I Agree completely, but this design was never aimed at optimal sound but an attempt at an acceptable compromise of sound and minimallistic design to achieve the appearance of a floating panel with the least amount of visible frame.
You can find out about the motivation on page 103, from post #1023 to #1028.
I'm definitely not dismissing framed designs, and If I built a vertically hanging or standing speaker I would likely try a frame again with the rubberfoam attached to the outer edge.
While fun to try and a good leaning experience, it turns out the compromise was not worth it in this case.
I'll move on to replace the rubberfoam with fishing line, and I'm thinking a discrete solution is to epoxy glue it on the edge of the board in each corner and attach the other end to the wooden frame using the screw hooks. I will need a test rig to suspend it horizontally.
The reason why I need the frame and not attaching the other end of the fishing line directly in the ceiling is because I have acoustic panels glued directly to islands of the ceiling, and I'm not going to put holes in those. I realize this makes my design more complicated and less useful for anyone else on here, sorry about that!
I'll try and test what impact the wooden frame has in my future horizontal mount test rig and share my experience.
If it has measurable impact I could probably remove some material from the MDF cross or put holes in it (or both).
Thanks for the feedback!
Last edited:
I cut two different size panels. First pair is 48inches height X20inches width. Second pair is 40inches height X 24inches width. As soon as I lifted up one of each panel in my hand it hit me. Guess which pair bends more?
I am getting old I think I need to record and write down everything Ive learned about these panels as my memory is not what it use to be.
I found the answer that I already knew but forgot. SMDH
Its just plain ole physics in height X width X thickness ratios. When height and width are equal there are no weak points so the shape of the material is stronger/stiffer. When the height and width are no longer equal then it creates weak points causing panel to bend more the further height and width the ratios become.
Taller aka longer in length panels bend more especially if the width is half of that of its length or less. Lesser the width the more it bends. Thickness also plays a major role in bending. Thinner the panel the more it bends.
Ideal panel material.
1. needs to be stiff or have a high density. Meaning you can step on it and it wont collapse. This will increase high frequency response
2. Needs to be very light pound for pound. This increases efficiency in SPL output.
3. Needs to bend easily. A materials height, width and thickness can contribute to its ability to bend more easily.
The two ideal materials that fit all three criteria is first Honeycomb composites and second is HIGH density EPS/XPS. Both materials are stiff, very light and bends easily.
Wood is stiff and hard so it meets the first criteria. Down fall to wood is that its not as light as HCC or EPS/XPS so less efficient. The second down fall is that wood does not bend as easily as HCC or EPS so it has less bass.
Great feedback DMLBES, many thanks for the detail, it is much appreciated. You are not alone in getting old! The difference in stiffness on length v width is a very interesting observation.
I am going to stick with ply for now as I have had good results from it to date and, probably due to my lack of experience, not as much success with EPS/XPS. I think that honeycomb composite is ideal but my pockets are not that deep.
Thanks again.
Burnt
I am going to stick with ply for now as I have had good results from it to date and, probably due to my lack of experience, not as much success with EPS/XPS. I think that honeycomb composite is ideal but my pockets are not that deep.
Thanks again.
Burnt
I cannot let this slip by un-noticed if it does make a fairly substantial difference. Why would flexible glue (2 part epoxy in this case) be better? My gut feel says the exciter should be glued as rigidly as possible to the panel in order to transfer vibrations accurately. Flexible glue will dampen this action somewhat, or am I missing something? What is best practice here?
Tall Story 3
This is the episode where Buzz Lightyear gets over confident and Woody comes to the rescue-again.
So this idiot Buzz was planning to release an Mpg4 to the thread so people could hear the sort of performance he was getting. Being Buzz Lightyear he aimed for Infinity and Beyond, leaping straight to the end solution without testing. He crashed and burned of course.
Woody, his faithful free standing ply panel, came to the rescue reminding Buzz that 40 hz had not been a fluke and also providing an afternoon of music to while away the time.
Here are the learns, or scars, as Buzz calls them.
For a framed panel dimension 1160mm by 300 mm using 4mm Birch Play BB grade
1. A non-framed or damped panel gives strong output down to 40 hz and will reach 30 hz
2. An 18mm wide strip of foamed double sided tape 1mm thick applied along both long edges reduced this to 70 hz, stealing nearly an octave. You can feel the width is too great making the bond too stiff despite the foam. I knew I had made a mistake the moment I did it but you can never tell Buzz any damn thing.
3. A 10mm wide strip of foamed double sided tape 4mm thick restores the framed panel to 40 hz. (phew!)
This is after an afternoon spent removing 9.280 meters of very sticky tape. yes, it stuck to both sides. It came away cleanly in the end, no damage to frame or panel.
So the trick is to support the panel to the frame with the minimum stiffness and ensure it has maximum rotational freedom at the join.
in summary, for this design of frame.
1. 18mm wide tape at 1mm thick applied along both 1160mm lengths gives 70 hz
2. 10mm wide tape at 4mm thick applied along both 1160mm lengths gives 40 hz
This actually corresponds very well to an earlier discovery when Woody was in the horizontal position in my first tests. I got deeper bass when the panel and the damping foam was lightly loaded with just the panel and exciter mass. When I applied heavier loading to the panel, via Wittgenstein’s Tractatuss translated into English applied as a reaction mass, I got less bass.
It is possible bass doesn't like German philosophy but I doubt it.
I forgot this important learn
What an idiot.
Things to consider
There are some other variations on mounting method that might be useful to trial. I can think of two noted below.
1. Use foamed tape to fix the panels to the corners of the panel but leave the panel sides clear. -You will need to use a spacer of circa 3mm as well as the 1mm thick tape to make sure the panel clears the frame, any contact between them and its Buzz city.
2. Use the 1mm thick tape on both long sides of the panel as above but only remove the plastic protecting the adhesive every 100mm for a space of one 20mm so that the panel is ‘spot welded’ rather than continuous. The adhesive will hold the panels and the non-adhered sections are free to move. The foam will still prevent the panel buzzing on the frame.
Thats all for now. Tomorrow if I get the chance I will make that mpg4.
I am going to stick Buzz back in his box and go to bed.
Burnt
This is the episode where Buzz Lightyear gets over confident and Woody comes to the rescue-again.
So this idiot Buzz was planning to release an Mpg4 to the thread so people could hear the sort of performance he was getting. Being Buzz Lightyear he aimed for Infinity and Beyond, leaping straight to the end solution without testing. He crashed and burned of course.
Woody, his faithful free standing ply panel, came to the rescue reminding Buzz that 40 hz had not been a fluke and also providing an afternoon of music to while away the time.
Here are the learns, or scars, as Buzz calls them.
For a framed panel dimension 1160mm by 300 mm using 4mm Birch Play BB grade
1. A non-framed or damped panel gives strong output down to 40 hz and will reach 30 hz
2. An 18mm wide strip of foamed double sided tape 1mm thick applied along both long edges reduced this to 70 hz, stealing nearly an octave. You can feel the width is too great making the bond too stiff despite the foam. I knew I had made a mistake the moment I did it but you can never tell Buzz any damn thing.
3. A 10mm wide strip of foamed double sided tape 4mm thick restores the framed panel to 40 hz. (phew!)
This is after an afternoon spent removing 9.280 meters of very sticky tape. yes, it stuck to both sides. It came away cleanly in the end, no damage to frame or panel.
So the trick is to support the panel to the frame with the minimum stiffness and ensure it has maximum rotational freedom at the join.
in summary, for this design of frame.
1. 18mm wide tape at 1mm thick applied along both 1160mm lengths gives 70 hz
2. 10mm wide tape at 4mm thick applied along both 1160mm lengths gives 40 hz
This actually corresponds very well to an earlier discovery when Woody was in the horizontal position in my first tests. I got deeper bass when the panel and the damping foam was lightly loaded with just the panel and exciter mass. When I applied heavier loading to the panel, via Wittgenstein’s Tractatuss translated into English applied as a reaction mass, I got less bass.
It is possible bass doesn't like German philosophy but I doubt it.
I forgot this important learn
What an idiot.
Things to consider
There are some other variations on mounting method that might be useful to trial. I can think of two noted below.
1. Use foamed tape to fix the panels to the corners of the panel but leave the panel sides clear. -You will need to use a spacer of circa 3mm as well as the 1mm thick tape to make sure the panel clears the frame, any contact between them and its Buzz city.
2. Use the 1mm thick tape on both long sides of the panel as above but only remove the plastic protecting the adhesive every 100mm for a space of one 20mm so that the panel is ‘spot welded’ rather than continuous. The adhesive will hold the panels and the non-adhered sections are free to move. The foam will still prevent the panel buzzing on the frame.
Thats all for now. Tomorrow if I get the chance I will make that mpg4.
I am going to stick Buzz back in his box and go to bed.
Burnt
Last edited:
He should of left out the word "flexible". All glues are flexible to a degree and all glues or any adhesive used will dampen it to a degree.
3m tape will fail over time, its best used when one is experimenting. Glues are used instead when one wants to mount a exciter permanently.
I said this many times before that even the type of adhesive used can affect the sound. Ive been using 5min. 2 part epoxy for over 4 years to permanently adhere my exciters to the panel.
When using any other type of glue make sure you remove the 3m tape first.
With EPS and XPS they are very sensitive to corrosive chemicals so certain types of glue with corrosive properties cannot be used.
Even Tech ingredients uses 2 part 5min. epoxy, but he uses it in slightly different way which is mind boggling and proves he aint as smart he thinks he is and or lacks common sense. LMAO He says he uses the epoxy to pre seal the surface of the balsa material so that the sticky tape can adhere better. WTF? Why not just use the 2 part epoxy glue as the adherent and remove the sticky tape? SMDH LOL
I 'guess' he wanted to be able to remove them later.
I don't think so or he would of mentioned just using 2 part epoxy if one wanted it to be permanent.
Sticky tape is used for experiments because it can be removed fairly easily and will fail eventually and fall off.
Glue is used for final product because its permanent and wont ever fall off.
Besides that like I said before different types of adhesive will sound different.
3m tape is soft....Epoxy is way harder then 3m tape. Harder surface equals better highs. If the double sticky tape is too thick it will over dampen contact area.
Any type of adhesive will result in damping to a degree some more then others depending on certain factors. If using more then one type of adhesive it will over dampen the contact area of the voice coil to the panel. So using epoxy combined with sticky tape will have more damping then just a single adhesive.
Thank you DMLBES, I think you are correct – the word flexible should have been left out. All 2 part epoxy glues I’ve ever used sets to rock hard. Pity one cannot use 2 part epoxy for solving other problems when getting older, but let’s not go there.
I think there is merit in doing that, especially if the panel surface is not perfectly flat (i.e. coarse plywood) or something softer like balsa.
Thanks for the updates Burntcoil. Waiting for future episodes. I had a look at Veleric's build again (around post 843 I think) and he did not use foam tape along the total length of the edges. So, I think your "spot welding" idea sounds like a good option to try. I think DMLBES also did not use foam along the total length of the edges.
Hi Burnt,
Thanks again for the additional info. As well as the width of the tape, is it possible the thickness (1mm v 4mm) makes a difference?
Regards, Ian.
twocents,
That's correct, on mine I used foam that covered only about 60% of the long edges and 75% of the short edges. Meanwhile I have done more testing with foam applied in many configurations (both less and more). In my tests, I didn't find much difference between the 60%/75% configuration and using foam around the full perimeter, both (and anything in between) seems to work best for me.
So far I did not find any advantage trying to use less support. Putting foam only at the corners, or only at the midpoints, or free hanging with no supports, all resulted in more sharp deep dips and big narrow spikes in the frequency response below about 200 Hz.
Eric
I for one suspect the thickness made a big (and maybe even bigger) difference. He might even want to try thicker foam.
Using thin, wide foam tape approaches what engineers consider to be a "clamped" support, where the edge can neither displace nor rotate. On the other hand, a thick narrow foam approaches a "simple" or "hinged" support, which prohibits displacement, but allows rotation.
I've modeled both edge conditions for a particular panel, and the fundamental frequency of the simply supported panel is about half that of the clamped panel. Additionally, the wide tape effectively makes the panel "smaller". So it's not surprising that using thicker, narrower foam provides audible response at lower frequencies.
Eric
Burnt,
For clarification, are you using REW or similar for your evaluations, or other?
For a long time I was using a series of test tones or frequency sweep with an frequency generator, and listening to the response to do my "testing". Then I started using white and pink noise with a spectrum analyzer. Finally I started using REW and realized how much detail and nuance that I was missing with the other cruder, methods. In particular, I was finding that there were some configurations that reached lower frequencies, but had holes in the response at higher frequencies, that I wasn't really picking up on.
I hope this does not sound critical or otherwise offensive. I just fear that you and others may be misleading themselves (like I was!) before I started using REW.
If you are already using REW or similar, nevermind!
Eric
For clarification, are you using REW or similar for your evaluations, or other?
For a long time I was using a series of test tones or frequency sweep with an frequency generator, and listening to the response to do my "testing". Then I started using white and pink noise with a spectrum analyzer. Finally I started using REW and realized how much detail and nuance that I was missing with the other cruder, methods. In particular, I was finding that there were some configurations that reached lower frequencies, but had holes in the response at higher frequencies, that I wasn't really picking up on.
I hope this does not sound critical or otherwise offensive. I just fear that you and others may be misleading themselves (like I was!) before I started using REW.
If you are already using REW or similar, nevermind!
Eric
You are correct I did this technique around 4 years ago and still do it today. Instead of one long continuous strip of foam tape I cut them into 2-4inch pieces spacing them 2-4inches apart. This way there is less damping and panels still have some freedom to move.
Like I said before in a previous post my design is like a sandwich with foam holding on both sides of the panel. Its not as stiff as a clamp and or as loose as a hinged support but more of in the middle to make it "JUST RIGHT".
Also one has to remember that except for the spline on the back side everything else has to be equal or else you will change the radiating sound pattern. I am not saying its a bad thing as its all personal preference but when you keep everything as equal as possible it will be the most coherent.
Last edited:
Podiums>width 23inch height 55inch
Melior one> width 22inchs height 45inchs
Bertagni various models width 19-22inchs height 29-55inches
Yamaha NS10/20 ear speakers width 24inches.
If you look at all those commercial built speakers width they are all around 19-24inches. I highly doubt its just coincidence. The height is also usually double the width.
Magnepan LRS is 14.5incheswidth. and 48 height.
Melior one> width 22inchs height 45inchs
Bertagni various models width 19-22inchs height 29-55inches
Yamaha NS10/20 ear speakers width 24inches.
If you look at all those commercial built speakers width they are all around 19-24inches. I highly doubt its just coincidence. The height is also usually double the width.
Magnepan LRS is 14.5incheswidth. and 48 height.
Feedback is always welcome Eric!
I am not using REW so your concern might be valid. I have a spectrum analysis result and some MP3's I have been trying to share to the site but I still have not worked out how to meet the site criteria yet. I am using jpeg and MP3 which should load but no luck yet. I will persevere and then you will have a recording of the raw panels, no eq, recorded in a room.
Back latter with data, I hope!
Best to all
Burnt
Sorry I thought I had responded to this. Yes I got much better results with thicker tape-3mm thick, 10mm wide. I used three strips per side each strip 100mm long. Way better response all round.
Burnt