Yup, the plan is to cross over to dedicated woofers at 400-600 hz, depending on how low the planar will go. I think I am leaning towards 500-600 hz since the distortion increases by a lot below 500 hz.
It is not that easy to see in the photo, but there is corrugation support foils both on the top and the bottom, like you have.
Very interesting!
That looks stiffer than my experiment, although I did not test how stiff it is without any aluminum. It might be worth for me to play around with increasing the temperature.
I have measurements from the foils in this post.
First the "default" 12 um kapton + silicone adhesive + 12 um alu:
Next the 10 um mylar + 15 um alu composite I got from @solhaga
Next the control 12 um mylar with lots of gaps:
Lastly the baked 12 um mylar, baked @ 140 degrees C for 20 minutes:
In conclusion, the mylar-alu composite is a lot more efficient than the kapton silicone adhesive variant. Also, while I did not feel much of a difference between the baked and control membrane, the baked one does seem to be stiffer since the dip is higher up in frequency. I still believe that a 23 um mylar variant would probably perform better, but this at least shows that baking does work. And maybe it will work even better if I dial in the temperature.
I also did distortion tests between the kapton-adhesive-alu and the mylar-alu-composite. Both of those is with the same signal and with a 370 hz LR4 roll off which is lower than I will probably go for in the end. But the interesting part here is that the mylar-alu-composite is not only more efficent, it also has lower distortion.
In fact the biggest weakness of the mylar-alu composite is that it is harder to source the foil 😆
If the gods of shipping smile upon me, then I will have another mylar-alu composite to test from @WrineX by next weekend where I hope to test it on saturday. It will be interesting to see how it compares. In that box I also expect some 23 um mylar that I plan to use and bake and see how it performs also.
23 um mylar should still be lighter than the composites, and might be stiff enough after baking. But if the 23 um mylar is strong enough then I will probably also buy a roll of 25 um kapton without any adhesive and see if it is also stiff enough. A problem with kapton though is that it that it has a much higher glass transition temperature so I don't think it is feasible to bake it, but it might be stiff enough without baking. So in the end, mylar might be a better option in this case.
Last edited:
I also contacted UKI because i was not sure if they still sold it, and they asked if i needed another roll of 50 meter of that 9/12 so if it works it is still possible to order it.
nope, i looked up what i paid previously but allot has changed. i know they raised there mylar prices a few times, and i think they where part of the EU back then. so it might be twice as expensive compared to before so maybe around 100-140 euro ex shipping and or VAT would be my guess
Last edited:
Ah, that wouldn't be too bad.
I was afraid it would be many times more, if you are correct then it is still way cheaper than kapton 🙂. And 50 m would be enough for many many membranes.
Hmm since we talk about the use of the 9-12 foil as carrier. i though i revisit a curved panel i made once. problem was/is magnets do not bend 🙂 and the NDFEB rubber magnet i used is still 1/3 the strength of a N35, but rotating the coil it means the N35 or N52 for that mather magnets are only 5mm wide and maybe 3 cm long.. you can make a rather steep curve. biggest problem is getting the shrinkage correct after corrugation of the foil. Well might be complete failure but ill try anyway. non related to you project but because you came up with the idea of using it as carrier foil 🙂
By the way i hope the package come in soon , i send out 2 one arrived in Germany i think by now, so yours would be coming soon as well
By the way i hope the package come in soon , i send out 2 one arrived in Germany i think by now, so yours would be coming soon as well
I take it you want to have a small radius? In my case I have a large radius, 150 cm so I could easily get away with longer magnets like 3 cm, the current 1.2 cm are complete overkill. They were pretty cheap for N52 magnets though 🙂
If I am very lucky it might arrive by tomorrow, but since it hasn't arrived yet it will probably arrive during next week.
If it doesn't arrive then I can still play around with baking temperatures of the mylar. Would also be interesting to make a non baked clone of the mylar membrane but with kapton instead. It will almost definently be awful, but it would be interesting to see if the dip is at the same frequency as the mylar or higher.
yeah true. i had a non shaded heavilly curved one in mind. not even for myself.. since i am not to sure if i like it. but a friend does. but that one has allot of efficiency problems between speakers (foam problems) and overal 🙂 so this might work for that. im sure it will be more efficient with a shitload off 50x 5x3, instead of rubber ndfeb 4x3mm, ndfeb rubber does follow the curve perfectly ofcourse (but it is insane expensive 🙂)
Last edited:
i made another version of the panel i used in my latest tests, but used the Laminate as carrier. but i made it 2 magnets wider to try to lower resonance down to the other panel... it overshoots.... so laminated has res of 280 or something and the regular tweeter variant 450-480hz. the laminated looks much better in the lower end. (even does look better in the top end but i think its due to its width still allows top end to add up. but dispersion sucks)
FR
Brown mylar and alu coil
Green laminated mylar with 2 more magnets 4 ohm compared to 2.8 of the tweeter.
Distortion regular mylar with coil
Distortion laminate version
of course the tweeter was never meant to be crossed this low 🙁. but its nice to see the extra stiffness does work pretty wel for a wider panel. i dont think it works nice for tweeters.... or line source since the mid low will already be much louder (unbless you remove that to just get lower distortion and you are ok with the max spl the 15Khz can put out.) then the top end. but in your case you have allot of unsported mylar. that might work wonders !! i hope it does !
here a video :
Cool!
A nice benefit is that the frequency response is also smoother, less dips and peaks.
And you could possibly do a tweeter that wide, if split the inner and outer coils such that you can drive them coaxially 🙂.
well it did. not sure if it has to do with the foil or the method i put it in the frame... then again the lower end looks better.
but i am sure it does lost top end, but because its wider it still managed to reach 20khz, making it less wide will ramp up resonance and make the top end fall off earlier. thats why in a longer / less wide version it might look different in the top end
well yeah to drive them coaxial they need to be even wider else my tweeters would be only 1-1.5 ohm or something 🙂
Last edited:
I just got the package with the 12+9 mylar-alu composite & the 23 um mylar samples from @WrineX.
The thicker mylar is a lot stiffer than the 12 um variant. It will be interesting to see how it and the composite perform, I have a gut feeling that the 23 um mylar might be the best of them all.
I will try them all out and publish the result next weekend + monday.
The thicker mylar is a lot stiffer than the 12 um variant. It will be interesting to see how it and the composite perform, I have a gut feeling that the 23 um mylar might be the best of them all.
I will try them all out and publish the result next weekend + monday.
Maybe not by itself no. But what if it is baked /annealed, maybe then it will? And who know, maybe it will also be stiffer or comparably stiff to the 9/12 composite at slightly less weight?
Maybe not by itself no. But what if it is baked /annealed, maybe then it will? And who know, maybe it will also be stiffer or comparably stiff to the 9/12 composite at slightly less weight?
No measurements, but some progress on the membranes with the foils from @WrineX
Here is the 12 um mylar + 9 um alu. Or at least that looks to be what we think it is. It is very heavy though. Either the mylar or alu is thicker or there is some adhesive or extra layer that adds a lot of weight.
Compare this to the supposed 10 um mylar + 15 um alu from @solhaga
Either the 9/12 is a lot heavier / thicker than the spec, or the 10/15 is lighter / thinner than we think it is.
I also played around some with the 23 um thick mylar:
Here is with and without the border. The border adds a lot of weight so if I go with the plain mylar, it might be worth playing around with using a thinner border to save weight.
I also tried to corrugate a 23 um mylar + 30 um aluminum without baking / annealing it, and it definently does not work! The mylar is too stiff so it does not hold the corrugations:
I might try with some thicker aluminum traces also, to use 50 um alu instead of 30 as used here but in the end I think baking might be a neccesary step when using this thick mylar. This probably also means that using plain 25 um kapton, which has too high glass transition temperature to be able to be baked / annealed without burning the acrylic adhesive to the alu traces, can be discarded as an idea.
Tomorrow I plan to make more membranes and to play around with baking the 23 um mylar.
Here is the 12 um mylar + 9 um alu. Or at least that looks to be what we think it is. It is very heavy though. Either the mylar or alu is thicker or there is some adhesive or extra layer that adds a lot of weight.
Compare this to the supposed 10 um mylar + 15 um alu from @solhaga
Either the 9/12 is a lot heavier / thicker than the spec, or the 10/15 is lighter / thinner than we think it is.
I also played around some with the 23 um thick mylar:
Here is with and without the border. The border adds a lot of weight so if I go with the plain mylar, it might be worth playing around with using a thinner border to save weight.
I also tried to corrugate a 23 um mylar + 30 um aluminum without baking / annealing it, and it definently does not work! The mylar is too stiff so it does not hold the corrugations:
I might try with some thicker aluminum traces also, to use 50 um alu instead of 30 as used here but in the end I think baking might be a neccesary step when using this thick mylar. This probably also means that using plain 25 um kapton, which has too high glass transition temperature to be able to be baked / annealed without burning the acrylic adhesive to the alu traces, can be discarded as an idea.
Tomorrow I plan to make more membranes and to play around with baking the 23 um mylar.
Oh that is indeed heavy!!! yeah so the datasheet might not be correct 🙁 that explains the loss in top end in my tests.
and baking 12 mic mylar ? i tried it yesterday. but because i do not have as much open area as yours. its hard to say if something changed. i mean corrugation is held already by the alu.
and baking 12 mic mylar ? i tried it yesterday. but because i do not have as much open area as yours. its hard to say if something changed. i mean corrugation is held already by the alu.