[arnaud]

Wow, I just happened to stumble upon this thread today and must say I am humbled by your efforts Wachara! I am going to ask a moderator to remove my contributor title and give it to you instead!!
 
As an acoustic engineer who spends his days simulating the vibro-acoustics of all kinds of equipment, I find fascinating the trial and errors that you've been through to come up with a good sound headphone.
 
Among the many things you've tried, I was wondering if you had some insight about the following? (I can make guesses based on my experience, but its virtually inexistent in the case of headphones so your input is worth so much more...
 
> Fim thickness: you've tried quite a few thicknesses and did not hear much differences at first but then found that 3 micron was better then 1 micron as far as bass rendering. What I can't get my head around is how you ensure that you have a reasonable cut-on frequency for the diaphragm (by that I mean the first free-air resonance of the diaphragm, which is typically in the 50-75Hz for Stax headphones I believe). Did you find out you had to increase the tensioning with thicker diaphragms?
 
> Earpad cavity: you've mentioned that from experience, increased distance from the diaphragm to the ear turns into a bassier tonal balance. Since the earcup cavity is most always seal type (leather or pleather pad), indeed the height of the pads governs the stiffness of this "air spring" and likely affects the low frequency driver response. But you don't seem to play with that variable as much as actual driver design? This is something I may be able to look into with simulation.
 
> Enclosure model: similar to the earcup cavity, the driver frame seems to act as an acoustic baffle to prevent acoustic cancellation of the front and back waves. Have you done experiments with different frame dimensions? This I could also potential look into with simulation.
 
> Stator design: per Stax, and intuitively, one has to pay attention to the stator design (it has to be open enough but on the other hand maintaining structural integrity is key to limit distortion). You seem to be using always the same material (1mm PCB?) but I wonder if such relatively thin part doesn't vibrate within the audio range. In which case, part of the benefits with limiting the perforated region is to maintain overall rigidity. This is something I could easily simulate if you give me stator dimensions (including perforations) and material.
 
> Dual diaphragm: we have been discussed on an other site about dual diaphragms and I was surprised to see you've already tried it! I am keen also on simulating this with an idea I had in mind (two concentric diaphragms between the same stators but with a rim spacer separating the two diaphragms and usage of a thinner diaphragm for the center "tweeter" part). I was wondering if you had any comment on this (like it makes no sense or you won't be able to assemble it...)
 
> Perforation size: it seems like that rule of thumb is to use a perforation diameter to match the stator/stator gap, any idea where that comes from? From a physics point of view, I imagine the perforated stators help to damp the diaphragm through viscous losses (although they're typically negligible for channels smaller than 1mm). In practice, using larger perforations brings nasty colorations in the midrange as mentioned by Birgir?
 
You may not feel like answering all these questions, I will totally understand . But if you have interest in what simulation may (or may not lol) bring to the table to help with improving the design, I'd be happy to contribute. Ideally, measurements by someone like Purrin would also dramatically help but I am not sure I can entice him into the venture .
 
cheers, arnaud

 

[AmarokCZ]

I also done some experimenting, so here is my point of view:
> Perforation size: I'm using 2mm holes with 0.6mm D-S gap and there is no coloration compared to STAX SR-3NEW (0.5mm holes and 0.3mm D-S gap).
> Film thickness: I tried few films and I found, that bass is better (more controlled) if the force required to move with diaphragm is larger (this applies only to some point). This, of course, negatively affects sensitivity of driver. One of my previous driver had free-air resonant frequency ~300Hz, but I think it's not problem since it's damped by closed cavity between head and driver.
Just for your information the free-air resonant frequency of SR-3NEW (old round earspeakers, 8cm^2 active diaphragm area) is ~160Hz and it affected low frequency response only when there wasn't good seal (see this frequency response graph: red line when resealed, yellow line is what I was listening to before resealing).
> Earpad cavity: I tried 15mm and 40mm thick earpads and with 40mm earpad the sound was little less bassier (!!) with less controlled bass and less bass punch.

 

[chinsettawong]

Hi Arnaud,
 
First of all, let me tell you that I don't really have much knowledge about Physics, Mechanic, or Electronics.  I make these headphones just for fun, and actually just to get me away from thinking about my day work.  I don't have any instruments for making sound measurement, and I don't do any calculations or simulations at all when I'm making a design. So, forgive me if I can't answer some of your questions, since I really don't know how to answer.  

 
> Film thickness: In my first few headphones, I have been using thicker spacers (0.6-1 mm) than what I use now (0.5 mm).  With thicker spacers, using thin film, the stability is good.  But when I reduce spacer thickness down, I need to increase tension to get good diaphragm stability.  With that, I lose bass.  Therefore, I have no way but to use thicker film.  With thicker film you can put more tension on it and still have good bass.  I am tempted to try 6 micron Mylar on my next one.
 
> Earpad Cavity:  Yes, putting the drivers a little bit farther away from ear, in my experience, gives a little better bass.  Thicker earpads are certainly better, IMO.
 
> Enclosure model:  I've never done any experiment like that.  However, I don't like to have anything obstructing the back side of the drivers.  Open back is certainly better than sealed back, IMO.
 
> Stator Design:  Yes, I have been using 1 mm PCB for my stators since the beginning.  I've only found out recently when I have a chance to open up a pair of Stax 007 that Stax also uses 1 mm PCB for its stators. I haven't had any problem with its rigidity.  I've made many different types of driver.  Which one's dimension would you like to have?
 
> Dual Diaphragm:  The last time I tried it, I didn't like it.  The sound that I hear was like echoing.  I think I must have done something wrong, because I tried stacking 2 drivers and listened to them the other day, and I didn't notice that echo sound.  I'll give it a try again in the near future.  I'm thinking of putting a thicker diaphragm in the back, and perhaps make some low pass filters for that diaphragm.  By the way, I don't quite understand the kind of dual diaphragms you try to explain.  The easier way to make a diaphragm sound different frequencies in different areas is to make segmentations on the stators.  You can make two separate circuits on the stators and put a high pass filter on the center circle.  That way you can make only a small center part of the diaphragm sound like a tweeter.  
 
> Perforation size:  I have't heard the rule of thump you mentioned before.  I always use 2 mm holes, and that's because I can find 2 mm drill bits easier.  To me the perforation size doesn't make much of a difference.  I don't think bigger holes puts coloration on the sound.  However, it might affect the efficiency of the drivers a little. 
 
Thanks for your offering to make simulation.  I have no knowledge about that.  But if you feel like doing it, let me know what info you need.  
 
Wachara C.

 

[arnaud]

Wachara and amarokcz, thanks for the replies!

Wachara, you've made several designs, how about your favorite one? All I need to get started is the parts dimensions, dsf files probably being the easiest to deal with. That and the materials used. I thing I want to try is evaluate the radiation from residual stator vibration relative to the diaphragm motion. Ideally, I would like to know the free air resonance frequency rather than tensioning but it seems you're doing everything by feel (and doing a darn good job apparently!).

I may start with the existing 009 model I made and evaluate the dual diaphragms (and/or split stators excitation) before going to your headphone model.

 

[chinsettawong]

Quote:

Wachara and amarokcz, thanks for the replies!
Wachara, you've made several designs, how about your favorite one? All I need to get started is the parts dimensions, dsf files probably being the easiest to deal with. That and the materials used. I thing I want to try is evaluate the radiation from residual stator vibration relative to the diaphragm motion. Ideally, I would like to know the free air resonance frequency rather than tensioning but it seems you're doing everything by feel (and doing a darn good job apparently!).
I may start with the existing 009 model I made and evaluate the dual diaphragms (and/or split stators excitation) before going to your headphone model.

I'll send you the drawings when I get back home tonight.  Please PM me with your email address.
 
Wachara C.

 

[dude_500]

Finished up building some quick headphones to go with the drivers today just so I could get some listening in on them before I leave for the rest of the month. They blow my HD600/beta-22 away. I'm not really good at describing headphone sound qualities, they're just incredible. I need to try to listen to some good Stax sometime just so I have a reference of comparison, since these are the first electrostatic headphones I've heard. I don't think I'll ever listen to magnetic headphones again (unless it's 100* out and my dorm room is on fire and 200 watts is too much to add in, which should describe the next two months quite nicely).
 
I haven't added any moisture shield in on the ear's side of the driver, which I imagine I'll need to do to keep them working in the long run. Do you typically tension that film or just leave it loose? I would think the film vibrating and hitting the stator could be a problem either tensioned or not.
 

 

[chinsettawong]

Hi Dude_500,

Congratulations! With those large drivers and a BH amp, I am very sure that your headphone's sound is already on par, if not better, than those top of the line commercial headphones. You know what? I stop listening to non-electrostatic headphones for a long time already. After this you might want to make yourself a pair of good ESL. And you'll never want to listen to those boxy loudspeakers again. I'm sure.

Yes, you definitely want to make a sweat protection diaphragm. The tension isn't crucial here. I just hard tight it and if it has any wrinkle, I don't really care. You put it on only the side that faces your ear. I don't put it on the back side, because it will trap the air and the bass isn't as good.

Wachara C.

 

[Micaiah]

After years of using Head-fi, this thread made me finally sign up. It's humbling how creative some people are and especially when they are willing to share their hard earned knowledge! Thank you!
 
I'd like to try making a pair of electrostatics, even if just something simple at first. Only problem is what to plug them into without jumping in head first, buying or building an expensive amp. Any recommendations for an amp that wouldn't be too costly? I don't care if it doesn't sound that good as long as it's cheap. Just something to experiment with to get my feet wet. Is it possible to convert a loudspeaker amp or to make a transformer that connects to an amp for not too much money?

 

[chinsettawong]

Hi Micaiah,
 
There are two ways to listen to electrostatic headphones.  The cheaper way is to listen to them thru the connection with step-up transformers and a normal loudspeaker amp .  You might be able to find a cheap used unit of Stax.  Depending on its age and condition, sometimes you can find it for less than US$100.  The other way is to build yourself a simple tube amp. I built 2 tube amps using a simple circuit, and they costed me around US$250 each. Either of these solutions can give you pretty good sound.  But if don't want to spend money and just want to try it out, you can also find some old, trashed toroidal transformers and put some wires in them.  Then they can be used as step-up transformers.  Together with a simple bias circuit, you can listen to the headphones for sure. They won't sound that good, but they can certainly get your feet wet. 

 
Let me know if I can be of further help.
 
Wachara C.

 

[spritzer]

Even power transformers wired backwards will work in a pinch.  The bias supply can be a simple voltage multiplier or just something simple off two back to back transformers. 

 

[chinsettawong]

Quote:

Even power transformers wired backwards will work in a pinch.  The bias supply can be a simple voltage multiplier or just something simple off two back to back transformers. 

Yup!  That will work too.
 
Wachara C.

 

[jgazal]

Quote:

> Dual diaphragm: we have been discussed on an other site about dual diaphragms and I was surprised to see you've already tried it! I am keen also on simulating this with an idea I had in mind (two concentric diaphragms between the same stators but with a rim spacer separating the two diaphragms and usage of a thinner diaphragm for the center "tweeter" part). I was wondering if you had any comment on this (like it makes no sense or you won't be able to assemble it...)
 
cheers, arnaud

 
 

> Dual Diaphragm:  The last time I tried it, I didn't like it.  The sound that I hear was like echoing.  I think I must have done something wrong, because I tried stacking 2 drivers and listened to them the other day, and I didn't notice that echo sound.  I'll give it a try again in the near future.  I'm thinking of putting a thicker diaphragm in the back, and perhaps make some low pass filters for that diaphragm.  By the way, I don't quite understand the kind of dual diaphragms you try to explain.  The easier way to make a diaphragm sound different frequencies in different areas is to make segmentations on the stators.  You can make two separate circuits on the stators and put a high pass filter on the center circle.  That way you can make only a small center part of the diaphragm sound like a tweeter.  
 
Wachara C.

 
I have tried to draw Arnaud idea:
 

 
Arnaud, is the rim spacer draw correct? Which material would you use with such rim spacer? Would the Mylar just be glued to the rim spacer?
 
It is easier to "build" things virtually...
 
Cheers 

 

[arnaud]

Quote:

Arnaud, is the rim spacer draw correct? Which material would you use with such rim spacer? Would the Mylar just be glued to the rim spacer?
 
It is easier to "build" things virtually...
 
Cheers 

 
That's the idea although I thought you could have the spacers (PCB or the like) mounted on both stators (the stators would have an outer and inner ring with the same radial stiffeners, the stator outer and inner rims supporting the annular spacers). If the DS gap is the same for both diaphragms, this not sound too difficult to build?
 
Now, to glue two different diaphragms to one of the stator/rim assembly, I'd imagine a 3 step process, such as this example (assuming same DS gap for both membranes):
1. Stretch the tweeter diaphragm and glue to the tweeter (inner) rim, cut off the excess material on the outskirt of the tweeter spacer rim
2. Glue the bass diaphragm (perforated with a hole of same size as the tweeter rim inner diameter) centered on the tweeter rim without tensioning (the tweeter rim could be made wide enough to ease that step) 
3. Stretch the bass diaphragm and glue to the outer spacer rim under tension, cut off the excess material on the outskirt of the bass spacer rim.
 
What did you use to make that cool animation? 

 

[chinsettawong]

Hi Arnaud,

While it would be fun to simulate, but what are you trying to achieve? For one thing, the sound from electrostatic driver shoots out like a beam, if you miss a sweet spot some of the frequencies are gone. I mean if you make a small tweeter and a woofer, when you move the cups around your ears, the sound you hear will change. As for woofer, the critical thing is the total active area. Given the same diaphragm tension and thickness, the larger it's active area the more bass you'll get.

As for double diaphragms, if you stack two drivers back to back, the total active area becomes 2 times as much. It is possible to achieve more bass without having to increase its physical width and height. It is also possible that the efficiency increases as well.

Wachara C,

 

[jgazal]

Quote:

 
That's the idea although I thought you could have the spacers (PCB or the like) mounted on both stators (the stators would have an outer and inner ring with the same radial stiffeners, the stator outer and inner rims supporting the annular spacers). If the DS gap is the same for both diaphragms, this not sound too difficult to build?
 
Now, to glue two different diaphragms to one of the stator/rim assembly, I'd imagine a 3 step process, such as this example (assuming same DS gap for both membranes):
1. Stretch the tweeter diaphragm and glue to the tweeter (inner) rim, cut off the excess material on the outskirt of the tweeter spacer rim
2. Glue the bass diaphragm (perforated with a hole of same size as the tweeter rim inner diameter) centered on the tweeter rim without tensioning (the tweeter rim could be made wide enough to ease that step) 
3. Stretch the bass diaphragm and glue to the outer spacer rim under tension, cut off the excess material on the outskirt of the bass spacer rim.
 
What did you use to make that cool animation? 

 
 
I cheated. I have used Google SketchUp and Microsoft Expression Encoder with Screen Capture. I did the easy way... 

 
Here is more and less what you describe:
 

 
And here are other crazy ideas:
 

 
 
 
 
I am with Wachara, perhaps the increased complexity worsens the reflections.