Relation between strecher and heatshrink in ESL building

Hi,

in short:

.... It depends

jauu
Calvin

You probably want the diaphragm as tight as possible to allow for the highest possible bias voltage. Stretching alone is best but it is very hard to do, you will need a big apparatus to do it right.

I would stretch the mylar as best as possible by hand with masking tape. Then shrink it.

Hi,

it depends simply on what You intend to do!
If You use the usual film that is sold by several sources and lets say You want to build a curved panel. Than thermal shrinking is no option (I don´t trust the method in R.Sanders Cookbook).

If You have to reach highest possible tension values (to allow for highest polarisation voltages for e.g) thermal shrinking is no option.

If You can live with a reduced value of tension (because You designed Your panel with that in mind) and if the type of panel allows for tempering, it is a neat technique indeed that reduces the break-in time and that allows for low tolerances.
Than mechanical stretching followed by tempering is enough.

jauu
Calvin

If the diaphragm is placed under very high tension (perhaps using a stretching frame like this), glued to the stator assembly, and then heated with a heat gun, does the applied heat increase or decrease the final tension of the diaphragm?

I know from experience that when the diaphragm is only loosely tensioned and then heat is applied, the tension increases. However, if the diaphragm is already strained (deformed because of the applied stress) when the heat is applied, I could imagine it simply stretching and losing tension. Can someone with hands-on experience report what actually happens? Does the resonance frequency of a highly tensioned diaphragm increase or decrease when heated?

Few

Hi,

tempering only tightens very loosely stretched films. Tight mechanically stretched films relax under heat and the Fs sinks.

jauu
Calvin

Thanks. That's what I expected, but I didn't want to assert it as fact without direct experience.
Few

polyester films...

are for the most part biaxially oriented. That means that they will shrink more or less equally in both length and width. How much they will shrink depends upon the actual film type as they are all different. The reason why shrinking diaphragms on curved panels is a no go is because with a curved stator you only want tension in the vertical plane (asume a horizontal convex curve) so that rules out heat tensioning because it will tension the film in length and width equally. That will cause the diaphragm to colapse onto the rear stator. The best way to imagine a curved panel diaphragm is the fishing line concept. Imagine the diaphragm as a constant arc made up from single thin fishing lines. You can tension each line all you want in the vertical plane and so the diaphragm will follow the curve of the stator. As soon as you place the diaphragm under horizontal tension you will pull the diaphragm flat.

In practice it's hard to stretch a panel to even tension everywhere. I use ER audio's method which is about the best out there (IMHO). With heat shrinking you may be able to do better. However, both take practice and neither is a clear winner IMHO.

One big advantage of the stretching technique is that if you use a stretcher like shown above, or ER audio's technique, it's very repeatable. With practice, so is heat shrinking, but it does take more practice.

The biggest advantage stretching has is that you won't melt a hole in the diaphragm after it's glued up. When heat shrinking, depending on our heat gun, there's a fine line between shrink, and melt. If something goes wrong with the stretcher, you are only out a piece of mylar. If you melt a hole in a diaphragm, you've got to clean up the glued stator.


Sheldon

My experience has been that if you want an ESL to be efficient which translates to playing loudly, you need to use high bias voltage. That means you need a LOT of tension on the diaphragm. That means you can't use heat shrinking.

I_F

High efficiency is good, but so are low resonant frequencies. You can play games with spacer and support geometry to get both low resonant frequency and good support for stability. But for a given geometry cranking up the tension can lead to other problems. If you are buildig a narrowish ESL and crank up the tension, you may have to deal with a resonant peak above where your woofer cuts over. You can use that to your advantage (as some older ML ELS's did I think) by using the resonant peak to counteract the frequency rolloff due to pressure feedback, that it's not a great solution.

But in general I_F is right, tension is good. In my experience either technique will lead to sufficient tension to be a satisfactory ESL. The technique that is best for you is the one that you are most consistent with.


Sheldon