I could feel cabinet resonance when I put my hand on, but pretty faint. There's just the one brace in the mid section of the cab. I've wanted to acquire a pair of the 1st series to compare but I've lost interest. The waterfall graphs of both are very similar. The 1st series were able to handle a bit more power and did not have the protection circuit that was thought to be it's limiting factor. Bass definition was considered to be better in the series 2. IME, I have not heard bass definition as sharp and articulate in a dynamic driver as those, seemingly perfect.
In 1997 I rebuilt my ATC SCM100As, and converted them tom ASLs. In this I braced the cabinets thoroughly, damped the panels, and soaked the front baffles in 50/50 polyurethane and white spirit where the MDF grain was open on the cut-outs. They took about a litre each, and stiffened and deadened.
I rounded all the port entrances and exits, and rewired with the thickest I could find.
I used 1mm EPDF rubber to make driver gaskets, for a slight decoupling/damping and a better seal, and made a 3mm PVC large gasket with which to bolt the detachable front panel to the cabinet main body. I theorised that these would only minimally affect transient response and information loss, but the advantages of reducing transfer of energy would be significant, but I made no attempt to measure it.
Bill Woodman asked me to submit to him a write-up of what I had done, enclosing samples of materials, and he didn't belittle what I had done.
I rounded all the port entrances and exits, and rewired with the thickest I could find.
I used 1mm EPDF rubber to make driver gaskets, for a slight decoupling/damping and a better seal, and made a 3mm PVC large gasket with which to bolt the detachable front panel to the cabinet main body. I theorised that these would only minimally affect transient response and information loss, but the advantages of reducing transfer of energy would be significant, but I made no attempt to measure it.
Bill Woodman asked me to submit to him a write-up of what I had done, enclosing samples of materials, and he didn't belittle what I had done.
This is why measurements are useful as I linked to earlier 😉. Why did you theorise that?
So, I'm going to an even softer foam about the driver basket - appx 6X6" - with a 2X4 stud glued to the cab back wall connecting to the back of the magnet through another foam piece cushion. How soft is it? The driver would probably rip right out of the box if you tried to ship it...
Understandably, this isnt as good as building a box with the complex structure that can hold the driver very firmly and distribute the reactive forces without introducing any resonant sound of its own. It's more of an experiment to see if I can make a really crappy 6 wall box with zero bracing sound OK.
I realize I'll lose whatever it is you get by providing the driver a relatively infinite mass to push against.
Understandably, this isnt as good as building a box with the complex structure that can hold the driver very firmly and distribute the reactive forces without introducing any resonant sound of its own. It's more of an experiment to see if I can make a really crappy 6 wall box with zero bracing sound OK.
I realize I'll lose whatever it is you get by providing the driver a relatively infinite mass to push against.
You guys should be taking your design cues from the Kef 105.2. It's well engineered and easily copied.
Try this:
Take a driver. Take three small blocks of wood.
Find a speaker or to accentuate the effect a dresser or tabletop or other speaker material like surface.
Put the blocks on the cabinet or top, so the driver frame will rest on the blocks.
Then facing the driver down (or up if the blocks are tall enough, and you are not worried about the cone) with sound playing through it.
Alternately press the driver firmly onto the blocks (simulating screws holding it in pressure) and lift it just enough, so that it doesn’t touch, but the sound reflection is the same.
Now you have a ballpark and quite visceral idea of how much the driver/cabinet connection contributes to the sound.
Take a driver. Take three small blocks of wood.
Find a speaker or to accentuate the effect a dresser or tabletop or other speaker material like surface.
Put the blocks on the cabinet or top, so the driver frame will rest on the blocks.
Then facing the driver down (or up if the blocks are tall enough, and you are not worried about the cone) with sound playing through it.
Alternately press the driver firmly onto the blocks (simulating screws holding it in pressure) and lift it just enough, so that it doesn’t touch, but the sound reflection is the same.
Now you have a ballpark and quite visceral idea of how much the driver/cabinet connection contributes to the sound.
I went in a different direction myself.
We have huge variations of humidity and temperature over here, so, trying to soft mount a driver with constant pressure on the gasket is quite the exercise.
So, to kill baffle resonances, I am using layered materials on the baffle attached to each other with epoxy style glue.
My best baffle has a combination of 3 very different materials, each having their own resonance, and each damping the other resonances.
The result is a very quiet baffle. Even with the woofers working overtime, putting a hand on the baffle and you can't feel anything.
We have huge variations of humidity and temperature over here, so, trying to soft mount a driver with constant pressure on the gasket is quite the exercise.
So, to kill baffle resonances, I am using layered materials on the baffle attached to each other with epoxy style glue.
My best baffle has a combination of 3 very different materials, each having their own resonance, and each damping the other resonances.
The result is a very quiet baffle. Even with the woofers working overtime, putting a hand on the baffle and you can't feel anything.
There is no getting away from it.
I understand the temptation to believe the driver should not move but I haven't found any evidence to support that belief.
If the materials are what I think they are any resonances would be higher in frequency than a woofer would excite. The only way to know would be to measure in some way.
Here is a good read about driver decoupling in the development of the B&W 800D
B&W Group North America Service & Support - Reference Materials
its in the appendix of the 800D white paper PDF
cheers, Arthur
B&W Group North America Service & Support - Reference Materials
its in the appendix of the 800D white paper PDF
cheers, Arthur
Thanks, Arthur - good to read. While they can go off and ultimately procure custom cast shape gel elastomers, I have to make do with the choice of whatever properties I can wring out of a few different yoga and other floor mats. Hopefully, these would be obtanium for most anyone, should they work out. The glue is setting and I can connect one speaker up tomorrow.
I botched the 3rd attempt at getting something together, this time wrecking the foam rubber baffle attempting to drill holes for the speaker mounting. Unfortunately, it didnt occur to me to heat a nail and plunge a hole, which worked great after being too late. So I tore off the soft foam and tried the hard foam again. (I believe I can retry the soft stuff again at a later time, should the hard turn out really horrible).
I cut a wood stud to the right length, glued it to the driver magnet. With a foot made of the soft rubber foam and glued at the stud's other end, I created something for the driver to push against, once glued to the cab's back wall. In addition to the basket being suspended upfront by the harder foam panel material.
The stud with the foam on the end makes quite a shock absorber. I discovered a prick-punch I had which is the so-called "automatic". You push the tip into the work - and keep pushing until it snaps. This snaping feature is very consistent, putting out the same impact every time. Something like
https://www.amazon.com/TEKTON-6580-Automatic-Center-Punch/dp/B0037UUO60
It makes for an interesting test, when you pop that punch into the side of a cheap speaker housing. Of course, you have to insulate the tip from the work in this case. When I tested the difference between the stud alone and the stud with the rubber foam pad, the attenuation into an unbraced box panel was quite significant. You can also put a rubber foot to the punch's output, which dampens it down a lot.
Last edited:
jjasniew, I find the idea of using an automatic center punch most interesting: a very accurate way to do the knuckle test repeatedly. Perhaps, in combination with a db meter, or a mic and laptop it could be a simple way to evaluate different cabinet construction methods.
Not quantified analysis, just based on sensibility; mass of basket + magnet far greater than cone + coil.
If a 12" driver was suspended by long strings, and given a high step voltage, would that (Xmax) cause any detectable reactional movement of the basket?
A close light could be used to amplify the movement, shown looking at its shadow on a distant wall.
An impulse can excite resonances way outside the spectral content of the impulse.
Yes, no doubt about it, but there doesn't appear to be any information loss or detriment to the transient response, in fact it seems the transient response improves.