Home-Made IEMs
Jun 1, 2012 at 10:32 AM Post #1,291 of 15,989
  Here is my first attempt: Sonion 2323 + Sonion 33a007    ...The result was not what I was expecting!


Nice! Not what you were expecting in what way? That's a TF10-like setup right there; was it unexpectedly unlike a TF10?


He didn't mention any crossover what so ever. He also didn't use any sound damper. It looks good tho.
 
Jun 1, 2012 at 12:44 PM Post #1,293 of 15,989
Jun 1, 2012 at 1:09 PM Post #1,294 of 15,989
This is the design I have followed:
 

1 uF cap and 10 ohms resistor
green damper for the 2323 driver
yellow damper for the 33a007 driver
(the dampers were removed at the time of the pictures)
 
Jun 2, 2012 at 1:06 PM Post #1,295 of 15,989
Great to see some life back in this thread. I've also started making my own iems. I'm using a TWFK with an ED23147 as the woofer.
 
I've done a pretty complicated crossover incorporating a lowpass filter and two highpass filters. Also an L-pad to drop the sensitivity of the woofer, with the added benefit of smoothing it's impedance changes a bit.
 
I've already cast a solid epoxy piece as a test. Now I need to figure out a method for casting a hollow piece, which is difficult as the epoxy has a very long cure time.
 
 
Jun 7, 2012 at 12:52 PM Post #1,296 of 15,989
I thought I'd join in on this thread. After all, "How hard can it be?"....
 
Here are some pictures of a hollow shell I managed to make:
 

 

 
Got a few ideas on the next steps but still unsure.
 
Jun 8, 2012 at 6:42 AM Post #1,297 of 15,989
It can be really difficult. Especially developing a proper crossover.
 
I found that it's almost easier to cast a solid piece and hollow it out with a dremel, than to try to cast a hollow one. This is with the epoxy I'm using though. It might be different with other materials.
 
Also if anyone is interested, I measured full impedance graphs from the TWFK and ED-23147 drivers. These are quite useful for figuring out needed component values for crossovers.
 
Jun 8, 2012 at 1:03 PM Post #1,299 of 15,989
Ok, here's a screencap from a spreadsheet I made. These measurements were made using the process mentioned here: http://sound.westhost.com/tsp.htm
 
As for the accuracy of my tests, you can compare these to the few values provided by knowles(impedance @500hz and @1khz). They're quite close.
 
These drivers have some pretty substantial impedance changes. Especially the ED-23147, those peaks are huge!
 

 
Jun 12, 2012 at 8:01 AM Post #1,302 of 15,989
I think I'm just gonna keep it simple. No crossovers as my soldering skills are rubbish.
As others have said, accoustic dampers should surfice (eg: Sony xba).

Just ordered a pair of TWFK and Ci with a grey filter for the rwfk and red for the ci driver.
 
Jun 12, 2012 at 8:19 AM Post #1,304 of 15,989
Quote:
I don't think the Sony XBA's have acoustic dampers, they diaphragm is specifically tuned to a different frequency in each driver.  Either way, they don't suffice, I can't listen to the XBA-3 or XBA-4 for more than 30 minutes, I can't stand them.


Hmm I listened to XBA3 for a few minus and I liked the sound. I listened it from clip+ though from Samsung galaxy S it sounded very lacking
 
Jun 12, 2012 at 1:46 PM Post #1,305 of 15,989
While I'm still working on my crossovers for my triple driver project, I made these:
 

 
 
I took drivers from and old pair of Shure E3C iems, that had a broken cable. Single Knowles BK-28562 driver inside. Good midrange, bass and a treble are a bit lacking, but there's only so much a single driver can do...
 

Users who are viewing this thread

Back
Top