Accelerometers to measure panel vibrations?

ACH-01

Several years ago, I played with an ACH-01, which you used to be able to buy a battery powered amplifier to go with it. You can find amp ref in bwaslo link in post #19. I used REW in an attempt to produce similar waterfall plots as shown in Stereophile. I used "Tacky Wax" to fasten it to the cabinet.
As several others have advised, it is not easy to calibrate and it does n't really tell much about what is going on.
Dependent upon how your rigid your cabinets are to start with it may be worth using the frequency generator in REW or similar, then using your hand or stethoscope to feel the vibrations.
For those wishing to DIY the amp, here is the detail.
 

Attachments

  • ACH-wiring diagram.jpg
    ACH-wiring diagram.jpg
    43.9 KB · Views: 357
...
For those wishing to DIY the amp, here is the detail.
It can actually be much simpler than that, just a 2.2k between the battery +9V and the "drain" connection. Connect the "source connection" to the "ground" connection. A 3.3uF cap from the "drain" to the soundcard hot input, and battery [-] connection to ground, and soundcard cold/hot connection to ground. Two other parts only.
The opamp circuit is if you want all the dynamic range you can get, but that isn't likely needed unless you are trying to drive a pro audio mixer input.
 
Ya just stick that one with...what the heck DO you stick it on with? I've read Atkinson's measurements for years, and is it Vance Dickason? But I don't remember the details.

My idea is have a professional build the cabinet shells, and then I experiment with all the internals. Like how about a rigid brace, versus a brace under compression but with a gel pad at one end. Well, something more durable than gel but with some vibrational elasticity which I someone opined was more effective than total rigidity.

Probably pretty much a waste of your time. You might be able to move the resonant frequencies up higher, but you really can't get rid of them in a box speaker.

Siegfried Linkwitz covered this subject quite well at Burning Amp 2017.

Here is link to the video:

Burning Amp 2017 - Siegfried Linkwitz - YouTube

Make sure you watch the whole thing. It's almost an hour long and he talks about this towards the end.
 
Last edited:
I both agree, and ( respectfully ) disagree, yes adding braces with shift the frequency higher rather than diminishing the vibration ( unless it's shifted above the crossover point of the mid/bass, in which case it's possible that a worth while improvement could be effected ), however I do believe ( and possibly Remlab would agree ) that box resonances can be significantly reduced, however you do have to make significant changes to box construction to do this, Eclipse TD, BBC ls3/5a, Remlab sandwich baffle or box in a box. I applaud head_unit's intention of implementing a methodical series of experiments, and hope to see the results.
 
Agreed. There is an awful lot of talk on the subject and a pitiful amount of data.

The question is not whether you can measure the resonances or not. They will be there and you can find them. The question is what can you really do about it. And particularly for the average DIYer.

SL with all of his time and talent finally concluded after a lot of testing that the answer to that question is - not much. Some may disagree, but I'm inclined to accept the conclusions of a man with his genius and time investment for understanding sound reproduction from speakers.

If you haven't watched the referenced video I highly recommend it. Not only for this subject, but for a lot of other material in the beginning that is very useful.
 
Last edited:
I recall Analog Devices releasing some accelerometer ICs many moons ago. Probably 15 years ago. One approach could be to start with one of the evaluation boards for these. At the time I was thinking to measure camera/lens vibration with them for the evaluation of various tripods.

Accelerometers are commonly stuck to surfaces using bees wax. The wax doesn't mar the surface and cleans up easily. Of course nothing prevents you from using epoxy or something similar, which could be handy for prototype builds where you want the accelerometer stuck in one place for a long time.

Tom
 
Member
Joined 2010
Paid Member
Here is a measurement with the ACH accelerometer placed on the panels of a subwoofer cabinet.
The side panels were very well braced and the back panel was not braced at all.
There is a difference of almost 20 dBs between them.

My opininon is that if you have the capability to make measurements, then there is a lot of things you could do to lower the vibration of a cabinet panel.
It will be a slow, step-by-step procedure but the final result will be much better.
And an accelerometer like ACH is a good start.
 

Attachments

  • Cabinet Vibration measurements.jpg
    Cabinet Vibration measurements.jpg
    123.7 KB · Views: 226
I should have been more specific, I was about to edit my post, I'm talking of edge diffraction, he mentions at the end that he doesn't like stepped baffle because of the diffraction. Diffraction from baffle edges

I'll read the material you referenced when I have more time. It's rather lengthy and SL isn't always easy to understand. Sometimes it takes several readings. At least for me.

But I think the quick answer to this particular video is that there is an alternative - namely electronic time delay. So if you can avoid any possible diffraction issues due to physical geometry why not do so electronically. SL is big on digital processing in general.
 
Box resonances start at the midbass driver and propagate through the cabinet. The most efficient and cost effective way to deal with this is a sorbothane gasket decoupling the midbass on the front of the baffle with rubber washers decoupling the screws on the back of the baffle(This part can only be done with a removable baffle though). As far as transient response concerns go, Laurence Dickie of Vivid has been doing this with virtually all his designs(From the woofers all the way up to the tweeters) since the B&W Nautilus with world class results. I took the concept further than that for a relatively huge cost($250 for each 7" x 12" removable baffle), but using the technique I just mentioned, you could get about 70% of the way there(Maybe more) for the cost of the sorbothane and some precision cutting with an exacto-knife. Once again, if Laurence Dickie uses this method, there is no debate. He's proven the efficacy of the technique beyond a reasonable doubt.
 
Last edited: