Jump to content

A DIY Shangri-la headphone for HIFI-Xmen


Kung
 Share

Recommended Posts

It started with repairing a badly damaged shangri la jr driver 2 years ago, I almost rebuilt the set from the diaphragm to the stators´╝îand of course the dust cover­čśŽ

IMG_2087.thumb.JPG.ad1ecd48399967fb2424961b6e962496.JPG

The first thing I discovered is that the 'nano-technology' used by hifiman actually comes from the covering used by indoors aircraft model enthusiasts´╝îthe name is OS film and it is still available on ebay or https://www.indoorffsupply.com/shop/os-film-25ft-roll-1

the driver of the hifiman estats is completely glued which is similar to the early stax lambdas, ┬ábut hifiman glue is significantly worse...after opening the driver i found that they only use OS film for their dust cover, the diaphragm is another PET film ~ a little bit thicker but with much better strength. In order to repair the diaphragm, I found a seller on Alibaba who can provide samples of Toray PET film with thickness below 2 microns at an acceptable price´╝îand use the inner tire stretcher to stretch the film┬álike any other diy estats┬á´╝îThen test it with a 10" subwoofer and minidsp umik-1 to find proper F0.

IMG_2747.thumb.JPG.19838a66d40854072ceaa8eb25d2c47e.JPGIMG_2746.thumb.JPG.5d049d3973e8d549eb911eebc64b383e.JPG

┬á ┬á ┬á ┬áconductive coating is another serious problem´╝îI've tried quite a few methods including antistatic agents like licron crystal, ANTISTATIK100´╝îfloor cleaner´╝îcarbon-based coatings such as graphene, and alcohol-soluble nylon, even tried nano-silver wire solution and metal evaporation...Some of them are sensitive to humidity, some arcs´╝îsome are ridiculously expensive, some have uneven surface resistance distribution, and some are sensitive to temperature changes. Yah~ finding the proper conductive coating is the most difficult part especially for a layman.┬áunexpectedly, I tried to search for hifiman patents, hoping to find something useful... here's a patent showing that they used some kind of metal oxide nano particles as conductive coating for their estats. emmm´╝îvery interesting...­čÖä┬áBy consulting with chemistry professionals, I decided to try a solution called ATO (antimony doped tin oxide), there are many antistatic agents I have used before have ATO ingredients though.┬áI got some ATO alcohol solution at concentration around 10~15%, it works really well, and┬ácoating appearance is somewhat identical to hifiman diaphragm!

IMG_2752.thumb.JPG.cefebee5972144efaec0e1883f9ee824.JPG

The last thing is to repair stators. hifimans probably use #150~300 brass mesh and solder their edges to a 2mm thick aluminum base plate PCB board, then glue the rest of middle part.┬áI've tried tensioning those brass mesh with embroidery frame and manual screen stretcher´╝îbut it is not tight enough to stretche brass mesh in this way, the mesh itself is too strong to tighten it up... then I found some brass sieves with different meshes on taobao and these machined copper screens are perfect for making DIY mesh stators...

IMG_2552.thumb.JPG.e30684a390b09f15bf954db7c8a0b017.JPGIMG_2753.thumb.JPG.d9247a43082d16b80903ab0b7d5ceb2d.JPG

after everything's ready´╝îi desoldered OG brass mesh, cleaned the copper foil, applied low temperature solder paste, and carefully soldered the new brass mesh with a heat gun.┬á

IMG_2750.thumb.JPG.9405fdacd2be6c16f0d7d6d347ca40ec.JPG

 

During the process of repairing shangri la jr, I started planning to make some bigger mesh stator estats... of course´╝îthe target is its big brother shangri la SR­čśłI received a set of the first batch of shangri la sr made in 2016´╝îhere is the headphone inside´╝îyap~inherited the tradition of hifiman with such horrible workmanship...­čś▒ however´╝îSR sounds not as technical as the X9000 though´╝îbut with more pleasant and passion.┬áSR uses 3mm thick aluminum base plate PCB instead of 2mm on JR version´╝î2 black things in the middle are to prevent the diaphragm from attaching stators´╝îThe D-S gap is around 0.7~0.8mm(thickness of glue included) which is identical to JR version´╝îactive area is close to 6400mm┬▓ ´╝îthat is pretty large diaphragm size for a commercial product┬áthough´╝îbut less 5300mm┬▓ of area on stator allows air to pass through.

IMG_2022.thumb.JPG.6411e6f34cb663738cd25da1311ddc81.JPGIMG_2020.thumb.JPG.54eec1aa28c3fdcb17b8a3fb73e70bd6.JPG

IMG_2598.thumb.JPG.53610c4a636cf7779555070648f6484c.JPG

I tried to use the 3mm FR4 board as mesh holder at first time´╝îhowever´╝îmesh stator disassembly is not an easy job...therefore´╝îI changed to use 1.5mm brass plates as the mesh holder and glued the copper mesh to it , brass mesh is also replaced with higher # red copper mesh for more convenient adhension. I spent most of time adjusting the tension of the diaphragm and finally found the tension needed for F0 similar to OG SR ,the clone driver has a slightly enlarged diaphragm(6839mm┬▓) ´╝î0.73mm fr-4 pcb spacer´╝î and everything is screwed to the housing cover.

IMG_2748.thumb.JPG.2065a0e8a2377d6f052c9478c9f79dcf.JPG

IMG_2694.thumb.JPG.5d7475737d8742e3f4c5c8c28d03ed0a.JPGIMG_2688.thumb.JPG.29d5daf5ba882cd76bf76e9be3b7c2b0.JPG

The frequency response of clone and SR are generally similar´╝îbut the sensitivity difference between the two is about 1.5dB actually´╝îSR is more diffused...clone one sounds a bit more clearer.┬áSR has obvious recession between 1-2kHz´╝îit may be caused by a sealing problem in the space between the driver and the earcup....┬áIt's unlikely to be solved without EQ, but it doesn't sound as bad as it looks from the graphics.

XSGL.thumb.jpg.4513d95a5fa2148e0740aeebad8c9859.jpg

IMG_2740.thumb.JPG.118ab214aeee39a437d6b25a0b81e107.JPG

one more thing...

 

Every shangri la must have a throne , so I made one for my mutation­čĄú

IMG_2749.thumb.JPG.b1b97863f346b27f9809f71cbb48dcbd.JPG

 

Edited by Kung
  • Like 28
Link to comment
Share on other sites

There will be plenty of demand for DIY repairs; this is what my second set of Shangri La drivers looked like after just over two years.  You will be one busy man.  The first set had a rattle in the right driver and was replaced, also.  Build dates were 2018, 2019, 2022.

The last Hifiman warranty replacement took 6 months to arrive.   Traded the new set for what I paid for them - a little more than a moldy bread sandwich.  At least the Hifiman sandwich comes with a 5 year warranty.

E12CFE3C-0024-4D1C-84C2-FA46C7ED75A4.jpeg

6B2839C1-CA28-4451-BF14-02B6479F3596.jpeg

CEC99B34-CAD7-4881-9894-CDD546736A16.jpeg

913EE92A-E759-4DE1-BEA2-705012795B94.jpeg

6B8E11E4-2ACF-47CD-B1F1-214DD3323B66.jpeg

  • Like 1
  • Confused 1
  • Sad 3
Link to comment
Share on other sites

7 hours ago, Kung said:

It started with repairing a badly damaged shangri la jr driver 2 years ago, I almost rebuilt the set from the diaphragm to the stators´╝îand of course the dust cover­čśŽ

IMG_2087.thumb.JPG.ad1ecd48399967fb2424961b6e962496.JPG

The first thing I discovered is that the 'nano-technology' used by hifiman actually comes from the covering used by indoors aircraft model enthusiasts´╝îthe name is OS film and it is still available on ebay or https://www.indoorffsupply.com/shop/os-film-25ft-roll-1

the driver of the hifiman estats is completely glued which is similar to the early stax lambdas, ┬ábut hifiman glue is significantly worse...after opening the driver i found that they only use OS film for their dust cover, the diaphragm is another PET film ~ a little bit thicker but with much better strength. In order to repair the diaphragm, I found a seller on Alibaba who can provide samples of Toray PET film with thickness below 2 microns at an acceptable price´╝îand use the inner tire stretcher to stretch the film┬álike any other diy estats┬á´╝îThen test it with a 10" subwoofer and minidsp umik-1 to find proper F0.

IMG_2747.thumb.JPG.19838a66d40854072ceaa8eb25d2c47e.JPGIMG_2746.thumb.JPG.5d049d3973e8d549eb911eebc64b383e.JPG

┬á ┬á ┬á ┬áconductive coating is another serious problem´╝îI've tried quite a few methods including antistatic agents like licron crystal, ANTISTATIK100´╝îfloor cleaner´╝îcarbon-based coatings such as graphene, and alcohol-soluble nylon, even tried nano-silver wire solution and metal evaporation...Some of them are sensitive to humidity, some arcs´╝îsome are ridiculously expensive, some have uneven surface resistance distribution, and some are sensitive to temperature changes. Yah~ finding the proper conductive coating is the most difficult part especially for a layman.┬áunexpectedly, I tried to search for hifiman patents, hoping to find something useful... here's a patent showing that they used some kind of metal oxide nano particles as conductive coating for their estats. emmm´╝îvery interesting...­čÖä┬áBy consulting with chemistry professionals, I decided to try a solution called ATO (antimony doped tin oxide), there are many antistatic agents I have used before have ATO ingredients though.┬áI got some ATO alcohol solution at concentration around 10~15%, it works really well, and┬ácoating appearance is somewhat identical to hifiman diaphragm!

IMG_2752.thumb.JPG.cefebee5972144efaec0e1883f9ee824.JPG

The last thing is to repair stators. hifimans probably use #150~300 brass mesh and solder their edges to a 2mm thick aluminum base plate PCB board, then glue the rest of middle part.┬áI've tried tensioning those brass mesh with embroidery frame and manual screen stretcher´╝îbut it is not tight enough to stretche brass mesh in this way, the mesh itself is too strong to tighten it up... then I found some brass sieves with different meshes on taobao and these machined copper screens are perfect for making DIY mesh stators...

IMG_2552.thumb.JPG.e30684a390b09f15bf954db7c8a0b017.JPGIMG_2753.thumb.JPG.d9247a43082d16b80903ab0b7d5ceb2d.JPG

after everything's ready´╝îi desoldered OG brass mesh, cleaned the copper foil, applied low temperature solder paste, and carefully soldered the new brass mesh with a heat gun.┬á

IMG_2750.thumb.JPG.9405fdacd2be6c16f0d7d6d347ca40ec.JPG

 

During the process of repairing shangri la jr, I started planning to make some bigger mesh stator estats... of course´╝îthe target is its big brother shangri la SR­čśłI received a set of the first batch of shangri la sr made in 2016´╝îhere is the headphone inside´╝îyap~inherited the tradition of hifiman with such horrible workmanship...­čś▒ however´╝îSR sounds not as technical as the X9000 though´╝îbut with more pleasant and passion.┬áSR uses 3mm thick aluminum base plate PCB instead of 2mm on JR version´╝î2 black things in the middle are to prevent the diaphragm from attaching stators´╝îThe D-S gap is around 0.7~0.8mm(thickness of glue included) which is identical to JR version´╝îactive area is close to 6400mm┬▓ ´╝îthat is pretty large diaphragm size for a commercial product┬áthough´╝îbut less 5300mm┬▓ of area on stator allows air to pass through.

IMG_2022.thumb.JPG.6411e6f34cb663738cd25da1311ddc81.JPGIMG_2020.thumb.JPG.54eec1aa28c3fdcb17b8a3fb73e70bd6.JPG

IMG_2598.thumb.JPG.53610c4a636cf7779555070648f6484c.JPG

I tried to use the 3mm FR4 board as mesh holder at first time´╝îhowever´╝îmesh stator disassembly is not an easy job...therefore´╝îI changed to use 1.5mm brass plates as the mesh holder and glued the copper mesh to it , brass mesh is also replaced with higher # red copper mesh for more convenient adhension. I spent most of time adjusting the tension of the diaphragm and finally found the tension needed for F0 similar to OG SR ,the clone driver has a slightly enlarged diaphragm(6839mm┬▓) ´╝î0.73mm fr-4 pcb spacer´╝î and everything is screwed to the housing cover.

IMG_2748.thumb.JPG.2065a0e8a2377d6f052c9478c9f79dcf.JPG

IMG_2694.thumb.JPG.5d7475737d8742e3f4c5c8c28d03ed0a.JPGIMG_2688.thumb.JPG.29d5daf5ba882cd76bf76e9be3b7c2b0.JPG

The frequency response of clone and SR are generally similar´╝îbut the sensitivity difference between the two is about 1.5dB actually´╝îSR is more diffused...clone one sounds a bit more clearer.┬áSR has obvious recession between 1-2kHz´╝îit may be caused by a sealing problem in the space between the driver and the earcup....┬áIt's unlikely to be solved without EQ, but it doesn't sound as bad as it looks from the graphics.

XSGL.thumb.jpg.4513d95a5fa2148e0740aeebad8c9859.jpg

IMG_2740.thumb.JPG.118ab214aeee39a437d6b25a0b81e107.JPG

one more thing...

 

Every shangri la must have a throne , so I made one for my mutation­čĄú

IMG_2749.thumb.JPG.b1b97863f346b27f9809f71cbb48dcbd.JPG

 

Very interesting!  I am very curious to learn about your coating.  Where do you buy it from?

 

Link to comment
Share on other sites

8 hours ago, chinsettawong said:

Very interesting!  I am very curious to learn about your coating.  Where do you buy it from?

 

A small bottle of sample I took from a local college friend, the ingredients included ATO powder with a particle size of 20-30nm, some surfactant, 
and the rest was some kind of alcohol solution. I was told to mix extra anhydrous alcohol at a ratio of 1:3 to dilute it for coating, but I didn't dilute it,
 just sprayed it on the sponge to apply the stretched film, and there were some traces of liquid on the surface of the film after drying , it looks the same as the hifiman's. 

 

4 hours ago, spritzer said:

That was excellent and what happened to that original Jr. set?  :)

that shangri la jr driver was later used as a test driver for testing different thicknesses of films.

  • Like 2
Link to comment
Share on other sites

12 hours ago, TammerDown said:

@Kung fantastic work. Would you consider making additional X-men (would be interested...)? What have you used for the headband?

I did make X-men round driver which has a similar stax 009/s driver housing but a bit larger´╝îand all X-men headbands actually┬áare made with hifiman's damaged headphone parts, however,┬áthe yoke is aluminum 3D printed...­čśł

IMG_2762.thumb.JPG.43665a1bd486547d6e62c209d171d39f.JPG

for comparison, the one on the right is the same size of 009/007 driver housing , the left one is the X-men with 1.5mm brass holder and #200 steel mesh stator in a 110mm driver housing.

IMG_2708.thumb.JPG.0ccc68742b29e6744972500fc2a26b54.JPG

the stator can cover diaphragm diameter┬árange from 84~92mm´╝îthe currently installed spacer is for 90mm┬ádiaphragm.IMG_2709.thumb.JPG.4b7581dd9430754a8dbce80b83d6abb5.JPG

I'm now running into some tolerances and earpad material problems....compared with perfectly matched sgl clone, the round estats have about 1dB consistency difference in sensitivity, and breathable fabric earpad causes roll-off of low frequency.

X0008.thumb.jpg.a2135b41dbca479b29df97b6c6e7439a.jpg

XSGL.thumb.jpg.e1721ab5f5fe8c8bd2137c9274fae928.jpg

IMG_2760.thumb.JPG.85d2a9215704484022891cc8c059d3a4.JPG

┬áthis estat is currently on hold...­čÖä

IMG_2701.thumb.JPG.d86f24755a0c904a7b3d4318a8b776a6.JPG

  • Like 8
Link to comment
Share on other sites

1 hour ago, Pirx said:

Really nice job. I would love to have such skills :)

How do you measure headphones?

I bought a 711 clone rig on Taobao, here is the link: https://item.taobao.com/item.htm?spm=a230r.1.14.18.b2407a72U22HcB&id=41416109881&ns=1&abbucket=3#detail  
Contains a clone coupler based on IEC-711 standard, a preamp, a silicone pinna and a calibration file based on B&K4128C.
the clone coupler has been confirmed by the headphone factory to perform almost the same as the G.R.A.S RA0045.
Unfortunately , the silicone pinna is not a standard pinna clone from G.R.A.S or B&K, resulting in an additional 5k~6khz attenuation in the frequency response diagram.

  • Thanks 1
Link to comment
Share on other sites

Updating more DIY mesh stators. 

 

Glue the brass holder to the copper sieve with ERGO 5500 or 5900

IMG_2767.thumb.JPG.ab5c3faf1d6b944bd3e5c2da3063afa0.JPG

 

then cut off the excess copper mesh with a nicking tool.

IMG_2768.thumb.JPG.dbbff15c61b09e5816f4dcbdccbea89c.JPG

IMG_2769.thumb.JPG.05019300930000feb484a6195d831257.JPG

IMG_2770.thumb.JPG.f4d828bcca306cb29491f679cb1f458f.JPG

 

Stators mesh size in the picture below ranges from #180 to #400´╝îIt is better to use a higher number mesh for making DIY mesh stators, the mesh with lower number size is difficult to glue...This may be the reason why stax no longer produce the original omega .┬áBoth omega and x9000 use about #80 mesh for their stator, ┬áand it is not easy to fix #80 mesh by hand, x9000 solves the problem by using thermo compression bonding...::)

IMG_2774.thumb.JPG.73e93c48385e666c840995424aeda603.JPG

 

 

  • Like 7
  • Thanks 1
Link to comment
Share on other sites

On 6/2/2022 at 2:13 AM, spritzer said:

Great work and Stax had endless issues with the SR-Omega... same as Sennheiser with the glass stators. 

It is likely that the #80 mesh is not easy to make a stator by hand, which leads to the failure of the omega driver, and the driver housing made by resin at that time was not as strong as later 007 009 though. There are some DIY estats such as Perun and Phenomenon in Russia, they use much lower number size copper mesh as stator. I have owned a pair of perun, the stators are made of about #10 copper mesh,  the mesh itself is very strong actually, and is glued to the PCB holder with structural adhesive.

As for the glass stator of the HE90, the glass is indeed rigid,┬áHowever, as a stator, it does not play the role in the Sennheiser advertisement at all. The only thing I can expect is that the glass is chemically stable and not easy to be corroded or oxidized, and Manufacturing microporous glass plate is extremely expensive┬áwhich also leads to the insane price of the HE90... Unfortunately, Sennheiser is now not planning to stop, they use even more expensive ceramic based PCB stator on HE1.­čśŽ

  • Like 5
Link to comment
Share on other sites

  • 3 weeks later...

@Kung, impressive work, looks quite time consuming to do in both carrying  out the repair  and sourcing materials , was it just to do a repair on a faulty pair you had or are you maybe thinking of doing this for others as well?

Link to comment
Share on other sites

12 hours ago, rod said:

@Kung, impressive work, looks quite time consuming to do in both carrying  out the repair  and sourcing materials , was it just to do a repair on a faulty pair you had or are you maybe thinking of doing this for others as well?

Just to repair my own headphones, cause I have not only hifiman cans but also several other old estats of Stax and Sennheiser.
it will be more reassuring to play with these old stuffs if I can do some repairs myself.:)

Link to comment
Share on other sites

  • 3 months later...

yes´╝îold ones are brass.┬á

i have considered carbon fiber or PMMA as mesh holder, after comparing the metal materials, I think the metal mesh holders have better looking though...

  • Like 4
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.