My measurements on the panel resonances in the Pensil 7 enclosure showed the greatest energy was between 100 - 500Hz, which is probably about right. That is unbraced.
Midrange is defined as the 4 octaves from 160 Hz to 2.5 kHz.
The energy available to excite a resonance is roughly inversly proportional to the square of the frequency.
Any primary resonance will also likely also resonate at integer multiples of that frequency.
The only sensible option, it seems to me, is to push panel resonances as high as possible.
Making it hi Q makes sense because it is less audiable and energy bandwidth needed to excite it is less, and music -- usually -- not being a steady state almost single frequency tone is unlikely to ever pump sufficient energy into the panel at the appropriate frequency.
I use those as a starting point for all my designs.
dave
The energy available to excite a resonance is roughly inversly proportional to the square of the frequency.
Any primary resonance will also likely also resonate at integer multiples of that frequency.
The only sensible option, it seems to me, is to push panel resonances as high as possible.
Making it hi Q makes sense because it is less audiable and energy bandwidth needed to excite it is less, and music -- usually -- not being a steady state almost single frequency tone is unlikely to ever pump sufficient energy into the panel at the appropriate frequency.
I use those as a starting point for all my designs.
dave
Hi All,
Well, box problems are made up of a number of things. Some of these are easier to tame than others.
Now, MDF does tend to have a high resonance, but it's really difficult to excite the panels into movement. My (total) guess would place the panel resonances up in the midrange. With high stiffness and high mass, acoustic energy just isn't going to transfer that much energy into the box system. So right off the bat, we're dealing with reduced excitation (that's a good thing). Going the other way, you aren't going to see much surface movement with MDF and darn near none with HDF. You've got to try working with this stuff at some point people. Don't forget the carbide blade! I'm am not kidding about that.
Hi Chris,
Never hold commercial boxes up as a good target to aspire to. They are normally 1/2 the volume they should be. The density of material is too light and the joints are often a complete joke (one broken bead of glue, or hot glue is not that uncommon). The interior surfaces are never treated with anything, they only have to last the warranty period. They certainly do not want them to be rebuilt into anything useful. I've worked with commercial speakers in a warranty capacity in various places since the late 70's on a professional basis.
There is simply no reason why the DIYer's out there shouldn't seal the inside of their boxes. It is not that difficult and really only represents a little more money and another step or two. You can tape out the joint areas before painting or spraying. Too easy.
Now, if you're going to accept that many unknowns in a listening test, best not even remember the outcome. My point also included other aspects that suffered from this construction that was done. As I mentioned earlier, if the design is not executed properly, both outcomes will be substandard. One design may easily fare better than another.
-Chris
Well, box problems are made up of a number of things. Some of these are easier to tame than others.
Now, MDF does tend to have a high resonance, but it's really difficult to excite the panels into movement. My (total) guess would place the panel resonances up in the midrange. With high stiffness and high mass, acoustic energy just isn't going to transfer that much energy into the box system. So right off the bat, we're dealing with reduced excitation (that's a good thing). Going the other way, you aren't going to see much surface movement with MDF and darn near none with HDF. You've got to try working with this stuff at some point people. Don't forget the carbide blade! I'm am not kidding about that.
Hi Chris,
Never hold commercial boxes up as a good target to aspire to. They are normally 1/2 the volume they should be. The density of material is too light and the joints are often a complete joke (one broken bead of glue, or hot glue is not that uncommon). The interior surfaces are never treated with anything, they only have to last the warranty period. They certainly do not want them to be rebuilt into anything useful. I've worked with commercial speakers in a warranty capacity in various places since the late 70's on a professional basis.
There is simply no reason why the DIYer's out there shouldn't seal the inside of their boxes. It is not that difficult and really only represents a little more money and another step or two. You can tape out the joint areas before painting or spraying. Too easy.
Now, if you're going to accept that many unknowns in a listening test, best not even remember the outcome. My point also included other aspects that suffered from this construction that was done. As I mentioned earlier, if the design is not executed properly, both outcomes will be substandard. One design may easily fare better than another.
-Chris
As i said earlier i used to use HDF exclusively. Ply just does the same stuff better.
dave
Greater density is not a positive attribute (by itself)
Resonant frequency goes up if a material is stiffer.
Resonant frequency goes up if a material is lighter.
So stiff & light is better than stiff and heavy,
dave
I would suggest a laminate of soft 22mm particle board, and 12mm MDF on each side
A quick calculation gives a board/plate thickness of 45mm
the combination of hard/soft/hard and greater thickness will be really hard to beat
oh, sorry, I see a bit late that we are in full range forum, and very thick baffles may not be so practical
A quick calculation gives a board/plate thickness of 45mm
the combination of hard/soft/hard and greater thickness will be really hard to beat
oh, sorry, I see a bit late that we are in full range forum, and very thick baffles may not be so practical
Hi tinitus,
Hi Dave,
Heavy and dense and stiff is hard to transfer energy into, light and stiff - not as hard. Keeping energy out of your box system to begin with makes a lot of sense. All things kept equal, you need less MDF to achieve the same effects as ply would be. Surface damping is always a good idea, so that's a constant unless you need to add more to the ply so as to increase it's mass. HDF is lovely stuff as long as you don't have to work it.
-Chris
I can't argue that!
Hi Dave,
Heavy and dense and stiff is hard to transfer energy into, light and stiff - not as hard. Keeping energy out of your box system to begin with makes a lot of sense. All things kept equal, you need less MDF to achieve the same effects as ply would be. Surface damping is always a good idea, so that's a constant unless you need to add more to the ply so as to increase it's mass. HDF is lovely stuff as long as you don't have to work it.
-Chris
Are you sure? To transfer energy in it has to be within the bandwidth of the resonance. What you are talking about is like a symptom not the cause.
If you have 2 panels of the same stiffness, but one is less dense (ie lighter), the the less dense one will have a higher frequency resonance. Since as frequency goes up there is less energy to excite the resonance so it is less of a problem.
MDF is heavy and not very stiff.
dave
Wrong. Well, you have to be careful with everything in this life, obviously, but I think you might be misunderstanding somewhat. All back loaded boxes function over a limited bandwidth; the Pencil 7s you have built and refer to are bass cabinets for example. They may have wide-bandwidth drivers installed in them, but that's not germane to their functioning. If you stick a high order XO on them at 400Hz, it's not suddenly going to change the behaviour of the enclosure. Stored energy falls at 1/f (or the square of frequency if you prefer), therefore if you use a material with a high Fs, then it is easy to push panel resonance well up, above the functional BW of the enclosure, where there is little energy available to excite its resonant modes, and these are easily damped with a minimum of material, if indeed they need damping at all.
Indeed. Average MOE specs. demonstrate 1 1/8in thick MDF panels are needed to ~equal the stiffness of 3/4 thick panels of a decent void-free ply (BB or similar), the price being greater mass of course.
Last edited:
Hi Dave,
What you are talking about is exciting the resonance of any panels or any other structure. That's correct and well documented. But this is not the entire story. If the panels are braced sufficiently, the resonance becomes very high and well up into the midrange. From a power perspective, there just ain't that much power up there compared to the lower registers where the woofer is operating. Midrange and tweeter drivers are often mounted with a decoupling compliance between the driver and it's mounting surface, which is the case for anything I have ever built. So you are not going to receive energy from a driver operating in the same band. That leaves the woofer, or a full range speaker (which would also be mounted with a compliance).
I think the most important thing to realize is that the actual amount of energy that makes in into the enclosure as a system will not be within resonance from a driver coupled to the enclosure, or in the air volume inside that system. Anything that is within resonance will be greatly attenuated, and anything going back out will be similarly attenuated as well. I am assuming that normal practices were observed and all the panels have differing resonant frequencies.
Basically, the concerns you are raising are non-issues if the enclosure is properly constructed. That and these issues will be a concern no matter what material is used to create and enclosure, and the same good construction practices will reduce each to levels well below ambient noise. The surface treatment using undercoating isolates the panel from the air as well. That stuff stays compliant and really destroys the coupling between the MDF and the air.
It doesn't seem to matter what project is attempted with DIY efforts. Often factors that are not meaningful (if good design practices are used) get the bulk of the design efforts. This happens at the expense of other aspects so that a design ends up with easily corrected problems. You also see this effect in small commercial concerns that build "high end" audio products. Same story, different box thingy.
Always look at things as a system, they are not isolated from the other parts that make the system up. Also recognize those things that no longer have a meaningful impact on the performance. MDF / HDF also have advantages in that there is no grain, and no end grain either. All those issues don't exist, and I like it that way for at least the inner layer.
Hi Scottmoose,
This is also the reason why plywood performs much better as shelving over MDF or HDF. And also why steel performs better than the previous examples.
I construct all my shelving from 3/8" ply (not void free - think cheap) and 1" X 1" bracing. The shelves end up small, light and very, very strong. None have sagged over the course of 10 years with well over 200 lbs on them on each section. However, that doesn't make that construction method suitable for loudspeaker enclosures. Those panels are easy to excite. The shelf piled up with transformers doesn't seem to vibrate though!
-Chris
From experience, yes. I'm pretty sure.
What you are talking about is exciting the resonance of any panels or any other structure. That's correct and well documented. But this is not the entire story. If the panels are braced sufficiently, the resonance becomes very high and well up into the midrange. From a power perspective, there just ain't that much power up there compared to the lower registers where the woofer is operating. Midrange and tweeter drivers are often mounted with a decoupling compliance between the driver and it's mounting surface, which is the case for anything I have ever built. So you are not going to receive energy from a driver operating in the same band. That leaves the woofer, or a full range speaker (which would also be mounted with a compliance).
I think the most important thing to realize is that the actual amount of energy that makes in into the enclosure as a system will not be within resonance from a driver coupled to the enclosure, or in the air volume inside that system. Anything that is within resonance will be greatly attenuated, and anything going back out will be similarly attenuated as well. I am assuming that normal practices were observed and all the panels have differing resonant frequencies.
Basically, the concerns you are raising are non-issues if the enclosure is properly constructed. That and these issues will be a concern no matter what material is used to create and enclosure, and the same good construction practices will reduce each to levels well below ambient noise. The surface treatment using undercoating isolates the panel from the air as well. That stuff stays compliant and really destroys the coupling between the MDF and the air.
It doesn't seem to matter what project is attempted with DIY efforts. Often factors that are not meaningful (if good design practices are used) get the bulk of the design efforts. This happens at the expense of other aspects so that a design ends up with easily corrected problems. You also see this effect in small commercial concerns that build "high end" audio products. Same story, different box thingy.
Always look at things as a system, they are not isolated from the other parts that make the system up. Also recognize those things that no longer have a meaningful impact on the performance. MDF / HDF also have advantages in that there is no grain, and no end grain either. All those issues don't exist, and I like it that way for at least the inner layer.
Hi Scottmoose,
Exactly, and well put sir.
This is also the reason why plywood performs much better as shelving over MDF or HDF. And also why steel performs better than the previous examples.
I construct all my shelving from 3/8" ply (not void free - think cheap) and 1" X 1" bracing. The shelves end up small, light and very, very strong. None have sagged over the course of 10 years with well over 200 lbs on them on each section. However, that doesn't make that construction method suitable for loudspeaker enclosures. Those panels are easy to excite. The shelf piled up with transformers doesn't seem to vibrate though!
-Chris
You are repeating what i already said. It is correct and argues against higher density being a positive factor. All other things equal, resonances are higher if the mass being moved is less, so material that is light & stiff is better than heavy & stiff...
This discussion being in FR means no mids and tweeters, it is a tougher job than when you have XOs to restrict the bandwidth of energy in each section of the cabinet.
I have already been thru all the arguments in an earlier thread http://www.diyaudio.com/forums/cons...ussion-what-materials-build-speakers-out.html. It took a while but eventually the light went on in John's head.
dave
Hi Dave,
No, I am agreeing with you, but also pointing out that the issues are of such a low level as to represent a waste of time being concerned about it. If the box hasn't been built well, then it's too early to worry about this. Fix the problems or construction methods and this will sort itself out.
Besides, I did state ...
Whatever generates the energy that you are worried about exciting a structure must apply a force to cause motion. If the mass is higher, the deflection (or distance moved) will be less, given the same force. In other words, the higher the mass is, the lower any displacement would be. That translates directly into the amount of energy you are transferring to the enclosure as a system. That can be extended to show that the lower the energy existing in a structure, the lower the energy is that will leave the structure if that energy isn't dissipated as heat (damped). By placing a compliant material between the surfaces of the enclosure and air volume, further isolation then exists and the amount of energy you can transfer to it is greatly reduced even further.
In short, you want high mass.
Dave, two cars with identical rolling friction. A Smart car and a Rolls Royce. Your job is to rock them back and forth using the same amount of force on each. Which one ends up with the higher amplitude of motion?
Okay, same situation except that now you are pushing these cars through layers of foam. Does the Rolls even move detectably? Is the smart car rolling back and forth as much?
Okay, same situation except that Cal jumps into the Smart car without you noticing (tall foam). Does that Smart car move as much as it used to?
The Rolls Royce is the MDF, and the Smart car is the plywood. I wonder if the Rolls gets wet, does it come apart?
-Chris
No, I am agreeing with you, but also pointing out that the issues are of such a low level as to represent a waste of time being concerned about it. If the box hasn't been built well, then it's too early to worry about this. Fix the problems or construction methods and this will sort itself out.
It's not unusual for a helper tweeter to be used in an otherwise full range design. Id argue that a common two way system is really a FR plus helper tweeter. Nothing to roll off to the woofer and a single cap to the tweeter. This describes most low wnd, two way systems - yes?
Besides, I did state ...
By including references to mids and tweeters, I'm merely including other system types where these issues are just as important. The information here is just as applicable to multi-way systems as it is to FR systems.
I'm sorry, but I strongly disagree with you here.
Whatever generates the energy that you are worried about exciting a structure must apply a force to cause motion. If the mass is higher, the deflection (or distance moved) will be less, given the same force. In other words, the higher the mass is, the lower any displacement would be. That translates directly into the amount of energy you are transferring to the enclosure as a system. That can be extended to show that the lower the energy existing in a structure, the lower the energy is that will leave the structure if that energy isn't dissipated as heat (damped). By placing a compliant material between the surfaces of the enclosure and air volume, further isolation then exists and the amount of energy you can transfer to it is greatly reduced even further.
In short, you want high mass.
Dave, two cars with identical rolling friction. A Smart car and a Rolls Royce. Your job is to rock them back and forth using the same amount of force on each. Which one ends up with the higher amplitude of motion?
Okay, same situation except that now you are pushing these cars through layers of foam. Does the Rolls even move detectably? Is the smart car rolling back and forth as much?
Okay, same situation except that Cal jumps into the Smart car without you noticing (tall foam). Does that Smart car move as much as it used to?
The Rolls Royce is the MDF, and the Smart car is the plywood. I wonder if the Rolls gets wet, does it come apart?
-Chris
On that we will have to dissagree.
Your analogy is not complete & incorrect. You do not have the same force to rock them As the energy available to rock the car is inversly related to the mass (increase mass, resonant frequency drops, available energy increase ~ by the square of the decrease), you have a 3 year old pushing the Smart, and hulk hogan pushing the Rolls.
dave
Hi Dave,
Seems we disagree on many things.
Okay, you run a sweep at 2 watts RMS constant. Sweep begins at 40 Hz and terminates at 2 KHz (for example). The identical amplifier, wires, signal source and driver are used. The audio source runs a programmed sweep and Dave is locked away from the controls so he can not modify any aspects of the test. That should take care of most the variables. Oh, assume a constant room temperature, voice coil temperature (I know - unlikely as heck) and air pressure. No funny stuff.
Now what we have is a driver exerting the same force exactly in all instances. The car analogy would hold if Dave is hypnotized and doing precisely what he is told (I know - unlikely as heck).
Dave, the energy stimulating the system does not change, else you have found a free energy source! The only thing that changes is the amount of effect this force has on each object. If you want to get silly and bury the concept in a myriad of details, at resonance, the impedance of the driver increases to a high level. The amount of energy delivered to that driver has now decreased. Let's not be silly and lose sight of the idea here.
-Chris
Seems we disagree on many things.
? I'm sorry but you lost me completely. Help me out to understand how your applied force is changing. That's the same thing as changing the rules as we play to be honest with you! Take note, the amount of energy we are sending into the system is constant. Even if you run into a resonant point, energy input remains the same, only the response to this energy would change as in the case of a resonant point that you are describing. In other words, energy in = some response.
Okay, you run a sweep at 2 watts RMS constant. Sweep begins at 40 Hz and terminates at 2 KHz (for example). The identical amplifier, wires, signal source and driver are used. The audio source runs a programmed sweep and Dave is locked away from the controls so he can not modify any aspects of the test. That should take care of most the variables. Oh, assume a constant room temperature, voice coil temperature (I know - unlikely as heck) and air pressure. No funny stuff.
Now what we have is a driver exerting the same force exactly in all instances. The car analogy would hold if Dave is hypnotized and doing precisely what he is told (I know - unlikely as heck).
Dave, the energy stimulating the system does not change, else you have found a free energy source! The only thing that changes is the amount of effect this force has on each object. If you want to get silly and bury the concept in a myriad of details, at resonance, the impedance of the driver increases to a high level. The amount of energy delivered to that driver has now decreased. Let's not be silly and lose sight of the idea here.
-Chris
to excite the resonance the energy has to be supplied within the bandwidth of the resonance, High frequency resonance, small bandwidth, music. Not much energy. Low frequency resonance, wider bandwidth, more energy available
Pesronally i listen to music. 2 watts constant RMS is artifical. Still the experiment would be interesting.
One could take the concept even further and try to place resonances in places where typical music has little energy using the well tempered scale.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.