could you please comment on the type of hot melt adhesive that you used and also on how you came to the number of contact points and if contact point orientation with respect to the panel has any impact? Thank you.
Hi Moray, Sorry for the delayed reply.
The adhesive was just a blob to secure the disk to the EPS. It was the closest thing to hand. It's the cheapest type you get at any hardware store.
The number of contact points started at many, and I simply pared them down until I got a decent response. There was no theory behind the solution except trail and error.
I did not test for orientation, but I imagine that, given the size of the disk compared to the size of the panel, that it would make much difference.
No worries I appreciate you responding. I was hoping to get an idea of the hardness of the hot melt adhesive that you used as they vary in hardness and in what they will stick to but hardness was foremost on my mind. So I will assume that this was a general purpose type of hot melt which is firm to hard when set. What prompted the question in the first place was discussion of using a compliant interface between the voice coil and the panel. forget now where I read that.
In this instance I see the hot melt adhesive as acting as a filter of sorts and I am suspicious that contact orientation with respect to the panel might have more impact than you think but short of trial and error I don't know how one would go about discovering what would be optimum. If that difference is a big one or a little one I could not say. Thanks for posting so much to think about and it is useful to hear the thoughts of others to expand the horizon.
In this instance I see the hot melt adhesive as acting as a filter of sorts and I am suspicious that contact orientation with respect to the panel might have more impact than you think but short of trial and error I don't know how one would go about discovering what would be optimum. If that difference is a big one or a little one I could not say. Thanks for posting so much to think about and it is useful to hear the thoughts of others to expand the horizon.
Yes, this must be analysed with an FEM modeller to find out exactly what's going on. The combination of the different masses, densities and surface areas of the materials concerned, and as you point out, possibly the compliance of the connecting legs between the components, makes this a horrendously complex system. "kots in n tuimeldroër", as they might say in the local vernacular. Due to the shortness/thickness/hardness of the hot-melt legs, I suspect that the filter effect is rather to do with the softness of the EPS where it interfaces to the hot-melt.
I think Jamie Nelson was doing something along the same lines, just a lot more civilised, with his ortho-planar spring solution?
that is the trouble when designs all of a sudden get a curve ball thrown at them and as the designer you don't know what to look at first because the multiple combinations of materials could respond very differently than any one of them alone would do so the situation could in the end be as simple as your first guess or it could be a real head scratch-er and or a real time waster.
You do a simple experiment to see what the hot melt is doing and make a hot melt gasket/washer to be place under the voice coil to simulate the multi leg hot melt attachment set up just to get a feel for how it makes a panel sound on its own. Not the whole picture I know but it could provide you with enough for a better sense of direction without spending a lot of time right away juggling a more complex situation. Will look forward to watching to see what you do next and how it all works out. Thanks for chatting.
You do a simple experiment to see what the hot melt is doing and make a hot melt gasket/washer to be place under the voice coil to simulate the multi leg hot melt attachment set up just to get a feel for how it makes a panel sound on its own. Not the whole picture I know but it could provide you with enough for a better sense of direction without spending a lot of time right away juggling a more complex situation. Will look forward to watching to see what you do next and how it all works out. Thanks for chatting.
I have actually used hot melt glue to mount the drivers to the panels. I get no difference in response between superglue, double-sided tape and hot-melt. The hot melt I use sets rather hard, there's no give to it at all.
But, because of the drop in sensitivity, I shelved this idea. However, I still go for multiple density solutions in the panels I'm trying to finalise. So many variables... I wish there was a Thiele Small equivalent for DML panels.
That is interesting to know. When you said there was no difference in the frequency response between these various adhesives when mounting the voice coil I assume you mean in the measured frequency response? If that is correct could I ask did you listen to each of those materials to compare what they sounded like as well? If so great.
I wanted to add that theses voice coil joints are all very thin in cross section and with the puck mounting block the gap between the puck and the panel is substantial. That is why I suggested making a gasket that has a substantial thickness to more closely simulate the mounting process of the decoupled puck.
Thread cleaned up. Some posts removed.I mean I shelved the idea of using a disk and EPS, not the idea of using hot-melt to mount the drivers.
am curious do you prefer driver mounting with hot melt to PVA adhesive? Thank you.
Moray, while I'm testing and prototyping I use cheap double-sided tape for easy-on, easy-off. For long-term adhesion I used to use hot-melt, but due to the high wattage I use, the driver/VC heats up, and the glue gets soft. These days, once I'm happy with material selection and positioning, then I use contact adhesive and let it cure overnight.
However, this past weekend I discovered a no-name (Chinese?) 1mm clear rubber double-sided tape which seems to be the same type that the drivers are supplied with from the factory. It's VERY strong, and I might be using this in the future for long-term adhesion if it survives the abuse I throw at it.
I did notice a bit less treble when using the supplied adhesive tape. I think it might help for HF reproduction, especially when using EPS, to have a hard bond to the plate which epoxy will give. I have a roll of 50mm wide 3M 9088 tape, which seems pretty much identical to the tape that the exciters come with. I think it is the thinnest double-sided tape they have at 0.21mm. A bit costly for a roll of tape, but has been worth every penny. Very useful for doing quick test mounting and for lots of other things 
I do think sticky tape can be able to handle a lot of abuse. It is not as hard and will give after a bit more, meaning it probably will cause the exciters to heat up less, and it will not crack from pressure.
I do think sticky tape can be able to handle a lot of abuse. It is not as hard and will give after a bit more, meaning it probably will cause the exciters to heat up less, and it will not crack from pressure.
I was picking up a 10kHz peak (or is it a wide-band 5kHz dip) when using 0.05mm double-side tape or hot glue or superglue or anything else with XT32's on 3mm Albasia.
The peak disappeared with the DAEX25FHE, but the dip did not.
I chased down the rabbit trail of cavity resonances. Wrong rabbit trail.
I changed the hard adhesives to the (1mm) softer tape which I suspect is the same as the original. I thought that the manufacturer might have specifically chosen this tape to kill the 10kHz peak.
Nope. No difference.
I eventually found the problem to be in the Albasia itself. I changed to Nidaplast/PVA/Resin, and I'm getting better curves even with the soft, thick tape.
Maybe the softer, thicker tape has non-Newtonian properties??
I imagine the thicker softer tape provides much more damping than hard adhesives. Having this small CLD patch near center of panel likely provides some improvement. With the right tape thickness, panel rigidity & lightness, and mount surface area the tape may not interfere too much with the translation component of the driver moving panel while still helping significantly to damp the ringing caused by the bowing of the panel.
Hi Andre, I have been having a peak around 9-10KHz with XT32's on my HD EPS panels too, I am using epoxy to attach them rather than the tape. What I found helped to tame it is a small piece of EPDM foam tape centred on the exciter. I am clamping the panels in the frame using 12w/12d/25l mm strips of the EPDM foam tape, 3 places long sides and 2 on short sides, I have found having the small strips to clamp good as I can tune it slightly by moving them and adding more places if it seems to help.
Attachments
Hello André,
It is possible I didn't read carefully enough previous posts...
What is "Albasia"?
Some readings I had recently recalled me different sources of HF problems :
- the peaks like the one seen at about 10kHz with EPS for example which is called "drum effect". It is the central area that acts like a drum. The panel is like a dual cone speaker. When you changed from the XT32 to the DAEX25, the diameter was reduced so maybe that frequency pushed to a higher value. What I didn't have in mind before those readings it is before the peak, this drum effect has a dip but it seems it is not the one you detected. Is it?
- the second source of HF limitation is in the chain that transmits the force from the coil to the plate. The first element pointed in a B Zenker's paper is the stiffness of the coil former. He didn't described against which element this stiffness react to create a filter or a resonator. The tape or any glue is also in this chain acting like a stiffness and a damper. In this paper, B Zenker recommands to make an impedance measurement with the panel laying on a support (a plan) with the magnet free. The frequency of the "accident" in the force chain appears as a small peak in the impedance curve where we are suppose to see only the coil inductance effect. To be tested maybe?
Hello Joska,
See my post just before that one. The 9-10k peak seems to be the drum effect of the central area. The difficulty with EPS, it is this effect is desirable to extent the band of the panel. Do you have a picture or a drawing of how the EPDM foam tape is added to damped it?
Last week I tested a 20mm thick light EPS I have had for a long time. It is 80cm long. The width started from 60cm and I cut it to 50, then 40 and finally 35cm. It was suspended the minimum as Steve recommends to start with a strip of masking tape and a sponge as damper. I was surprised by the smoothness of the response, not flat but a nice smooth curve and by the fact it didin't go really below 300Hz (from memory). Your FR is not smooth and goes quite low... Could you remind me the dimensions and thickness? Have you tested it with a simple strip of tape?
Christian
Hi Christian,
I do vaguely remember people discussing the drum affect but it was before I had played around making them and had a mic to measure so didn't pay much attention.
The foam tape has adhesive one side so it is just stuck central on the VC area on the front of the plate. These are approximately 300/500mm 20mm thick 28kg/m eps with a light hide glue + shellac coating. From memory I have tried one in free air held between fingers and had not a lot of output below 300Hz.
The high pass is just a first order capacitor in series to limit the low bass energy they don't produce sound from anyway, seems to only drop 3db but might help I thought.
Yes the FR looks bumpy at this scale but they sound fantastic and the peaks/dips are less than 5db from a central average so don't seem extreme. If I zoom out to a 120db scale and make it 1/3 smoothing they look better haha. Also a quick measurement indoors so probably could be done more carefully.
Joska
I do vaguely remember people discussing the drum affect but it was before I had played around making them and had a mic to measure so didn't pay much attention.
The foam tape has adhesive one side so it is just stuck central on the VC area on the front of the plate. These are approximately 300/500mm 20mm thick 28kg/m eps with a light hide glue + shellac coating. From memory I have tried one in free air held between fingers and had not a lot of output below 300Hz.
The high pass is just a first order capacitor in series to limit the low bass energy they don't produce sound from anyway, seems to only drop 3db but might help I thought.
Yes the FR looks bumpy at this scale but they sound fantastic and the peaks/dips are less than 5db from a central average so don't seem extreme. If I zoom out to a 120db scale and make it 1/3 smoothing they look better haha. Also a quick measurement indoors so probably could be done more carefully.
Joska