Doesn't even look like any glue is used, just the mylar held in place by the sandwiched precision spacers.  I probably could have saved money on glue.  Still waiting for mylar and static spray. It appears everything sits passively under pressure in the chamber from the rubber washer on the top piece.

I do have another question for 

chinsettawong.

Do you apply the antistatic solution to one side of the mylar, or both sides?  I would think one side would be OK since the stuff probably sinks in.

Thank you, Wachara

I can actually practice a bit with saran wrap.  The way things are layered, I can fail a few times before getting it right and not be any worse off.

Well, I did a trial run with saran wrap.  Got the 1.4 micron mylar today, no antistatic yet, that will come next week, but experimented for now just with saran wrap.

I stretched the saran with tape, put one half of the mylar spacer under, pressed it with my finger and no glue.  I then placed the other half of the spacer over the bottom one, pressed both spacers together with saran in between, then cut around it with a sharp scalpel.  

Surprisingly, this produced a tense, stable configuration without glue.  I pinged it with my finger, and it produced a note nearly identical to the Stax mylar membrane.  I placed the mylar back into the headphone shell, it all fit perfectly that way.

My thought is to tension the 1.4 mylar, put the antistatic on it, let that act as glue, let it dry with pressure overnight with both spacers together, cut with scalpel, reassemble the headseat, and try it out and see how it does.

If that doesn't work, I will strip out the mylar and try it with glue.  I think this system is intended to work without glue, I couldn't find any trace at all of glue, but maybe they use some kind of ultra thin adhesive that doesn't show very well.

I got a pair of working diaphragms into the headseats.  I have to admit, I like Stax better when they are "magic boxes" rather than fixer uppers.  I am learning more about stators than I ever wanted to know.

I also identified the original source of my problem at last.  Although the holes punched in the mylar by some voltage aberration did not do any good I am sure, there is also a point of vulnerability in the stators where a crimp exists at the stator attachment.  I guess the insulating glue detached, the juncture flexed, and the metal fatigued and broke, giving an intermittent connection in the right channel. It never had anything to do with bias.

I rigged the headsets up to allow the broken stator a kind of temporary passive contact with a scraped spot with the new diaphragm in place.  

I got them playing with normal bias quite nicely, balanced, and without break up. They sound nearly as good on normal bias as my normal bias SR 3n/5n.

However, I think I will re-do the diaphragms with very light glue as suggested by chinsettawong to get the most stable tensioning, then I will have to devise a way of reconnecting the fractured stator.

They started out very inefficient, but eventually became louder with use.  I have only used the normal bias SRD 7SB so far, which probably has about 200v bias.  I won't have my pro socket transformer back until next week, and need several hours for the project.

Hi, Wachara,

Thanks.

A couple of years ago I was trying different thicknesses of metal and aluminum foils for ribbons in super tweeters, I noticed advantages to both thicker and thinner.  Thinner=faster but maybe not so natural.  Thicker=slower, but fuller and in some ways more pleasing.

I suppose the thickness would be to taste per individual.  After i get through this project with a working set of headphones again, experimentation is a possibility, especially since  I have the  stuff to do it with, now. It looks like standard mylar saran wrap must be something in the 5 micron vicinity.

Carl

PS it is hard to take a picture of the "crimp" because my camera does not focus that close.  The working ones are still glued and imbedded, it is basically a  partial triangular bend close to the stator body from the soldered lead. It looks like one side of the lead has rubber cement, which insulates it from the diaphragm/bias ring. It just broke off at the crimp point when the rubber cement  allowed some motion in the stator lead.

Well, I have had a chance to try my 507's with new mylar (first try) and kind of jimmy rigged the stator on the R. I finally listened using the pro socket 7 transformer I brought back from Santa Cruz.

Wow!  The 507 with these ad hoc membranes and high bias are amongst the most revealing headphones I have ever heard.  They make my previous speed champs, the normal bias 3n/5n sound slow by comparison.  And they sound good, not tizzy or harsh, very smooth as well.  The re-mylared 507s bring out high frequency detail and modulations I can't hear on standard bias Lambdas or the 3n/5n.  Soundstage is huge. Bass is quite decent, but not as deep and hard as the standard 507 when they were working, but not that far off, and they seem to be getting better and a bit more efficient/louder as they get some use.

It is too soon to say that the new mylar improves them over the previously working stock 507, but it seems quite promising, the detail retrieval is greater than I have heard from headphones I have had reasonable experience with.  I may just get the stator attached better and listen to them like this for a while because I am almost afraid changing out the mylar again might make them worse.

When the 507s in standard issue were working, they had more detail than normal bias lambdas, but maybe a shade less detail than the 3n/5n that I have.  They had larger soundstage, but fuller bass and colorful tone quality.

All I can say, soundoholic, is that your work leaves me in complete awe.  Those appear to be gorgeous, world class units compared to my modest breadboard effort.

It is nice to hear from somebody who enjoys the benefits of transformer coupled stax with top amplifiers.

Hi, Wachara,

Since the last report without glue, which produced a usable result with 1.4 micron mylar, I have re done them twice with glue to hone the method.  I used the 1.4 micron from outdoor specialties that you recommended. Redoing them twice, I am becoming a bit more handy with them.  I glued a first trial set, but did not like their tension, so re-did them once again.  

There are a lot of subtle things to do to make sure the installation goes OK, and several pitfalls that I learned as I went along, but I think I now have a nice working pair of membranes.  The 507s sound so far at least as good as when they were presumably working OK, but I think that they have more detail, they sound quite good for DIY. Faster sounding than normal bias 3n/5n and more dynamic than standard bias Lambda, so far so good.

Also, they sound great as standard bias phones, but don't get that last push in detail and dynamics unless the pro bias is used.

Thank you for your help, I will keep posting as I try them out over time.