My DIY electrostatic headphones
Apr 1, 2018 at 2:19 PM Post #2,911 of 4,061
I still haven't built the headphone housing for the electrostatic drivers I've made yet, but I thought it would be fun to cut some acrylic frames for a set of PC/desktop speakers for them. The speaker frame came out looking nice, and they sound really nice with the volume turned all the way up... not really enough bass to be the perfect desktop speakers, but the transparency and soundstage because of the electrostatic drivers sound really good to me. I used some tiny-XLR sockets for the connections at the speaker frames' back -- there seems to be some capacitance issue with the way I've run the wiring inside the case... Take a look at my photos.


Very nice!!! Where you get the flat cables?
 
Apr 1, 2018 at 10:44 PM Post #2,914 of 4,061
Diaphragms 0.8 mm and 0.5 mm.



Ear cups. used t= 0.2 mm (0.5 Oz) lambskin.


Hi Inuponken,

Are you using those clips as weights? I don't think you have enough tension there. I once tried water bottles weighing around 0.75 KG each and the tension was still not enough.

Your wooden frame and ear pads look really nice. Wow!

Have you put them together and played some music yet?

Wachara C.
 
Apr 14, 2018 at 3:34 AM Post #2,916 of 4,061
Hi Matt,

I have CNCed 1mm & 0.8mm aluminium sheet and also tried Gerber files to fabricate PCB on aluminium substrate board (I wrote about them earlier in this thread a while back, probably over a year ago). All these were not as good as my 1mm double sided FR4 stators. Predominantly flatness,... Still the best, in my experience (and most expensive) FR4 is the Bungard brand.

But, give the aluminium sheet a go, you may have different ideas and techniques which may produce better results....

Mmmm,... aluminium may not be a good choice for the spacers since they must be insulators.

David.
 
Last edited:
Apr 14, 2018 at 6:32 AM Post #2,917 of 4,061
That's what in my mind too, hence using a paper between the stator and spacer as insulators.

My main problem actually is the spacers being too thick, at 1mm using copper clad boards(PCB)

Also, anybody tried household food/cling wrap?
 
Apr 14, 2018 at 7:13 AM Post #2,918 of 4,061
Yes, 1mm spacers will be too thick for electrostatic headphones, bigger speaker panels ok.... although at this thickness you will not get squeal from dust!
There are many suppliers of FR4 in 0.5mm, 0.6mm, 0.8mm.
PCB fabrication houses will do 0.6, 0.8 and 1mm as standard, just send them your Gerber files. If you send to PCBway.com, the one I use, they are high quality and cheap.

I have, along with many other DIYers, tried ‘clingfilm’ or any of the other brand names. My results have been very good. Easy to tension, easy to glue to spacers and since it has elasticity it is good with bass. The longevity is not as good as Mylar. My choice is 2um Mylar, but others use various other thicknesses.

David.
 
Apr 14, 2018 at 8:00 AM Post #2,919 of 4,061
Would it be okay if I ask what would be the consequences if I use 1mm spacers? Also, the standard PCB here is 1.5mm. would it be okay to use that for the stators or best to stick with 1mm?

I actually don't want to do shipping for my materials. It's too much of a hassle here in our place that's why i want to source things locally.
 
Apr 14, 2018 at 8:55 AM Post #2,921 of 4,061
That's what in my mind too, hence using a paper between the stator and spacer as insulators.

My main problem actually is the spacers being too thick, at 1mm using copper clad boards(PCB)

The performance of stators is a wonderfully complex field with boundaries imposed on transmission by thickness of the plates, percentage of open area though the plates, and distance between the holes These three boundary parameters interact to generate

1. Resonances analogous to the transmission variations in a Fabry–Pérot optical interferometer, improved by decreasing the spacing between the diaphragm and the stator - which has severe limits, obviously.

2. Exaggeration [spreading] of the resonant effects by coupling between the air-masses of the closely spaced tubes and by interaction between the hole-end effects on inner and outer stator surfaces, improved by moving the holes apart - but then no sound gets through, and by making the plates thin - but thin enough means the plates cannot support themselves, which is bad, and become self-resonant, which is bad also.

3. Diffraction effects, where sound from each element of area traveling outwards through a directly facing hole is interfered with by sound that has travelled from the same element via the holes surrounding the primary hole and arrives with slight delays and subtracts or adds to the output of the main hole depending on frequency - It's a complex two dimensional comb-filter system and messes with the HF end.

The best-case combination of all these parameters is actually a very thin plate with a large proportion of open area generated by very large numbers of small holes that are very close to each other - screen-printing wire mesh is absolutely ideal acoustically but it has no mechanical strength. The mechanical issue of fixing such a stator in space is a problem that has been addressed in the Hifiman Shangri-La [actually it's quite a crude version of the mesh], so it can be done. The standard argument about lots of holes diminishing field strength is a red-herring with fine wire mesh - look at the protective guards round ultra-HV test facilities, the artificial lightning doesn't get through even though the mesh is high open area.

So, to focus on your idea, thin aluminium plates are a valid way to go - thin is good - but keeping them flat is an issue - it's dealt with in the Stax SR009 by bonding the thin plates to a thick but substantially open spider of the same metal, and is one approach I'm taking - incidentally, I'm amused by a number of 009-style designs in this forum that omit holes where the 009 has ribs - the ribs aren't there to block holes, they're there to reinforce the plates; in plates that do not need to be reinforced the rib positions should be filled with holes!! The possibility that the blank areas are used to control diaphragm resonances, which some may have thought, is certainly not true in the case of the 009 where the spider layout is symmetrical and would tend to favour some higher order resonances if such a mechanism were at work.

Also, anybody tried household food/cling wrap?

I'm more inclined to wonder why no one has hit on the static dissipative films that have the right level of surface resistivity without needing coatings. Hard to use them in production because minimum purchasable quantities are ridiculous, but blagging samples is not too difficult and even an A4 scrap is several diaphragms.

Ed Form
 
Last edited:
Apr 14, 2018 at 9:53 AM Post #2,922 of 4,061
The performance of stators is a wonderfully complex field with boundaries imposed on transmission by thickness of the plates, percentage of open area though the plates, and distance between the holes These three boundary parameters interact to generate

1. Resonances analogous to the transmission variations in a Fabry–Pérot optical interferometer, improved by decreasing the spacing between the diaphragm and the stator - which has severe limits, obviously.

2. Exaggeration [spreading] of the resonant effects by coupling between the air-masses of the closely spaced tubes and by interaction between the hole-end effects on inner and outer stator surfaces, improved by moving the holes apart - but then no sound gets through, and by making the plates thin - but thin enough means the plates cannot support themselves, which is bad, and become self-resonant, which is bad also.

3. Diffraction effects, where sound from each element of area traveling outwards through a directly facing hole is interfered with by sound that has travelled from the same element via the holes surrounding the primary hole and arrives with slight delays and subtracts or adds to the output of the main hole depending on frequency - It's a complex two dimensional comb-filter system and messes with the HF end.

The best-case combination of all these parameters is actually a very thin plate with a large proportion of open area generated by very large numbers of small holes that are very close to each other - screen-printing wire mesh is absolutely ideal acoustically but it has no mechanical strength. The mechanical issue of fixing such a stator in space is a problem that has been addressed in the Hifiman Shangri-La [actually it's quite a crude version of the mesh], so it can be done. The standard argument about lots of holes diminishing field strength is a red-herring with fine wire mesh - look at the protective guards round ultra-HV test facilities, the artificial lightning doesn't get through even though the mesh is high open area.

So, to focus on your idea, thin aluminium plates are a valid way to go - thin is good - but keeping them flat is an issue - it's dealt with in the Stax SR009 by bonding the thin plates to a thick but substantially open spider of the same metal, and is one approach I'm taking - incidentally, I'm amused by a number of 009-style designs in this forum that omit holes where the 009 has ribs - the ribs aren't there to block holes, they're there to reinforce the plates; in plates that do not need to be reinforced the rib positions should be filled with holes!! The possibility that the blank areas are used to control diaphragm resonances, which some may have thought, is certainly not true in the case of the 009 where the spider layout is symmetrical and would tend to favour some higher order resonances if such a mechanism were at work.



I'm more inclined to wonder why no one has hit on the static dissipative films that have the right level of surface resistivity without needing coatings. Hard to use them in production because minimum purchasable quantities are ridiculous, but blagging samples is not too difficult and even an A4 scrap is several diaphragms.

Ed Form


Or MattMatt,
You could just dive right in and try out your idea, adjusting, trying again like I did after stumbling onto this post almost two years ago. It’s part of the fun of DIY. For me, I’ve settled on .5-ish mm FR4 pcb stock as spacers and Mylar membranes just like Muamp had described earlier.
 
Apr 14, 2018 at 6:24 PM Post #2,923 of 4,061
Hi Ed,

I am one of the DIYer’s who has made FR4 stators with the spokes in a style of the 009s, but with my stators the spokes were not there for the purpose of reducing holes in the stator as you thought, no, they are there to add to the stiffness and rigidity of the stator, the stator being the thinnest FR4 I can get away with whilst still being rigid enough.

I have listened to the Shangri La and I was amazed at how good they sounded, with the wire stators similar to how you intend to make, but I am also very impressed with the 007s. The 007s are still one of my all time favourites along with my DIY Spyders. I have learnt the real test is in the listening.

Static dissipative films that I have seen are all too thick to be diaphragms. I did think of trying some, but not ever found any suitable.

David.
 
Last edited:
Apr 15, 2018 at 1:37 AM Post #2,924 of 4,061
I still have problems with acquiring 0.5mm thickness FR4 for spacers. I think I can source here stainless steel and have it laser cut. I'll use FR4 with 1.5mm thickness for the stators. Because of stainless steel is also a conductor, I'll make the cut the inside ring slightly bigger than the active area of the stators. Would that be okay? Here's my design as of the moment for the spacers and stators.
upload_2018-4-15_13-36-38.png


upload_2018-4-15_13-37-10.png
 
Apr 15, 2018 at 1:19 PM Post #2,925 of 4,061
Hi Ed,

I am one of the DIYer’s who has made FR4 stators with the spokes in a style of the 009s, but with my stators the spokes were not there for the purpose of reducing holes in the stator as you thought, no, they are there to add to the stiffness and rigidity of the stator, the stator being the thinnest FR4 I can get away with whilst still being rigid enough.

:ksc75smile: I was only joshing - you guys have what are plainly fine-sounding, working cans: I've got lots of background knowledge and big dreams!

I have listened to the Shangri La and I was amazed at how good they sounded, with the wire stators similar to how you intend to make,

I've already done some work on using mesh, but I've not yet had any success with the process of stretching the mesh out flat and uniting it to a support structure. I attended the MACH advanced manufacturing show on Thursday and was intrigued to find that my two piece stator design of a chemically milled perforated plate and a laser cut support spider could actually be replaced with a one-piece 3D-printed aluminium part. Several companies were willing to make the item - but some objects I saw displayed gave me new ideas for a much lighter stator with tall, slender ribs embracing much more of the plate area while occluding a distinctly smaller area than the clumsy laser-cut spider design I had gone for up to this point.

Ed Form
 
Last edited:

Users who are viewing this thread

Back
Top