[raptor18]

  I got some Sonion 26-e25 drivers which have a 2mm wide port. I bought some 1.6mm pipe and fits nice and tight just with the pressure. However the sound filter is 2mm wide too and sonion recommend to put it about 1/3 deep in the lenght of the pipe. Do you guys have any advice about how to push it through the pipe without breaking it?
 
My other options would be to use 2mm pipe, which could work fine but I guess I'd have to see if it changes the sound signature and I'll have to glue it to the driver. I could also put the 2mm filter just at the end of the pipe rather than pushed at third way through.
And should I put a dirt filter at the end of my pipe in addition? If so, which material would do the trick?
 
I also see that there was discussion about using pipe as a horn form 1.5mm to 2.5 in the precedent page. So would it be better to put a lenght of 1.6mm pipe and then a lenght of 2mm? And does PVC tubing does the trick or teflon really has some advantages?
 
Thank you if you have any advice.

 
I'm having the same issue. Even if i manage to push the damper in a bit, it is not aligned in the tube.
Any tips? I really want to try a horn. 1,5 mm -> 2 mm ID.
 
I know i could cut the 1,5 mm ID tube and insert a 2 mm ID tube with the damper in between but i have seen other manufacturers pushing in dampers in 1,5 mm ID tubes.
 
 

 

[MuZo2]

Do you have tool for dampers? there is special tool on one side you can push it while on other side there is screw with which you can pull it.

 

[raptor18]

That was an expensive tool.
 
http://www.digikey.com/product-detail/en/knowles/BF-1778-000/423-1217-ND/4867480
 
Then we have Sonion damper tool.
 
http://www.sonion.com/Products/Micro%20Mechanicals/~/media/Files/Products/Data%20Sheets/Electromechanicals/SSD/SSD_DS_rev013.ashx
 
This is way cheaper.
 
Do you have any of these? Is it confirmed that they can push dampers in 1,5 mm ID sound tubes? I'm asking because sonion and knowles state their dampers are for 2 mm ID sound tubes.
 
Because if you look at the drawing:
 
http://www.knowles.com/eng/content/download/3162/37787/version/2/file/BF-1778.pdf
 
The tips is 2,6 mm, a lot more than 1,5 mm.

 

[MuZo2]

I bought the tool from taobao for 5$
 
 

 

[raptor18]

Super
 
I'll try to get one.

 

[raptor18]

  Opps I forgot Peter is still lurking around here lol
 
Anyway, the problem with making (long) horns for tweeters is that silicone tube has to be used no matter what, and I'm not a fan of it. It's common knowledge that high frequencies are affected by transmission path the most. I don't have the exact figure but I'm pretty sure absorbing coefficient of silicone isn't that ideal compared to, say titanium (I'm a big fan of that material btw). Plus it's always recommended to have a straight transmission path for tweeters, which is virtually impossible with CIEM shells. To me, when using long horns with CIEMs you are actualy amplifying degenerated high frequencies due to the use of long silicone tube (no offence here Peter, I like your work, it's just me being not comfortable with silicone, that's all).
 
Btw that prototype up there used no silicone for its tubing
 
And if you guys look at how IEMs are designed these days, treble extension is not the top priority out there, unless you want to create a mini HD800. Most people above thirty can't really detect frequencies above 16khz during normal listening so why aim for something higher than that? The only obvious advantage I can see in horns is their capability of lowering the impedance at the tweeter's nozzle to smooth out the upper region. However CIEMs in general don't really need that feature because of their deep insertion.
 
A short horn is do-able, but there's not much point in having one though.

 
@tranhieu
 
I know this is an old post, but i have a very relevant question.
 
Since high frequencies are affected by transmission path, isn't it better to use as large diameter tube as possible to lower the effect of the tube attenuating high frequencies than to use a smaller ID tube, which will attenuate high freqs a bit, so a stepped horn can be created? 
 
To me, horns are not wise, just use maximum possible ID tube all the way if physical space permitts.. But then again, even Sonion keep using them in their design guidelines. So at the end of the day, i don't understand

 

[tranhieu]

   
@tranhieu
 
I know this is an old post, but i have a very relevant question.
 
Since high frequencies are affected by transmission path, isn't it better to use as large diameter tube as possible to lower the effect of the tube attenuating high frequencies than to use a smaller ID tube, which will attenuate high freqs a bit, so a stepped horn can be created? 
 
To me, horns are not wise, just use maximum possible ID tube all the way if physical space permitts.. But then again, even Sonion keep using them in their design guidelines. So at the end of the day, i don't understand

You are partially correct. However, if you look into it carefully having a 'maximum possible ID' is also another kind of stepped horn with the spout of the driver being the 1st step.
 
Back to why horns are being used here, there's a paper on Etymotic's page (or you can google it for any relevant audiology book) talking about boosting specific (high) frequencies using horns. That's kind of similar to recovering the part of the frequency range that has been lost due to being absorbed by the tubing material.

 

[roughington]

I see a lot of drivers on taobao that go for a decent amount less than they would on mouser or digikey I was wondering if anyone could vouche for them?

 

[Hutnicks]

I see a lot of drivers on taobao that go for a decent amount less than they would on mouser or digikey I was wondering if anyone could vouche for them?

There are altenatives. Sony's MH1 and HPM 62 and 64's can be had dirt cheap. All have very good drivers to be salvaged.

 

[Xymordos]

You can even get the JVC FX650&850 drivers on taobao. I've also seen the CKR9/10 driver and the Sony BA drivers being sold. . 

 

[s1190]

There's even IE800 driver on taobao, but I do doubt it's real.

 

[s1190]

You are partially correct. However, if you look into it carefully having a 'maximum possible ID' is also another kind of stepped horn with the spout of the driver being the 1st step.

Back to why horns are being used here, there's a paper on Etymotic's page (or you can google it for any relevant audiology book) talking about boosting specific (high) frequencies using horns. That's kind of similar to recovering the part of the frequency range that has been lost due to being absorbed by the tubing material.

The same apply to using different material of tubing (etc stainless steel or titanium). Read this somewhere on knowles application note. Metal tube enhance HF around 2db. I have tried metal tube without much success.

Main problems:
1) bending the metal to fit inside the shell ( K&S metal bendable tube may work which I haven't try).
2) how to fit an ascoustic filter into the metal tube ?

I also read that for JH audio, they use smaller ID(1.2mm if I recall correctly) for exiting tube. Common tube size (3mm x 2mm) work well for 1 or 2 bore. For triple bore and above, the use of smaller ID is necessary. Tubing that is small enough to bend and doesn't collapse. Otherwise using one big horn exit, this approach work well with people with smaller canal.

 

[Clarence1]

   
@tranhieu
 
I know this is an old post, but i have a very relevant question.
 
Since high frequencies are affected by transmission path, isn't it better to use as large diameter tube as possible to lower the effect of the tube attenuating high frequencies than to use a smaller ID tube, which will attenuate high freqs a bit, so a stepped horn can be created? 
 
To me, horns are not wise, just use maximum possible ID tube all the way if physical space permitts.. But then again, even Sonion keep using them in their design guidelines. So at the end of the day, i don't understand

 
I can't help you, but if we look at Sonion's design guidelines, page 9, graphic on the right, we can see that the 1.6mm tubing (purple line) allows higher frequencies than the 2mm tube once it's past over 10khz, but a "drop" at around 10khz. 
http://www.sonion.com/~/media/Files/Academy/Academy_ProAudio__DocCode_304_V_001_Web.ashx
 
Otherwise, I'll try to go with 2mm diameter to avoid problems with the damper insertion I think, or I will try a stepped horn shape, as some here suggest.

 

[raptor18]

I honestly still don't understand, so i have a problem and will need to read more about this 

 
I understand we are boosting high freqs with a horn, but we are at the same time killing the high freqs with a smaller ID tube so we can create that horn.
 
@Clarence1
 
They only use the IEC711 coupler in those graphs, so we cannot trust anything above 8 khz.
Just look at the more recent knowles drivers tech specs, they provide a graph from both iec711 and 0,4cc coupler which is more trustworthy for higher frequencies. Very well done by Knowles.
 
Edit: I found it, will read it.
http://www.etymotic.com/media/publications/erl-0048-2003.pdf
 
Very good site with excellent material for us!
 
http://www.etymotic.com/publications

 

[jbr1971]

  I've been using Clear Balistics for a few weeks and I only achieved a bubble-free negative when I put the molds in the oven at 265F for about 2 hrs.  The bubbles slowly rise and you get a crystal clear negative.  However, this means you can't use wax.  Good stuff though...

 
Thanks again for the tip on putting the mould in the oven. It took a few tries to get the result I wanted (once I figured out how to compensate for the gel contracting during cooling), but it worked great. A perfect cast with very few bubbles.
 
I took a picture comparing the mould I used for my first set of IEMs alongside the new one. The air bubbles in the first one made for a lot of extra time and work fixing the shells. I cannot wait to see what the new shells will look like...
 
Edit: The shells came out perfect. Almost no bubbles, clear, and fit perfectly first try. Build time has significantly dropped from the first set already. Less than 2 days to re-mould and cast good shells, as opposed to 3 to 4 weeks the first time.