Home-Made IEMs
May 30, 2016 at 10:29 AM Post #4,788 of 15,989
Start reading backwards 
wink.gif

 
Quote:
  For those who don't know, here is a Japanese blog with a list of drivers used by in various CIEMs
 
http://84audio.blog27.fc2.com/ 
https://docs.google.com/spreadsheets/d/1jguxmYfEmZdHEGIUDvQUnfm3vhxo51WeNKKG0_oRtMA/pubhtml
 
It gets updated regularly, though it doesn't have Xovers info, it's a great source 

 
Jun 1, 2016 at 6:09 AM Post #4,789 of 15,989
  Subbed. I made my speakers via a DIY kit and some custom work, and I'd love to make a custom IEM in the same manner.
 
Since there is no OP with a summary of info, what's the best starting point for me to learn crossover assembly, favorite designs/parts, etc?

Coil is right, try to go backward, there is a googledocs made by Furco at some point collecting various posts which explains all the basics.
https://docs.google.com/document/d/1jU37oWefx2dGS8dWw5TVkugEtwJh_bh3fuUBNe2HrQg/edit?pref=2&pli=1
 
There is a also a subreddit much younger than this thread (with much less info and less to go through) that gather information on CIEMs. https://www.reddit.com/r/DIEMs
 
It is a long thread to go through, I did when it was about 240 posts already, took me 2 weeks but was worth it. If you already know crossover electronics from speakers design then it will be easier for you to grasp the concept.
 
Here is a Sonion (Sonion is a brand, manufacturer of balanced armature drivers)  brochure explaining the basics 
http://www.sonion.com/~/media/Files/Academy/Academy_ProAudio__DocCode_304_V_001_Web.ashx
 
Sonion and Knowles are dominating the market of BAs, among the many drivers they have, some are quite similar (with almost the same frequency response and impedance).
Sony lately started  to produce their own BAs mainly for their latest products.I think they are also used by Klipsch in their latest IEM lineup but cannot be bought individually. Other manufacturer are based in China and Japan (more info here).
 
You could try to open the QT5, I am sure you could learn a lot about designing IEMs and also share with us diyers the secret sauce.
 
Jun 1, 2016 at 7:21 PM Post #4,790 of 15,989
So finally i got CI-22955 with 10 Ohm in 22 mm 1,5 mm ID , and ED-29689 with 2,2uf and white damper with the suggested horn in a 12 mm long tube. Tha sound is nice in bass, quite driven and punchy. Mids are really good imho, but heights still a little low, and not at all "shimmering". Would it need another driver to get better heights? What does half coil and reversing polarity do for a difference?
 
Jun 2, 2016 at 2:48 AM Post #4,792 of 15,989
Driving high frequencies is very difficult due to the nature of BA's (and speakers in general) where the impedance raises a lot compared to the rest of the register. The higher the frequency the higher the impedance which makes it more difficult to drive*.
That is why you want to have a BA with relatively low impedance so you can lower the high frequency impedance to reasonable level.

It is described in this article (graph 2)
http://www.knowles.com/eng/premiumsound/Resource-center/The-Science-of-Premium-Sound-Using-Miniature-Transducers

Running the ED-29689 half coil lowers the impedance by half, so you will have more reasonable high frequency impedance to be able to drive the high freqs.
 
However, when doing so, you will drive a very low impedance BA which can be bad for your damping factor and even your player which needs to output a lot of current.
 
That is why you need to create a crossover quiet high so you don't "play" the lower areas where the impedance is low.
 
*Higher impedance isn't harder to drive, it just creates a big difference between the high freqs and the rest of the frequency band which has a lower impedance and will play louder than the high freqs.
 
Jun 2, 2016 at 6:35 AM Post #4,795 of 15,989
Most resins react and "exhale" toxic compounds, cyanide included. I have worked/studied some resins and worked with them while working with carbon/glass-fiber molding and these reactions should be carried out ONLY in ventilated room/chamber! 
 
Jun 2, 2016 at 10:28 AM Post #4,796 of 15,989
So finally i got CI-22955 with 10 Ohm in 22 mm 1,5 mm ID , and ED-29689 with 2,2uf and white damper with the suggested horn in a 12 mm long tube. Tha sound is nice in bass, quite driven and punchy. Mids are really good imho, but heights still a little low, and not at all "shimmering". Would it need another driver to get better heights? What does half coil and reversing polarity do for a difference?

Pete, I was up most of the night working on this.  I'm stubborn and determined to use the DTEC.  I had probs with CI and DTEC cancelling other out and switching polarity helped this out a lot! (Thanks Piotrus!!!)  In fact, I think switching polarity on CI and ED helped too.  But, the real fireworks happened when I did away with resistors on CI and DTEC and ran them wide open and put a resistor on the ED.  WOW!  What a difference!  I got CI and DTEC to play nice but every time I hooked up ED, the sound turned to crap.  After I incorporated a 10 ohm resistor to ED at half coil, it mellowed it out and harmony and bliss filled my ear holes.  Half coil bumps SPL by a few dbl, maybe 6?  You do not need another driver, ED to my ears gets way up there!  Happy tuning!
 
Jun 2, 2016 at 1:28 PM Post #4,799 of 15,989
  Pete, I was up most of the night working on this.  I'm stubborn and determined to use the DTEC.  I had probs with CI and DTEC cancelling other out and switching polarity helped this out a lot! (Thanks Piotrus!!!)  In fact, I think switching polarity on CI and ED helped too.  But, the real fireworks happened when I did away with resistors on CI and DTEC and ran them wide open and put a resistor on the ED.  WOW!  What a difference!  I got CI and DTEC to play nice but every time I hooked up ED, the sound turned to crap.  After I incorporated a 10 ohm resistor to ED at half coil, it mellowed it out and harmony and bliss filled my ear holes.  Half coil bumps SPL by a few dbl, maybe 6?  You do not need another driver, ED to my ears gets way up there!  Happy tuning!

Great to hear it. My thoughts are that you were still cancelling ED, if you got a "crap" sound but none the less. It would be great idea if you could post schematics of your design I could check it out in spare moment and let you know if there's something you could do better with implementing ED.
 
If you decided to post graphs, please post one after your new solution and before so that we can see what exactly you'd changed.
 
EDIT: Oh I see you posted a schematics, but more details would be nice - right now I cannot tell polarity of the drivers. - You can mark either spout or coding (numbers) side so that we can see what position your drivers are.
 
Jun 2, 2016 at 1:45 PM Post #4,800 of 15,989
  Great to hear it. My thoughts are that you were still cancelling ED, if you got a "crap" sound but none the less. It would be great idea if you could post schematics of your design I could check it out in spare moment and let you know if there's something you could do better with implementing ED.
 
If you decided to post graphs, please post one after your new solution and before so that we can see what exactly you'd changed.
 
EDIT: Oh I see you posted a schematics, but more details would be nice - right now I cannot tell polarity of the drivers. - You can mark either spout or coding (numbers) side so that we can see what position your drivers are.

Sorry, I realized my DTEC was wrong.  This should be more accurate...
 
https://docs.google.com/drawings/d/12XHs-AufgU2IE7Yzq-O08dijiVQDOBchOBTqqkaIfQw/edit
 
I really can't give a graph of before because I tried wiring it any way possible before coming to this.  Thank again, Piotrus!  Us newbies are very fortunate to have you around!   
 

Users who are viewing this thread

Back
Top