[chinsettawong]

Hi OndesX,
 
Yes, the Mylar can be a headache to work with.  The thinner you go, the easier it'll break.
 
The correct tension value, in your case, will be the amount of weight you put on.  Isn't it?
 
Before you can get the right value, you'll just have to do a lot of experiments.  Too much tension, there will be less bass.  Too little, the diaphragm will not be stable.  

 
Wachara C.

 

[ondesx]

Hi Wachara,
 
Yes, but I still have to define a method to determine beforehand the needed tension (i.e. the mass) for a given plastic film...
 
I suppose that this will be determined through several trial and error, since I haven't any idea about how to build a computer simulation or an analytical study of such a problem...
 
Any suggestion is welcome, but I suppose you already have thought about this question.
 
OndesX

 

[chinsettawong]

Trial and error will be your very good friend.

 

[davidsh]

I read on a DIY esl page that as a guideline the Mylar should be stretched 1.5%.
Just passing it on..

 

[chinsettawong]

Stretching the diaphragm that much is way too much, unfortunately.

 

[n3rdling]

That's about the most you can stretch it before it tears.

 

[ondesx]

Hi all,
 
Here is some maths about the estimation of the tension for the diaphragm.
 
The resonance frequency of a stretched membrane is a function of the tension and the form of the diaphragm. For a circular membrane, it's known that the tension and the mass have the following relationship :
 
T=M(f x R^2/0.146)
 
where :
T is the tension in dynes/cm
M is the mass of the membrane in g
f the resonant frequency
r the radius of the membrane in cm
 
It is assumed that a 500 cm^2 - 12µ Mylar membrane has a mass about or less than 1 g and for a 4 µ thin film the value should be 3 times lower i. e. about 0.3 g. A 10 cm stator have a 5 cm radius, thus a surface of 80 cm^2. Then the mass will be about 0.05g. Thus, knowing the diaphragm radius and for an estimated 100 Hz resonant frequency, the tension might be calculated now...
 
Any comment is welcome...
 
OndesX

 

[ondesx]

Another formula is given here :
 
http://hyperphysics.phy-astr.gsu.edu/hbase/music/cirmem.html
 
For the Mylar the density is 1.3 g/cm^3 and 5.2 g/m^2 if one uses a 4µ Mylar film. For a 10.5 cm of diameter diaphragm, the calculated Tension on this page is about 90 g/m...
 
Something seems to be wrong.
 
OndesX

 

[chinsettawong]

I think it's rather difficult to calculate the right tension. Even if you can, how can you make sure that you can get the exact right tension?

I always stretch the diaphragm so that it's free of wrinkles or almost free of. The elongation is probably only around 0.5%. However, since the tension isn't so much. It's rather difficult to get a consistent result. I eventually have to do it a few times before I'm happy with the result.

I suggest that you just try it and you'll see what I mean.

 

[ondesx]

If we must stretch the diaphragm so that it's almost free of wrinkles, the rings alone will do the job in my case. I mean additional weight will be useless.
 
But, if I correctly remember one picture you posted on this thread (March 2, 2013), you have shown several bottles of water pulling the film... 
 
http://www.head-fi.org/t/498292/my-diy-electrostatic-headphones/720
 
You suggested at that time that for a 3µ film, you could use at least 800 to 1000 g for each weight. On the picture, each bottle weighs 650 g and you used a total of 8 bottles, i. e. more than 5 kg  !... I suppose that you might remove the ripples for much less weight isn't ?
 
Please, can you comment this point ?
 
OndesX

 

[arnaud]

Assuming you want to match a pair of assembled transducers, how about this:
- Download a free / cheap SPL analyzer app on your smartphone.
- Run some white noise through assembled transducer and check the open air resonance frequency using the SPL analyser (1/12 Octave band or some similar setting) by placing the phone nearby either sides of the transducer
- As expected,  tensioning will directly relate to this frequency and it's fairly easy to visualise as a free transducer has very little self damping so you'll get a nice bump (anywhere between 80Hz and 150Hz I guess).
- I guess the next job is to build multiple samples and pair them. 
- Now, unfortunately there is some complication as the free air resonance frequency is quite different from the coupled resonance frequency once the transducer is "acoustically" loaded by the earpad cavity system
- If it's a non leaky pad, you can expect a severe change in the first coupled resonance, like dropping from 120Hz-150Hz to 40-50Hz in case of Omega / SR009 type frames (see measurement link in my rig, I think I compare free air vs. loaded response of SR009 and SR007 phones)
 
Just stating the obvious, I probably missed an important point why you are not already doing it this way...
 
arnaud

 

[chinsettawong]

Quote:

  If we must stretch the diaphragm so that it's almost free of wrinkles, the rings alone will do the job in my case. I mean additional weight will be useless.
 
But, if I correctly remember one picture you posted on this thread (March 2, 2013), you have shown several bottles of water pulling the film... 
 
http://www.head-fi.org/t/498292/my-diy-electrostatic-headphones/720
 
You suggested at that time that for a 3µ film, you could use at least 800 to 1000 g for each weight. On the picture, each bottle weighs 650 g and you used a total of 8 bottles, i. e. more than 5 kg  !... I suppose that you might remove the ripples for much less weight isn't ?
 
Please, can you comment this point ?
 
OndesX

 
Using the method I posted certainly works, but I haven't spent enough time doing trials and errors to get the right weights. 
 
Whatever method you choose to use, you'll just need to try, try, and try until you get it just right.  It's not so easy.
 
Wachara C.

 

[chinsettawong]

Hi Arnaud,
 
To make a match pair, I just stretch the Mylar that is just big enough for one pair.  That saves a lot of time and headache.
 
If I need to find whether they're really match, I simply knock them against the edge of a table and compare the sounds.  The match pair should sound very similar.
 
Wachara C.

 

[arnaud]

That's right, I always forget about all the tricks you apply Wachara. Somehow they're always so much more reasonable ideas than what an engineer can come up with  

 

[ondesx]

  Quote:
 
Using the method I posted certainly works, but I haven't spent enough time doing trials and errors to get the right weights. 
 
Whatever method you choose to use, you'll just need to try, try, and try until you get it just right.  It's not so easy.
 
Wachara C.

 
Yes I understand that I have to try and fail several times... But, my question was about the intensity : removing simply the ripples doesn't need 5 kg around the diaphragm... 
 
Thus, with your more recent stretched membranes in which range are you ? around one kg or more than 5 kg for à 3 µ Mylar ?
 
OndesX