[chinsettawong]

I don't use weights to stretch Mylar anymore. I use my inner tire stretcher instead.

I stretch it so that the wrinkles are almost all gone. The elongation is just around 0.5% or less.

It's rather difficult to describe. I encourage you to really try it and you'll understand.

 

[ondesx]

I don't use weights to stretch Mylar anymore. I use my inner tire stretcher instead.

I stretch it so that the wrinkles are almost all gone. The elongation is just around 0.5% or less.

It's rather difficult to describe. I encourage you to really try it and you'll understand.

 
OK but do you remember your advice to somebody in this thread :
 
"If you can find a helping hand, I think the stretcher using clips and weights is better.  It's easier to repeat the process too."
 
Thus, I'll give a try to a 3µ Mylar with my jig since it's possible, this way, to "quantify" the weight used for stretching it.

 

[chinsettawong]

I agree that using weights, once you get it right, you can have good repetitions.

Your jig looks good. Go for it!

 

[BillG]

I don't use weights to stretch Mylar anymore. I use my inner tire stretcher instead.

I stretch it so that the wrinkles are almost all gone. The elongation is just around 0.5% or less.

It's rather difficult to describe. I encourage you to really try it and you'll understand.

Hi Wachara et al,
I'm following this discussion with great interest. Some day soon I'll be ready to do some stretching too.  I raise some issues and ideas which apply to the topic and hope they help in getting the diaphragms we need.
 
Once the film is stretched to a desired tension, adhesive is applied to the spacer ring and the ring is placed on the tensioned mylar. We hope that the adhesion is great enough so that tension is maintained after the adhesive sets and the covered spacer is removed from the larger mylar sheet.    In making a matching pair of diaphragms for a pair of phones, some adjustment with the hot air gun on the diaphragm with the lower vibrating pitch is possible.  So far so good.  My concern is that the mylar is stretched, on whatever jig is being used, to some arbitrary tightness and simply held there for further adjustment followed by  the glueing. No heat shrinking has taken place yet.  Does that leave the diaphragm in an unstable state? Perhaps pre-shrinking of the mylar should be done to stabilize it.  Preshrinking may not be needed, I don't know yet.
 
The next parameter to consider is how tight to make it initially.  I'll explore that in the next posting. Gottago now, back later.
 
BillG

 

[BillG]

Hi All,
This reply/post is really a continuance of my #1789.  Something I was just getting to was whether or not the mylar film shrinkage or even "coeficient of stretch" is the same for direction of applied tensioning force.  I think I saw something indicating it may stretch more in one direction than the other.  If this is true, what is the result?  Does it even matter?  How to compensate for this is a question.  Maybe prestretching could  be helpful?  
 
In selecting the desired tightness of the finished diaphragm it seems to be the general opinion that resonant frequency of the diaphragm in free air is the parameter easiest to observe and control.  To those of you all that have built ESPhones what RFs have you found to make the best sounding phone(s)?   Has there been any attempt to find how much changing this RF one way or the other, say up or down, by 10 or 20 Hz makes any real difference in the overall sound quality? 
 
I suspect that  parameters  other than diaphragm tension  may make as much or more difference in the final sound output.  Tension seems to be the one which is least able to be defined or controlled. This is why I've chosen to begin my  thinking here.
Regards to all,
BillG

 

[chinsettawong]

There are really two ways to tension the Mylar.  You either stretch it or heat shrink it.  However, mechanical stretching gives far higher tension than heat shrinking.  I've tried many, many times to use just a hot air gun to heat shrink the diaphragm, but it just doesn't yeild enough tension.
 
Once the Mylar is mechanically stretched and glued to the spacer ring, if the glue is good, the tension stays stable for a long, long time.  So, please don't worry too much.
 
As for me, I make the headphones just for myself.  I don't worry too much about the theory, and I just do it.  The test results of my headphones aren't that bad too:   http://www.innerfidelity.com/content/extraordinary-diy-electrostatic-headphones-chinsettawong
 
Wachara C.

 

[ondesx]

Something I was just getting to was whether or not the mylar film shrinkage or even "coeficient of stretch" is the same for direction of applied tensioning force.  I think I saw something indicating it may stretch more in one direction than the other.  If this is true, what is the result?  Does it even matter?  How to compensate for this is a question.  Maybe prestretching could  be helpful?    
In selecting the desired tightness of the finished diaphragm it seems to be the general opinion that resonant frequency of the diaphragm in free air is the parameter easiest to observe and control.  To those of you all that have built ESPhones what RFs have you found to make the best sounding phone(s)?   Has there been any attempt to find how much changing this RF one way or the other, say up or down, by 10 or 20 Hz makes any real difference in the overall sound quality? 
 
I suspect that  parameters  other than diaphragm tension  may make as much or more difference in the final sound output.  Tension seems to be the one which is least able to be defined or controlled. This is why I've chosen to begin my  thinking here.
Regards to all,
BillG

Hi BillG,
 
It's true that there are 2 perpendicular stretching directions for PET films, one direction is more difficult to stretch than the other : you might try to stretch manually a small piece of PET to convince yourself.
But, this isn't a paramount point, as far as you make an homogenous tension around all the film.
 
The resonance frequency isn't so important, but it must be maintained around 100 Hz more or less. Ten or 20 Hz of variation in this RF isn't an issue.
 
The physical problem behind the diaphragm tension is that the electrostatic force must overcome this tension on one hand, because if the force can't overcome this tension, no sound will happen ! But, on the other hand, if one uses a too low tension, it results an instability of the diaphragm which collapses on the stators on one side after the other. This is why the value of this tension is so delicate to define : we have to optimize these two contradictory constraints !...
 
Just to comment a few on the major factors needed to obtain, at least, an usable pair of headphones : I think that the quality of the gluing for the spacers and the diaphragm is paramount, not too thick and as even as possible. The thickness of the spacers and the stators, the insulation of the stators, the efficiency and the lasting of the diaphragm's coating and the rigidity of the whole system are, probably, much more important than the RF.
 
This is my personal vision of this project, of course.
 
OndesX

 

[davidsh]

Just want to give an update. I actually finished the esl in my physics project. Very very crude and simple construction. The stretching of the food film was done by hand by twisting metal rods around the 4 sides of the film then stretching, one person per side.

 

[BillG]

I agree that using weights, once you get it right, you can have good repetitions.

Your jig looks good. Go for it!

I'm sorry to continue poking at this but it appears that we are not responding to the question that is being asked here.    I too agree that one should be able to get good repetitions using weights.  The problem is that no one seems to be able to recognize what "right" is once you get it.  IMHO those of us who are pushing on theoretical points and definitions etc, are just trying to recognize "right" once we get it.  If we can do that it should be possible to reduce the number of random iterations and zero in on "right" in a reliable manner.
BillG

 

[davidsh]

But what is right, though? We could start there. What do you think?

 

[BillG]

But what is right, though? We could start there. What do you think?

Hi David,
You read my post correctly and translate it to a small number of words.  In coming up with more knowledge of "right" we could start by establishing a tension that just keeps the diaphragm from touching the stator. This would be a best case for safety.  But would it be the best case for sound?   What other controls could be applied with reasonable effort which would be better than just not touching the stator?  By the way, I'm trying to simplify the process.
Regards,
BillG

 

[davidsh]

Well, for the diaphragm not to touch the stator you'd want the restoring force to be great than the electrostatic force at all times. The restoring force depends on the elasticity which depends on the material used and the tension. I think the esl cookbook has something to say about it, though there are no equations as such. I'm not really sure there are any means of calculating this in the first place. Whether the diaphragm will collapse depends on a lot of factors I think. Besides, there are many other types of instability such as crackling or hissing sounds, imbalance and stuff like that. In principle any diaphragm can be made stable if the spacers are big enough but I guess we are assuming that one uses pro bias voltage here.
I am still to figure out how to calculate the forces on the diaphragm, though I'm getting there.

 

[ondesx]

...read my post correctly and translate it to a small number of words.  In coming up with more knowledge of "right" we could start by establishing a tension that just keeps the diaphragm from touching the stator. This would be a best case for safety.  But would it be the best case for sound?   What other controls could be applied with reasonable effort which would be better than just not touching the stator?  By the way, I'm trying to simplify the process.

 
Unfortunately, there is no easy way to determine, from a theoretical point of view, the "best" tension of the diaphragm for a given polarization voltage.

First of all, we have to define what we mean by "best". I think that we're looking for the tension of a given film, that allows the loudest audio signal with as much bass as possible... Since the constraints are opposite, none recipe can define a priori this "best" tension. The optimization of this value requires an iterative process between a given tension and several polarization values. Progressively, we might derive this value for a set of parameters (the thickness and the composition of the film, polarization tension and physical data like DS-distance and geometry of the stators).
 
Anyway, it must be reminded, as mentioned in a previous post, that the quality and durability of the coating of this film, the thickness of the adhesive, the regularity of its deposit, the parallelism and stiffness of all the components are of paramount importance to derive an usable cell !... 
 
Of course, we state from here that we have already defined the "good stators and spacers", i. e. with the correct number, distribution and alignment of holes, a good thickness of all components without Helmholtz's resonators, and obviously with an homogenous electric field for the whole surface of the film... 
 
Well, as usual, the more we penetrate in a given technical domain, the more we discover new issues ! 

 
OndesX

 

[wakibaki]

Hi guys.
 
Sorry I haven't made a contribution for a while. Getting stators and spacers manufactured in bulk has just proved beyond my energy.
 
Regarding the stretching problem, there are a couple of things that might be done.
 
A stretching jig might be devised with some kind of repeatability and mechanical graduation.
 
A way of estimating the tension might be devised.
 
In reverse order, you could apply a load to the diaphragm, and observe the displacement in the bowing of a reflection of a straight edge or slot source (lamp). You could apply the loading electrically or maybe just with a weight.
 
As far as a stretching jig is concerned, I was thinking of an arrangement like an embroidery ring to hold the membrane, hand tensioned to remove slack. This would then be inverted (membrane on the bottom) and forced over a considerably smaller inner ring to a graduated depth. I've actually tried this with embroidery rings, but the grip on the membrane was inadequate. PTFE (Teflon) might be good for the surface that bears on the sheet and stretches it. Graphite might be a benign lubricant.

 

[chinsettawong]

  I'm sorry to continue poking at this but it appears that we are not responding to the question that is being asked here.    I too agree that one should be able to get good repetitions using weights.  The problem is that no one seems to be able to recognize what "right" is once you get it.  IMHO those of us who are pushing on theoretical points and definitions etc, are just trying to recognize "right" once we get it.  If we can do that it should be possible to reduce the number of random iterations and zero in on "right" in a reliable manner.
BillG

 
Hi BillG,
 
As for me, I trust my ears more than anything.  By "just right", I mean it sounds good to my ears.  I really don't know how to quantify the feeling.
 
Anyway, you can look at the FR of the 2 pairs of headphones that I sent to Tyll for measurement.  If that helps, that's what I meant by just "right".
 
Wachara C.