Home-Made IEMs
Jan 4, 2019 at 11:51 AM Post #7,892 of 15,989
in most of KA's datasheets especially for pro-audio drivers there's provided inductance
well, i need ED29689 following data

DC Resistance of the driver[ohms]
Voice coil Inductance of the driver [H]

for basic Zobel Circuit Layer 1


Second parallel layer is Resonance peak equalizaion in impedance

for that i need graph

here it goes, as per my theory
Zobel circuit has flattened the high freq impedance rise giving me an equalised R1.

the peak, mostly in bass region or mid region, called as resonance peak, has a top point which is R1+R2(Zmax).

from there we can find value of R2

Then we mark an impedance value |Z*| at a frequency f* higher than fo. As a rule we choose impedance values somewhere in the middle between R1 and |Z|max.

R3 = [R1(R1+R2)]/R2

we got our second circuit resistor value

We may need to zoom in a narrow frequency range around the resonance frequency. In this way we will be able to get a good estimate of fo, |Z|max, f* and |Z*|.

x= f*/fo (Variable 1 for Q factor)
y=|Z*|/R1 (Variable 2 for Q factor)
ymax=|Z|max / R 1(Variable 3 for Q factor)
w=2 pie fo(assistive variable, i forgot the exact name)

Q= {x/(x^2)-1} * [root{ (ymax^2) - (y^2)/(y^2)-1}]

L=R2/Qw
C=Q/R2w

Lreq=(R1^2)*C
Creq=L/(R1^2)



We need R3 then Lreq and then Creq in series attached parallel to Basic Zobel

For flat Impedance, capacitance graph plus making single BA more extended on both side of response chart and with more stability.

just making impedance curve flatter wont do
capacitance graph and peak resonance graph of single BA matters

Technically fixing the spout and Internal BA resonance with Impedance rise. This will make the BA less sensitive but from my point, it will draw less power, be always linear no matter what we do plus protect from random signal shots

Something like your Flat impedance but in dual layer

1. Flattened whole graph without getting the driver customised
2. More safe to powered burst
3. More heat division
4. Very low distortion if we dont add the component distortion, but the distortion should still remain lower compared to stock driver


now problems are
1. knowles gives scarce data for all the things i am trying to do
2. Inductor coil saturation is still an issue

@Slater

then i add small dynamat in the ending like you do bro

i mean, laser cut dynamat....lol to a BA



@piotrus-g

after all this, i will add Edgar Mid range horn with 2mmx2mm mouth(straight) or 4mm libby horn with 3mmx3mm mouth(folded in curve, so stopping pressure is low)

Edgar midrange horn is good for fixing conventional horn resonance

Libby horn will raise 2kHZ to 5kHz by 10dB and all high freq by 5dB, but green damper will bring my mids back while retaining extra air in treble region


all these method can reduce ear strain by magnitude and driver volume would be compensated back while having all the benefits of zobel and my circuit
 
Last edited:
Jan 4, 2019 at 2:00 PM Post #7,893 of 15,989
Hi @Furco, @Bassiklee,

For my previous question, here's a photo of my universal.

I tend to sand and polish in this area due to resin overflow and when I do sand and polish, the MMCX also gets sanded down.

Also, for spot curing, do you guys get that tacky residue? I get them even when I only adhere my sound tube to the nozzle.

upload_2019-1-5_2-58-58.png
 
Jan 4, 2019 at 2:24 PM Post #7,894 of 15,989
I use a different MMCX connector, never had a issue. As for the inhibition layer, yes that will happen. Alcohol will wipe it off. I use resins inside the shell that don't produce that layer. Everything on the outside of my shells is one of three finish lacquers, so they don't get tacky if fully cured. I'm still looking for the perfect lacquer. I like the Egger, because it doesn't yellow. I like the Dreve Lak3, because it flows better. I've tried the Pro3Dure lacquer as well. It behaves more like Lak3, including the yellowing. Really only an issue on clear shells.
 
Jan 4, 2019 at 8:33 PM Post #7,895 of 15,989
I need someone who has a universal design in .stl format

I am tired using aliexpress semi universal shellz plus they come for 40$ and feel very cheap.

It would be better for me to print one for myself

But dont know how to design according to ears.
 
Jan 4, 2019 at 11:48 PM Post #7,896 of 15,989
I use a different MMCX connector, never had a issue. As for the inhibition layer, yes that will happen. Alcohol will wipe it off. I use resins inside the shell that don't produce that layer. Everything on the outside of my shells is one of three finish lacquers, so they don't get tacky if fully cured. I'm still looking for the perfect lacquer. I like the Egger, because it doesn't yellow. I like the Dreve Lak3, because it flows better. I've tried the Pro3Dure lacquer as well. It behaves more like Lak3, including the yellowing. Really only an issue on clear shells.

May I see the MMCX you use? Oh, any recommendations for resins that has no inhibitation? I have problems sourcing lacquers here :frowning2:
 
Jan 4, 2019 at 11:49 PM Post #7,897 of 15,989
I need someone who has a universal design in .stl format

I am tired using aliexpress semi universal shellz plus they come for 40$ and feel very cheap.

It would be better for me to print one for myself

But dont know how to design according to ears.

Why not use other universal shells as your master then cast one yourself?
 
Jan 5, 2019 at 12:00 AM Post #7,898 of 15,989
Why not use other universal shells as your master then cast one yourself?


Shell
these are the shell i am using bro
so can you tell me how to work on this. I mean to say, i dont know how to edit 3D files like .STL(PITA)
how do i scan these on budget, well i was in financial crisis previous year and i am still recovering so cannot buy a 3D scanner for time being

Thanks for the reply though:smile_phones:
 
Jan 5, 2019 at 12:02 AM Post #7,899 of 15,989
These are what I use. As far as resins with no inhibition layer, look at stuff marketed at arts&crafts, or fishing/surfing. Again, I don't use these on the outside, so skin contact isn't an issue.


I'm not sure why it's not letting me add a picture. The thumbnail was there. I swear. PM me, and I'll send you a picture of what I use.
 

Attachments

  • 54B18585-9EF7-474C-B50D-3DCB271257C2.png
    54B18585-9EF7-474C-B50D-3DCB271257C2.png
    1.3 MB · Views: 0
Last edited:
Jan 5, 2019 at 3:56 AM Post #7,901 of 15,989
Alright here's the second one more fancy, more neutral (still quite fun! but more on the sub-bass side)
It's 4 way design for those who don't fear soldering since it's more complicated.



My recommendation ... don't use two CI-22955 if you plan on using only one TWFK
:D

Can we replace the "CI with R3" resistor with ED29689 and keep everything same

i mean damper and R3 value
You know, i have problems with tube measurement, so i am not changing that
 
Jan 5, 2019 at 4:58 AM Post #7,902 of 15,989
Do look up photogrammetry. You just need a smartphone, pc and some free software. I think a even normal consumer 3D scanner can’t capture such fine details.

Shell
these are the shell i am using bro
so can you tell me how to work on this. I mean to say, i dont know how to edit 3D files like .STL(PITA)
how do i scan these on budget, well i was in financial crisis previous year and i am still recovering so cannot buy a 3D scanner for time being

Thanks for the reply though:smile_phones:
 
Jan 5, 2019 at 8:50 AM Post #7,903 of 15,989
Do look up photogrammetry. You just need a smartphone, pc and some free software. I think a even normal consumer 3D scanner can’t capture such fine details.
Used scann3d....ask for money
Clicking photos, dam this is hard

And even if the accuracy is to the point, the issue is the bottom of the shape

It produce some nasty bottom, which is hard to clean

I have to make a rotating jig and light box for 3D scanning.....

Taking to much time also

Any other way, easier than this

Or I have to stick with this


By the way, thanks a lot
 
Jan 5, 2019 at 12:01 PM Post #7,904 of 15,989
I think mobile app can’t get the precision that you want.

There is this tutorial about a paid Pc/MAC software. She demonstrate scanning a rock in a upright position, then flip the rock around and scan again. She then merge the 2 sets of photo to form one seamless 3D model. I believe you can find free softwares that can do it, or somehow get your hands on that paid software.

If I’m not wrong, this YouTube channel call prusa upload a updated process for photogrammetry using a free software. Have not watch the video yet.

If you don’t want to do too many hard work. Try approach your local 3D printing and scanning community and ask them for advise or even do for you.

This last method really depends on your luck. Try Google ear impression 3D scan at your neighbourhood. You might find some businesses that actually make your ear impression for the usual price and then 3D scan them for a small fee. You can try negotiate with them to do just the 3D scanning part. You can either process your shell a bit then bring them for scanning or make a negative mold followed by a positive mold then bring it for scanning.

This topic is going quite in-depth about 3D. I only have very basic knowledge about 3D printing, scanning and modelling. Couldn’t help very much.


Used scann3d....ask for money
Clicking photos, dam this is hard

And even if the accuracy is to the point, the issue is the bottom of the shape

It produce some nasty bottom, which is hard to clean

I have to make a rotating jig and light box for 3D scanning.....

Taking to much time also

Any other way, easier than this

Or I have to stick with this


By the way, thanks a lot
 
Jan 6, 2019 at 9:07 AM Post #7,905 of 15,989
My first project in progress: an Etymotic style shape + single RAB-32257.

I've printed a 3D shape created by some user, the original file in here, now I'll finish sanding and paint it to try looks a bit more refined. Also, I'm just trying to figure out how is the best way to seal/filling up the armature in this cylinder, will something like that with blue tack works fine? If someone has a better idea would be helpful too!

get that 3D design and do a stainless steel CNC with Stainless steel tube and circular tubing
man that will look dope

BluTack works fine by the way





by the way,

does anybody have impedance graph of any of the stated driver

Knowles ED-29689
Knowles GQ-30783
Knowles CI-22955

Hmm
i dont have measuring rig, and no multimeter in hand
plus i had 4 ED, which are already in iem, which i dont wanna take out
hmmmmm

So I went to the CANJAM convention in Denver the other weekend. Thought I might share some comments about looking at CIEMs there, with the caveat that I’m still learning and am now even more aware of how little I know. Perhaps would be of interest, especially to beginners in my same, or similar situation. Sorry about the length and rather meandering narrative.

  • I spoke with a bunch of people, and listened to A LOT of in ear monitors from the likes of 64 Audio, JH (Jerry Harvey) Audio, Noble, Unique Melody, and quite a few others (Ultimate Ears was not there). All had most of their CIEM models there, but with universal outlets so they could be sampled with silicone/foam tips.

  • Some observations and conclusions first: Strangely, I was kind of wondering if I’d be discouraged by how much better the main line CIEM manufacturers products might sound compared to what I’ve made in my basement. In the end, I felt just the opposite. I decided that I’m pretty satisfied with some of what I’ve already built and felt gratified that they, at least to my ears, they compared favorably with some rather pricey offerings. For example, the “Low Pass CI / Bass CI / Mid-High TWFK” recipe I built as suggested earlier in this thread (~p265?) by piotrus-g (thank you!). Also decided that the set I made using the Knowles GV quad drivers still sounds very nice and much like many of the 4 driver setups I sampled. BTW, of the above, the first sounds a bit more “speaker-like” in the lows and the GV more “clinical” and tight. I like them both.

  • Of the big name brands I sampled, ALL had sets I thought were really good. Clearly, they know what they’re doing. On the other hand, I was pretty unimpressed with offerings from several lesser known, or no-name companies. At one such booth the rep prodded me several times to tell him honestly what I thought. I had to say I couldn’t find many positives other than lower price, - apologetically suggested he may want to spend some time listening to the competition and then come back and compare... (right, like I’m really an expert?)

  • I quickly got the notion that in that crowd, BEATS seem to really be discounted as being “headphones for dummies” and for clueless people who have no idea as to what they should be hearing. However I found it interesting that quite a number of companies had sets they described as having a “fun” profile which seemed really to be just another “V” shaped profile with bass emphasis…..um.. like BEATS? (ok, I admit I kind of like them too)

  • I know that I have some “age/abuse” hearing reduction, however I was really pleased that I could still hear and identify profile differences and identify quite a few of the claimed characteristics of the ones I sampled.

  • In general, more drivers per ear did not necessarily make me like a set any more. In fact I found myself pretty indifferent about some of the high dollar 9, 10, 12 driver models. I actually sampled several pretty good sounding doubles, but after 3 or 4 (some cases 6) drivers the incremental quality of sound seemed to mostly disappear. Admittedly, this could be due to my current level of hearing capabilities. If I was a true audiophile with young ears, I may have felt differently. For now, I’ll stay away from attempting more than 4 or possibly 5 drivers for fear of wasting a lot of time screwing up what I don’t fully understand.

  • I found little correlation between a manufacturer’s frequency response graphs and what sounded good to me. 2 almost identical graphs might sound completely different to my ears. Yeah, I do understand how that can be…

  • I only listened to a handful of dynamic driver based earphones, but there was one particular single dynamic driver setup that I really liked from Periodic Audio (Beryllium model, not so much the Titanium).

  • My mind kept going back to pondering how important it really is to be so precise. In the end, the sum of everything that the drivers create must pass through a single, convoluted, purely mechanical path before being sensed and sent to the brain to be interpreted as sound. I.e. eardrum>hammer>anvil>stirrup>oval window membrane, then once delivered to the inner ear, “wiggle” many thousands of hair cells connected to nerves, creating signals send to the brain to be interpreted as sound. Crazy.

  • In the end, the more I listened, the more selective I became. I had started this adventure to make some CIEMS for stage monitor use. I first made several 2 driver sets for myself that live in my gig bags. Upon returning home, I recalled hearing some pretty great sounding 2 drivers at the show and decided that the preconfigured Knowles “HE” dual driver set I first made was nice but had room for improvement. Also decided the other 2 driver setup I made for myself, reconfigured once already, while perfectly functional, is no longer satisfying at all. I cut both of them apart to modify/reconfigure. BTW, the latter one is now a really nice sounding 2 driver setup, based yet again on another Piotrus-g suggestion for a starting point. It’s now w 2way CI 30120, 15 ohm, yellow, 1.5 mm id tube / Bellsing 30095 “WBFK”, 1.5uf, white, 2mm id tube. A budget $ keeper to my ears.

  • To the other new learners out there. Read this thread, but be aware it is dangerous. You could get hopelessly sucked in and lose a lot of sleep. On the other hand, if you pay attention to processes and recipes from the veterans such as Shiloh and Piotrus (and others), pretty cool results are possible. Thanks so much!
Periodic Audio Be....you liked it

Hmmmm...cool

But please try Dita. These set blow me away
I talked to Dita and they told me everything they did with driver and how the pressure regulate inside the chamber

Other is Sony EX1000 with spinfit CP240



















well for all new comers into this DIY
1.FURCO book pdf download

https://docs.google.com/document/ex...Os1CA:1545412179251&includes_info_params=true

(If Headfi redirect comes and take time, go to this link....it will take some time, but you will get pdf download.....Its a redirect download link)

2.Deeper Crossover Understanding

3. Notes about tubing and damper effect(I dont have one for now, go read Knowles and sonion guide)
 
Last edited:

Users who are viewing this thread

Back
Top