Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › My DIY electrostatic headphones
New Posts  All Forums:Forum Nav:

My DIY electrostatic headphones - Page 73

post #1081 of 2527

Umm..OK.  I hope I didn't hurt your feelings as it seems you took offense for some odd reason.  I think I made it clear what I considered the pros and cons of each method.


What do you want to know about the forces?  Constant charge, charge migration, and displacement are nothing new and have been discussed on HF in the past (I can't recall the threads but I'm sure this thread has information on those already). 


I agree that there will be people that want to get rid of any potential distortion and that's a good thing.  I was just pointing out that the displacement is so small that the added distortion is going to be very small as well.  Here's the thing though: displacement will still exist to some degree depending on how evenly you coat the sides.  I think it would be really hard to get the exact same amount of coating on each side of the membrane by hand.  In fact, you could theoretically end up with more displacement than if you had just coated one side from the start.


I'd wager that the coating outweighs the membrane pretty significantly.  Of course this will depend on which coating you decide on as well as how much you apply.  If you can provide me with exactly how much coating you are using on a single side I can provide you with numbers. 


Regarding my last statement, I'm simply talking about the uncoated side of the membrane (in conjunction with the stator on that side) being much more immune to things like dust and humidity since the membrane is essentially insulated on that side.  wachara can correct me if I'm wrong but I think this is why he chooses to coat one side. 


I'm not trying to discourage you from coating both sides.  This thread is all about experimentation.  I think you should try both methods and report your findings.  :beerchug:

post #1082 of 2527
Originally Posted by n3rdling View Post

Umm..OK.  I hope I didn't hurt your feelings as it seems you took offense for some odd reason. 


No offence taken. I simply had a technical disagreement with bidoux. I asked the question about coating both sides. He replied with a totally spurious answer regarding voltages and spewed up a load of irrelevant equations in justification of his answer. I came across a simple discussion of the issues with a mathematical description of the situation that I could show easily with a link. Then you jumped in with an attempt to dismiss the issue, which I thought ill-considered and unsubstantiated. I'm not interested in what you're prepared to bet on, sight unseen. If the weight of the coating is significant, then perhaps we need to think about the nature of the coating.



If these issues have been discussed previously no-one went to the trouble to provide links to where they were discussed or to simply sum up the conclusions.



post #1083 of 2527

As long as you don't put gobs of coating (depending on which type of course) on both sides it shouldn't make much audible difference - the 3 dB roll off point will still be supersonic.  As you said though, "perfectionists will never be satisfied".

post #1084 of 2527

This is very interesting. I'm doing a rigorous physics analysis to see exactly how important this single-sided coating distortion is, and to find out if it is irrelevant or not compared to driving forces.


I've been working the math here and taking it a bit further in analysis for a few hours and as best I can tell the paper is spot on, and putting a coating on both sides will remove a source of distortion (it is crucially important that both have their own bias resistors, though, or all benefits are gone... I wonder if any retail headphones have ever done this?).


Does anyone know what the typical diaphragm displacements are at a reasonable listening level? I don't have even a ballpark estimate, and it is of great importance to the analysis of this source of distortion as the distortion is highly non-linear with displacement



One thing this brings out is that lower bias and more drive voltage has to have less distortion, because you increase the drive force to distortion force ratio. Of course this would increase amplifier distortion, so it's hard to say what the sweet spot would be without enormous analysis, and it would be specific to every amplifier topology.

post #1085 of 2527

As n3rdling has pointed out, there are some difficulties in getting the coatings on both sides identical. Then there's the problem that the membrane itself should be an insulator (although I imagine Mylar is), and the question of feeding through 2 resistors.


The diaphragm supports (spacers) will need to be polished free of the slightest trace of swarf, as a tiny piece of copper could penetrate the membrane and produce a short. Perhaps it would, in fact, be easiest to simply coat one side...


There's a suggestion in the papers I linked that the membrane should be untreated and that reliance should be on the surface resistivity of the diaphragm being high, but not as high as the bulk resistivity of the diaphragm material.


As regards the diaphragm displacements, they're less than +/- 0.5 mm, certainly in the case of Wachara's, he mentions that occasionally the diaphragm contacts the stators when things go wrong.


I'm just trying to glue up my first membrane, I'm using the water-based carbon glue that's advertised as 'wire glue'. I don't even know if it will dry when it's covered with Mylar.



post #1086 of 2527
I'd have to look at the simulations I ran again but my recollection was that you could get in the +/- .3mm rms displacement at high SPL around the first resonance (50Hz or so) for the sr009. So much so I once thought the hole in the center of the stator had some to do with this...
post #1087 of 2527
Thread Starter 

Hi W,


Are you going to use the carbon glue so that both sides of the diaphragm can be energized?  Since the diaphragm will be very highly stretch, I doubt if this water base glue is going to do the job.  The best glue I would recommend is the contact cement or the yellow rubber glue. 


I've never tried coating on both sides of the diaphragm, but it certainly will be fun to find out if it will be beneficial.


Wachara C.

post #1088 of 2527
Thread Starter 

Regarding the spacer thickness, it all depends on your driver's dimension.  You can go as thin as you want, but then you need to stretch your diaphragm so much that it'll not collapse to one side of the stators.  When you do that, you lose the bass.  To my experience, for the diaphragm width of 78 - 80 mm, 0.5 mm is probably the thinnest you can use.   With Stax Lambda type of driver which has the diaphragm width of only 50 mm, the spacer thickness can go much thinner.  And that's why Stax Lambda series are more efficent than the Omega series.


Wachara C.

Edited by chinsettawong - 10/11/13 at 10:27pm
post #1089 of 2527
Originally Posted by wakibaki View Post


There's a suggestion in the papers I linked that the membrane should be untreated and that reliance should be on the surface resistivity of the diaphragm being high, but not as high as the bulk resistivity of the diaphragm material.


It's a great suggestion, but unfortunately it just doesn't work, the surface resistivity is simply too high at least for mylar. I've tried uncoated diaphragms, it never charges to make any sound at all.

post #1090 of 2527
Originally Posted by chinsettawong View Post

Stax Sigma sounded very interesting.  To clone it would be fun.  dt880smile.png
Dear Wachara
I aufgestellt this teo Times :-))
In The next Weeks I'll try

Regards Geoeg
post #1091 of 2527

I have received two of Wachara's creations here and color me impressed!


Only little time spent so far, with the Omega clone, and I like what I hear. It's sort of a mix of Omega 2 mk1 and SR009 through my amplifier.


I'll start with the cons, which all seem to point to a lack of juice on the amplification. It seems like a tough cookie to drive, I don't get much extension at both ends of the frequency range, the imaging is quite Left/Right/Center, and it overall sounds quite sleepy on my stock 727 amp.


Now for the pluses. I absolutely love the natural voicing and non-agressive nature of the phone. It reminds me very much of the 007mk1 in that regard. When listening to the 009 just after, I do hear much more detail but the tonality does seem off with some voices in the sibilance region (as if I could hear the same dreaded nasal coloration of some more recent generations of Lambda headphones). I have to spend more time to get a handle on this but, so far, the Omega clone is very gentle in the way it presents treble, very different from the 009. Even though I cannot get a proper seal, the bass is not bloated like it gets with an 007mk2. Overall the tonal balance is quite good, very pleasing and non-agressive, good presence and body, very far away from the stereotypical ethereal attribute of stat phones actually. 


I suspect the observations above on the Omega clone would change drastically with some beefier amplification but that's all I got with me for now (it might change soon).


I haven't even started to play with the open baffle can, looks like some busy schedule ahead ;).


MANY THANKS to Wachara, the man has got talent and great ears too, I could not dream of being to make and tune that stuff all by ear myself!



post #1092 of 2527

Just started to listen to the open baffle phones, oh my... Wachara, you're a genius at this! This time the imaging is fantastic as well as the treble extension. The bass rolls off but the mid-bass quantity is plenty to make it sound quite balanced overall.


Also, on the Omega clone, I get totally different response by pressing the cups a little so as to ensure the seal. In particular, it then goes down right to the sub-bass region. I think I will be making a few different measurements later on (the open baffle might be tricky though, as suspected).



post #1093 of 2527
Thread Starter 
Hi Arnaud,

I need to point out to you that the phones you were hearing weren't Omega clone. They're more or less my own version of the Orpheus. The phones are of Oval shape. I actually prefer the sound from these than the Omega clone.

When you compared them to Stax SR009, you'd noticed that the SR009 was more efficient. My phones are more like SR007 and should need a beefier amplifier, unfortunately. biggrin.gif

The open baffle phones would need even more power to drive. Have you got any other more muscular amplifier? tongue.gif

Wachara C.
Edited by chinsettawong - 10/14/13 at 8:10pm
post #1094 of 2527



Thanks for the clarification, I keep mixing things up. Ok, Orpheus clone makes more sense given the shape & size ;).



Originally Posted by chinsettawong View Post

Have you got any other more muscular amplifier? tongue.gif


Working on this ;). Honestly speaking, I can feel the potential and it's pretty darn awesome already!

post #1095 of 2527


Tonight's project was a bias knob on my PSU. A knob that everyone making electrostatic headphones should add to their power supplies immediately! It is extremely fascinating to be able to scan the bias from zero to high without having to re-wire anything.


The following data was collected at a variety of biases. One channel was tested, and the test fixture was completely untouched between all data collections, so this data is not contaminated by experimental error. The earpad was sealed to the text microphone panel. The results really surprised me, especially the THD!


It seems that as bias is lowered, treble gain increases at the 500hz crossover point (whatever that crossover is, seems to be in all electrostatic headphones, DIY or not). In fact that is exactly what I first noticed in listening. As bias increased, treble harshness decreased and relative bass amplitude increased.




I'm now loving my new headphones even more at 780V bias. A noticeable and significant improvement over 400V!

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › My DIY electrostatic headphones