Home-Made IEMs
Jan 17, 2018 at 6:56 AM Post #7,261 of 15,989
Ok guys I need some help. I’ve been around watching for years a tried and tried again. I have a 36watt nail UV setup and I’m just not getting consistent shells. I have weak spots and my curing time is upwards of 8 minutes. Im using krystaloid (its 2 years old and used probably 10 times). Help please!

I have a 36W nail lamp as well and anything above 2 minutes gives me completely solid chunks of Fotoplast. Are you using Fotoplast (or similar)?

Given the age and usage of your krystaloid, how clear is it? It may be "cloudy" enough that the full strength of the light is not getting through. Do you have a picture?

Also, I found the underside portions of my shells were not as thick as the rest, so whenever I make shells I set the lamp on a mirror which has helped a lot in fixing those issues.
 
Jan 17, 2018 at 9:30 AM Post #7,262 of 15,989
I have a 36W nail lamp as well and anything above 2 minutes gives me completely solid chunks of Fotoplast. Are you using Fotoplast (or similar)?

Given the age and usage of your krystaloid, how clear is it? It may be "cloudy" enough that the full strength of the light is not getting through. Do you have a picture?

Also, I found the underside portions of my shells were not as thick as the rest, so whenever I make shells I set the lamp on a mirror which has helped a lot in fixing those issues.

What are your normal cure times? I just ordered some more krystalloid but I have the gelatin I'll try while i'm waiting for it's arrival. looking at this it looks pretty cloudy to me.
 

Attachments

  • Gel.jpg
    Gel.jpg
    226.2 KB · Views: 0
Jan 17, 2018 at 11:06 AM Post #7,265 of 15,989
Contact Chris Perkins at Lightning Enterprises (cperkins@lightningenterprises.com)
He'll make it happen You can order in 1 lb increments

Here is where Chris P. sells it by the pound...

http://www.fusionet.com/

He also has Lightning Enterprises but I think he just sells it by the tub on that site. However, it has more products in general.
Good luck!
 
Jan 17, 2018 at 6:25 PM Post #7,266 of 15,989
What are your normal cure times? I just ordered some more krystalloid but I have the gelatin I'll try while i'm waiting for it's arrival. looking at this it looks pretty cloudy to me.

The initial cure is 45-50 seconds (I cover the top with a quarter so I don't have to pry a solid top off), dump the excess Fotoplast, then final cure in alcohol.

That gives me a good shell thickness while leaving me room for parts.

Keep in mind, I have big ear canals, so if yours are small, you will want to try a shorter time to make sure you don't end up with solid canal portions.
 
Jan 17, 2018 at 11:51 PM Post #7,267 of 15,989
I think your negative form is to big, so the light couldn’t reach the fotoplast evenly and this is why your shells are not so good.

Also try to use the nail lamp upside down. Put your form into a height that the side lamps are on the same height.

But the main reason is in my opinion the diameter of the form.
 
Jan 18, 2018 at 2:15 AM Post #7,268 of 15,989
I think your negative form is to big, so the light couldn’t reach the fotoplast evenly and this is why your shells are not so good.

Also try to use the nail lamp upside down. Put your form into a height that the side lamps are on the same height.

But the main reason is in my opinion the diameter of the form.

That’s a really good thought. Im going to make some smaller forms tomorrow and see what that does. Im currently testing out the gelatin as I threw out all my krystaloid due to it being so cloudy.

On another note, does fotoplast have a shelf life? Mine is 2 years old. Wondering if that’s the problem. I have shells cooking for 9 minutes right now and I’m barely getting a wall.... they are colored with alcohol ink, just in case that matters
 
Last edited:
Jan 19, 2018 at 12:00 PM Post #7,269 of 15,989
I need some help now too. I have ordered some Knowles BAs, some GV-Quads (TWFK+HDTEC), GKs (TWFK+CI), CIs and TWFKs for a 2*TWFK+CI combination or something like that. Now I want to build universal IEMs with the GK and maybe the quad, which I want to use with Comply or silicone eartips. So here come the questions:

1. Now there are some DIY Shells for sale on Aliexpress, most of them look like crap, some of the ones sold as Shure 535 shells seem so be okay in quality and in fact I got myself some of them. Now how to I combine the 2 driver outputs, each with different dampers into the small sound tube of a universal IEM? Should I just use some heatshrink construction at the very end (take the 2 small tubes, than a bigger one stuck on a small one and just heatshrink over all of it)? Or try to get two small soundtubes next to each other in the bore, maybe widen it a little more with a small dremel tool? Or draw something up in CAD and have it 3D-Printed (I don't own a printer myself sadly - not yet..)?As I see it, most manufacturers use more or less a single piece to route the soundtubes together (or take a different approach like Logitech did with the UE900 and the two bores). I just don't know how much of a hassle it will be to implement this kind of solution in a DIY IEM.

2. If I am not satisfied with those shells after all, which relatively cheap China IEMs or so have a relatively universal fit, a larger housing and would therefore be a good template for making my own universal IEM shells with Dreve Fotoplast? KZ ZST look okay, but I saw some pictures here with DIY universal IEMs that had a slightly different, less edgy shape to them. So if you have suggestions, I would be very happy to hear them!
 
Jan 19, 2018 at 2:49 PM Post #7,270 of 15,989
2. If I am not satisfied with those shells after all, which relatively cheap China IEMs or so have a relatively universal fit, a larger housing and would therefore be a good template for making my own universal IEM shells with Dreve Fotoplast? KZ ZST look okay, but I saw some pictures here with DIY universal IEMs that had a slightly different, less edgy shape to them. So if you have suggestions, I would be very happy to hear them!

I did use the KZ ZST's as template for my universals and it worked great! I used the same process as you would use building a custom iem but instead of using ear impressions I used the KZ's, as simple as that
 
Jan 19, 2018 at 4:30 PM Post #7,271 of 15,989
Hello again! I have been trying to finish up the electronical notch-filter and it has worked pretty well! My problem, however, is that the high frequencies gets affected aswell.
Here is a graph of the headphones before and after the use of the filter:

No filter.png Nya komponenter v1.png

The changes from around 4-5kHz all the way to around 13kHz is quite alot.
I used a notch-filter calculator and the components I used was a polyester capacitor at 4.7uF along with an inductor of 560uH. The sound is good, but I wouldn't mind keeping more of the high frequencies to be able to tame them further with a screenfilter if needed.

I used this calculator: http://sim.okawa-denshi.jp/en/RLCBesample.htm

So, my question is this: Is there any other way of making this kind of a notch-filter without affecting the high-frequencies? Or can I modify the circuit to preserve it better?
I tried connecting the components in parallel, but it didn't do much difference at all (it only cut it 4-5db).

Thank you very much!

/Anders
 
Jan 19, 2018 at 8:58 PM Post #7,272 of 15,989
Hello again! I have been trying to finish up the electronical notch-filter and it has worked pretty well! My problem, however, is that the high frequencies gets affected aswell.
Here is a graph of the headphones before and after the use of the filter:



The changes from around 4-5kHz all the way to around 13kHz is quite alot.
I used a notch-filter calculator and the components I used was a polyester capacitor at 4.7uF along with an inductor of 560uH. The sound is good, but I wouldn't mind keeping more of the high frequencies to be able to tame them further with a screenfilter if needed.

I used this calculator: http://sim.okawa-denshi.jp/en/RLCBesample.htm

So, my question is this: Is there any other way of making this kind of a notch-filter without affecting the high-frequencies? Or can I modify the circuit to preserve it better?
I tried connecting the components in parallel, but it didn't do much difference at all (it only cut it 4-5db).

Thank you very much!

/Anders

Looks amazing!

Did you try just adding a high pass filter before/after the notch, perhaps around 1k? This could help tame the lows and relatively balance it out again. I'm not sure if this will cause unintended interactions between the 2 filters, though.
For multi-driver designs, maybe a resistor on the woofers could work, too.
 
Jan 19, 2018 at 9:26 PM Post #7,274 of 15,989
Hey guys! Thank you very much! :) Right now, the way I have it set up is to a breadboard. However, SMD components are on the way, and will hopefully be able to produce a similar response (let's hope).

I only use a single BA driver and at the moment I think the balance between bass and mids are pretty good. A high pass might also lower the overall volume, but maybe not significantly? However, it is a good suggestion and I will try it!

If it is possible, I want the headphones to be driven from a phone without a problem. Right now it can for the most part, but don't want to lower it any further.

I will get back when I get the rest of the components to show the progress. Thanks for the support! I really appreciate it and needed it to be able to work even better.

/Anders
 

Users who are viewing this thread

Back
Top