reports are slightly exaggerated. Xylophones are rarely made of concrete I believe.
I live in an apartment in central Vienna, few opportunities for holes in the ground. And also lots of rain. A 60kg concrete box will do its job just fine. The ringing reports are slightly exaggerated. Xylophones are rarely made of concrete I believe.
Only to the subjectivist mind. I'm more a science guy... I like to see proof that it does or does not matter.
I think I worded that poorly... my apologies.
What I was driving at is that without analysing the baffle vibrations you actually have no way of knowing their effect on the overall sound. The results then become subjective (open to interpretation) because they ignore key facts.
For example...
I recently ran into a problem with a new installation that had both my friend and I quite confused... there was a 10db bulge in the SPL scanns of his new speakers. But where the heck was it coming from?
It took a few days to figure out, but it was a wall that was being extremely reactive to sound. At about 100hz it vibrated like a drum head... but only from certain spots in the room, and it echoed voices rather badly.
My point was that you probably do need to analyse the baffle as well, to know what it's effects (or lack there of) really are.
Due to the complex nature of the vibration, warping producing peaks ("positive" and "negative") and nulls?
Yes Scott, I get the theory - but how much does it matter? That's what I want to know.
You could do the pretty things shown in the salty video with drivers as well (if they were flat), but what does it tell you about the sound? A lot really, but that is an issue for Drs Kurt Mueller and Klippel. There must be a reason we measure loudspeakers with mics as opposed to accelerometers - because they show us just where those peaks and nulls happen in 3D, at listening positions.
Douglas, room effects will be largely excluded because I use gated measurements - same proven method used for measuring and then building speakers. I also don't expect useful measurements much below Schroeder Frequency, maybe from 250Hz up.
Maybe I didn't explain my Isolator™
well enough - the whole point is the isolation of the one aspect I think we are discussing here: to what extent does a loudspeaker enclosure act like an unwanted transducer and therefore interfere with the sound of the product? And how can that be minimised? I guess that makes me an engineer rather than a scientist, but certainly not a subjectivist.
You could do the pretty things shown in the salty video with drivers as well (if they were flat), but what does it tell you about the sound? A lot really, but that is an issue for Drs Kurt Mueller and Klippel. There must be a reason we measure loudspeakers with mics as opposed to accelerometers - because they show us just where those peaks and nulls happen in 3D, at listening positions.
Douglas, room effects will be largely excluded because I use gated measurements - same proven method used for measuring and then building speakers. I also don't expect useful measurements much below Schroeder Frequency, maybe from 250Hz up.
Maybe I didn't explain my Isolator™
well enough - the whole point is the isolation of the one aspect I think we are discussing here: to what extent does a loudspeaker enclosure act like an unwanted transducer and therefore interfere with the sound of the product? And how can that be minimised? I guess that makes me an engineer rather than a scientist, but certainly not a subjectivist.
Very little or there would be more interest in it, and it would only be an issue if resonances affected the FR unduly