Serious DIY speaker builders work very hard at reducing cabinet vibrations. And they do it with good reason. Cabinet resonances can blur and swamp the signal being reproduced by drivers. This often causes that "boxy" sound no one wants. In fact, it has been reported that in some instances cabinet resonances are produced at higher SPL's than the signal the drivers are producing! This is not good.
There are lots of theories on methods of reducing cabinet resonances. They include cabinets that are: high mass, low mass, stiff, floppy, lossy, etc., etc. I will leave the theory to others. What this post is about is an inexpensive method of testing the success of your cabinet vibration reduction techniques.
The method I suggest is to use an accelerometer. You stick it on a cabinet and play pink noise through the speakers then take the output of the acclelerometer and feed it to your acoustic measurement system. You can then look at a FR plot of the cabinet's "response". While the pink noise is hissing away you can move the accelerometer to different points on the cabinet and try to determine what is what. You will be surprised at what you will discover. In effect, you are using an electronic stethescope that allows you to see what the cabinet is doing in different places on its surface.
The problem with this method is that sensitive commercial accelerometers are quite expensive. I will show you a very sensitive tool that you can make for virtually nothing out of what you probably have in your shop. Even if you have to buy the parts it will only cost a few bucks. The idea came from an Audio Amateur magazene article years ago. And it works like a dream.
The heart of the tool is a piezio element from one of those cheap piezio tweeters (just take apart the T and extract the element). The piezio element has special properties. If voltage is placed across the element it will flex. Indeed, that is the way it produces sound. But the converse is also true. That is, if you flex the element it will produce voltage. So in a sense it can be made to act like a microphone for cabinet vibrations.
What must be done is to provide a method of attaching the elemet to the cabinet in such a way as to allow the cabinet vibrations to flex the element. This can be done by supporting the element in the middle with a rod that can be attached to the cabinet. As the cabinet vibrates the vibrations are passed to the center of the element and its outer edges are then free to flex.
The picture below shows my version. The green thing is a smallish wire nut like those used in construction. I filled the base with epoxy to give a flat spot to stick to a cabinet. I epoxied the element to the top and bound the thin RS coax to keep from stressing the lead out wires.
To use it, simply put a tiny bit of Blue Tac, clay or plumber's putty on the flat bottom of the wire nut and stick it to the speaker (press hard). Run pink noise through the speaker, feed the signal to your measurement system (I have the IMP) and see your cabinet's resonances as a FR curve.
Some comments are in order:
This unit is not calibrated so you cannot come up with absolute SPL levels but it will show relative db's on the FR curve.
I have not used it with measurement systems other than the IMP. But it should do just fine. Some measurement systems might need a pre-amp to raise the signal level, but maybe not. Try it and tell us.
Surprisingly, even though a tweeter element is used, it is quite a wide band tool and does a marvelous job.
No, I do not have any graphs to show. I threw out the old ones years ago. But I am building some speakers now and should have some plots to show in the near future. If you beat me to it, post some plots.
The best way to develop a cabinet is to build a prototype, measure it, then start adding/changing braces, dampening, etc. until you get what you want. You may find that what you are accomplishing is only raising or lowering the frequency of resonances or going from high Q to low Q, etc. Then you must choose what you want. Engineering is all about compromises.
Hope you find this useful.
Enjoy!
There are lots of theories on methods of reducing cabinet resonances. They include cabinets that are: high mass, low mass, stiff, floppy, lossy, etc., etc. I will leave the theory to others. What this post is about is an inexpensive method of testing the success of your cabinet vibration reduction techniques.
The method I suggest is to use an accelerometer. You stick it on a cabinet and play pink noise through the speakers then take the output of the acclelerometer and feed it to your acoustic measurement system. You can then look at a FR plot of the cabinet's "response". While the pink noise is hissing away you can move the accelerometer to different points on the cabinet and try to determine what is what. You will be surprised at what you will discover. In effect, you are using an electronic stethescope that allows you to see what the cabinet is doing in different places on its surface.
The problem with this method is that sensitive commercial accelerometers are quite expensive. I will show you a very sensitive tool that you can make for virtually nothing out of what you probably have in your shop. Even if you have to buy the parts it will only cost a few bucks. The idea came from an Audio Amateur magazene article years ago. And it works like a dream.
The heart of the tool is a piezio element from one of those cheap piezio tweeters (just take apart the T and extract the element). The piezio element has special properties. If voltage is placed across the element it will flex. Indeed, that is the way it produces sound. But the converse is also true. That is, if you flex the element it will produce voltage. So in a sense it can be made to act like a microphone for cabinet vibrations.
What must be done is to provide a method of attaching the elemet to the cabinet in such a way as to allow the cabinet vibrations to flex the element. This can be done by supporting the element in the middle with a rod that can be attached to the cabinet. As the cabinet vibrates the vibrations are passed to the center of the element and its outer edges are then free to flex.
The picture below shows my version. The green thing is a smallish wire nut like those used in construction. I filled the base with epoxy to give a flat spot to stick to a cabinet. I epoxied the element to the top and bound the thin RS coax to keep from stressing the lead out wires.
To use it, simply put a tiny bit of Blue Tac, clay or plumber's putty on the flat bottom of the wire nut and stick it to the speaker (press hard). Run pink noise through the speaker, feed the signal to your measurement system (I have the IMP) and see your cabinet's resonances as a FR curve.
Some comments are in order:
This unit is not calibrated so you cannot come up with absolute SPL levels but it will show relative db's on the FR curve.
I have not used it with measurement systems other than the IMP. But it should do just fine. Some measurement systems might need a pre-amp to raise the signal level, but maybe not. Try it and tell us.
Surprisingly, even though a tweeter element is used, it is quite a wide band tool and does a marvelous job.
No, I do not have any graphs to show. I threw out the old ones years ago. But I am building some speakers now and should have some plots to show in the near future. If you beat me to it, post some plots.
The best way to develop a cabinet is to build a prototype, measure it, then start adding/changing braces, dampening, etc. until you get what you want. You may find that what you are accomplishing is only raising or lowering the frequency of resonances or going from high Q to low Q, etc. Then you must choose what you want. Engineering is all about compromises.
Hope you find this useful.
Enjoy!
Comment