How about a membrane that is firmly attached in its sides wrt. how much it can actually move there compared to the centre. Is it really OK to inject the same force close to the fixed sides as in the centre? Does trying to move a membrane 1mm in both centre and rim (same current / same foil / same F) actually cause some sort of distortion for the close-to-rim case? This is something I have pondered on for some time...
//
//
If you look at the first posts, you'll see that there is a fair amount of distance between the mount of the membrane and the moving parts.
So there will be some suspension due to that distance alone. But, yes that will be a huge source of distortion.
But I hope counter measures such as dampening of the free membrane will mitigate that.
The mid corrugated membrane will have an effective width of 8 cm. The membrane area will then be 220x8 = 1760 cm2, that is, comparable with an 18' driver.
Although not directly comparable as my speaker is a line source, a 18' speaker only needs to move a fraction of a millimeter to reproduce 100 dB SPL:
My concern is, as I want a rigid moving membrane which I hope to achieve with corrugation, the membrane already are at +/- 0.5 mm due to the corrugation to begin with. Hence the +/- 1 mm simulations.
The tweeter will only have an effective area of 0.5 x 220 cm = 110 cm2, which corresponds to a 4.5' driver.
So I will roughly need the same amount of movement:
As for bass membranes, where for example the suspension would be a even greater problem, I think that you're better off with traditional speakers.
I myself rely on my six AE TD15H to deliver the bass and sub bass with really low distortion:
I used this calculator.
So there will be some suspension due to that distance alone. But, yes that will be a huge source of distortion.
But I hope counter measures such as dampening of the free membrane will mitigate that.
The mid corrugated membrane will have an effective width of 8 cm. The membrane area will then be 220x8 = 1760 cm2, that is, comparable with an 18' driver.
Although not directly comparable as my speaker is a line source, a 18' speaker only needs to move a fraction of a millimeter to reproduce 100 dB SPL:
My concern is, as I want a rigid moving membrane which I hope to achieve with corrugation, the membrane already are at +/- 0.5 mm due to the corrugation to begin with. Hence the +/- 1 mm simulations.
The tweeter will only have an effective area of 0.5 x 220 cm = 110 cm2, which corresponds to a 4.5' driver.
So I will roughly need the same amount of movement:
As for bass membranes, where for example the suspension would be a even greater problem, I think that you're better off with traditional speakers.
I myself rely on my six AE TD15H to deliver the bass and sub bass with really low distortion:
I used this calculator.
Last edited:
First attempts, one fourth of the aluminium foil of mid and tweeter membranes.
I wanted to find out if the Silhouette Cameo would cope with a 220 cm long membrane, that is four times longer than this test.
I also wanted see if the theoretical resistances matched the real ones.
It has been a long time since I last cut aluminium foil, so the quality wasn't that good; I didn't manage to get the aluminium foil glued properly to the paper.
The mid part that has two sections is to the left and the tweeter is to the right.
Both parts have 10 mm aluminium strips on each side. They are for mounting purposes and will also help to keep the corruation.
Based on this test cut, I would say that 220 cm long membranes are feasible on the Silhouette.
One problem tough, there must a 220 cm long clearance in front and at the back of the cutting machine.
The real resistance one of the two mid sections was just shy of 4 Ω.
The four times longer full membrane will then be 8 Ω if the sections are coupled in parallel.
In theory the resistance should be 2.8 Ω for one 220 cm long and 3 mm wide aluminium strip.
That length correspond to one section of the mid above:
The tweeter part to the right was supposed to be two 1.5 mm strips 0.5 mm apart.
That didn't go well; only one strip survived.
But I now know that a 220 cm long and 3 mm wide strip has a resistance of 4 Ω, so I don't have to worry about that.
I wanted to find out if the Silhouette Cameo would cope with a 220 cm long membrane, that is four times longer than this test.
I also wanted see if the theoretical resistances matched the real ones.
It has been a long time since I last cut aluminium foil, so the quality wasn't that good; I didn't manage to get the aluminium foil glued properly to the paper.
The mid part that has two sections is to the left and the tweeter is to the right.
Both parts have 10 mm aluminium strips on each side. They are for mounting purposes and will also help to keep the corruation.
Based on this test cut, I would say that 220 cm long membranes are feasible on the Silhouette.
One problem tough, there must a 220 cm long clearance in front and at the back of the cutting machine.
The real resistance one of the two mid sections was just shy of 4 Ω.
The four times longer full membrane will then be 8 Ω if the sections are coupled in parallel.
In theory the resistance should be 2.8 Ω for one 220 cm long and 3 mm wide aluminium strip.
That length correspond to one section of the mid above:
The tweeter part to the right was supposed to be two 1.5 mm strips 0.5 mm apart.
That didn't go well; only one strip survived.
But I now know that a 220 cm long and 3 mm wide strip has a resistance of 4 Ω, so I don't have to worry about that.
But I haven't been able to do one yet, I only stated that based on the 55 cm test it looks feasible.
I just got a new paper crimper. It is 21.6 cm wide (that's 8.5'' for non SI guys) compared to the old one (10.5 cm) that I used for my ribbons.
The new one only has 5 mm between two tops while the old has 4 mm, but I guess that's of no significance.
The new one seems to make stronger corrugation though; looking at the pocket depths they are slighter deeper, 0.7 mm compared to the old one's 0.5 mm.
I don't know if that's is enough. Otherwise I'll have to make my own corrugator.
I just got a new paper crimper. It is 21.6 cm wide (that's 8.5'' for non SI guys) compared to the old one (10.5 cm) that I used for my ribbons.
The new one only has 5 mm between two tops while the old has 4 mm, but I guess that's of no significance.
The new one seems to make stronger corrugation though; looking at the pocket depths they are slighter deeper, 0.7 mm compared to the old one's 0.5 mm.
I don't know if that's is enough. Otherwise I'll have to make my own corrugator.
i believe (and mine is not correct either 
) a less deep corrugation would be better for tweeter duty. i only have one size that i use. the old one is even bigger never measured them to be honest well since i got a 3d printer i think thats the first thing ill try... not sure if it has the accuracy though , but you could make funny shapes to
) a less deep corrugation would be better for tweeter duty. i only have one size that i use. the old one is even bigger never measured them to be honest well since i got a 3d printer i think thats the first thing ill try... not sure if it has the accuracy though , but you could make funny shapes to
Yes, horizontal. That is, from left to right when the membrane is mounted in the finished speaker.
The goal with the corrugation is to make the membrane from left to right to be rigid as a beam so that the non aluminium parts between the strips move exactly as the aluminium parts do.
Sorry båndsei, but I can't that happening in these membranes.
I agree, but the shallow rhomboid pattern reduces the aluminium's own voice (very important IMO). The second "usual" corrugation can give more stiffness at cross direction, keeping the first's sonical benefits (I hope).
The membranes that båndsei posted I believe are ribbons with aluminium foil strips almost on the entire membrane.
So the non aluminium parts will probably not pose any serious problems; the aluminium's own voice being a bigger problem.
But for the planar I'm planning the aluminium foil strip only covers 30 %; 3 mm aluminium foil and then 7 mm plastic only.
So there will for sure be several eigenmodes if the membrane is not corrugated as I have described above.
For sure, I do not know at this stage if the corrugation will be enough. That's why I will test different plastics and even sandwich designs.
So the non aluminium parts will probably not pose any serious problems; the aluminium's own voice being a bigger problem.
But for the planar I'm planning the aluminium foil strip only covers 30 %; 3 mm aluminium foil and then 7 mm plastic only.
So there will for sure be several eigenmodes if the membrane is not corrugated as I have described above.
For sure, I do not know at this stage if the corrugation will be enough. That's why I will test different plastics and even sandwich designs.
This is absolutely relevant opinion. The Magnepan covers the planar-membranes a lot of elastic layer. This provides coupling between driven and undriven areas, but there is too much added moving mass. At first, you can listen the small prototype without any surface post-treatment. I guess, that will be surprisingly good.
After a week of trials with the corrugation, I realize that it is not feasible to corrugate a 220 cm long membrane with 1 mm creases.
Even if, the corrugation doesn't hold in the bare plastic parts; those without the aluminium foil.
So it will probably have to be a tensioned membrane:
Before I start tedious experiments, I would like to simulate the membrane's behaviour with respect to amplitue and frequency of the driven aluminium parts.
The membrane could be BOPP, PEN or PET of various thicknesses.
The aluminium foil is 7 µm.
The fastening should not pose any problem to set up, the dampening does.
The aluminium parts are subject to a force that results in a movement that varies with amplitude and frequency.
How the non aluminium parts of the membrane behaves is the goal of the simulation besides finding a suitable dampening.
There is also the aspect of the tensioning and for sure other aspects that I'm currently not aware of.
Can this simulation be set up in MATLAB/Octave or does it require more powerful simulators such as COMSOL?
Even if, the corrugation doesn't hold in the bare plastic parts; those without the aluminium foil.
So it will probably have to be a tensioned membrane:
Before I start tedious experiments, I would like to simulate the membrane's behaviour with respect to amplitue and frequency of the driven aluminium parts.
The membrane could be BOPP, PEN or PET of various thicknesses.
The aluminium foil is 7 µm.
The fastening should not pose any problem to set up, the dampening does.
The aluminium parts are subject to a force that results in a movement that varies with amplitude and frequency.
How the non aluminium parts of the membrane behaves is the goal of the simulation besides finding a suitable dampening.
There is also the aspect of the tensioning and for sure other aspects that I'm currently not aware of.
Can this simulation be set up in MATLAB/Octave or does it require more powerful simulators such as COMSOL?
Last edited:
Not the answer to the above but yet another visualization of the force on a membrane when it moves from zero position to 2 mm, that is, the full gap.
Here's the LUA script:
A video of script running:
Output file:
Here's the LUA script:
-- Calculate the force on the membrane.
file_name = "mid 4 columns 5 mm apart gap 4 mm"
femm_file = file_name.. ".fem"
result_file = file_name .. ".txt"
showconsole()
mydir="./"
open(mydir .. femm_file)
print(" mm, N\n")
result = openfile(result_file,"w+")
write(result, file_name .. "\n" .. " mm N\n")
mi_saveas(mydir .. "temp.fem")
mi_seteditmode("group")
for n=0,20 do
mi_analyze()
mi_loadsolution()
mo_showdensityplot(1,0,1,0,"bmag")
mo_groupselectblock(1)
fz=mo_blockintegral(12)
-- if (n<20) then
mi_selectgroup(1)
mi_movetranslate(0,0.1)
s = format("%.2f, %f",n/10,fz)
print(s)
r = format("%.2f, %f",n/10,fz)
write(result, s .. "\n")
-- end
end
mo_close()
mi_close()
closefile (result)
A video of script running:
Output file:
Code:
mid 4 columns 5 mm apart gap 4 mm
mm N
0.00, -0.319454
0.10, -0.319171
0.20, -0.319331
0.30, -0.319249
0.40, -0.319226
0.50, -0.319402
0.60, -0.319457
0.70, -0.319382
0.80, -0.318672
0.90, -0.318313
1.00, -0.317683
1.10, -0.317006
1.20, -0.316217
1.30, -0.314860
1.40, -0.312948
1.50, -0.310126
1.60, -0.308473
1.70, -0.304684
1.80, -0.301662
1.90, -0.296963
2.00, -0.292433When trying to do the above for a any other than a 3 mm wide foil, I get errors.
Here's how that look like:
In the video I abort with ESC key when the progress bar stops.
I've tried every tip I could find including shorten the problem depth from 2200 mm to 22 mm; nothing helps.
Here's the zipped lua script and femm-file if some one would care to take a look at it.
Here's how that look like:
I've tried every tip I could find including shorten the problem depth from 2200 mm to 22 mm; nothing helps.
Here's the zipped lua script and femm-file if some one would care to take a look at it.