Latest Thread Images

Home-Made IEMs
Jul 11, 2019 at 2:04 AM Post #9,556 of 15,996
dhruvmeena96 said:
For CI, ED and SWFK

I think I might use SWFK32254(so dampers are unneeded)
For CI and ED

I might see

I was thinking of using HE for a moment though
Knowles HE and SWFK



MASM legend update for perfect never hot treble

GQ grey damper swapped with brown.

More smoother and flatter treble
Click to expand...

I've never used a damper for the 31736 either actually. But I think the ED29689 will definitely need a damper.

Oh, I just thought of something. You could try this setup for CI+ED+SWFK:

CI22955, C=6.8uf R=20ohms, white filter [1mm x 10mm PVC]
ED29689, C=2.2uf R=10ohms (need to test this), white filter [1mm x 10mm PVC]
SWFK, C=0.13uf

Should get you a pretty flat response.
 
Jul 11, 2019 at 10:54 AM Post #9,558 of 15,996
upload_2019-7-11_16-53-56.png

Tubeless custom iem from FIR audio using DD, BA and electrostatic.
 
Jul 11, 2019 at 10:55 AM Post #9,559 of 15,996
Xymordos said:
I've never used a damper for the 31736 either actually. But I think the ED29689 will definitely need a damper.

Oh, I just thought of something. You could try this setup for CI+ED+SWFK:

CI22955, C=6.8uf R=20ohms, white filter [1mm x 10mm PVC]
ED29689, C=2.2uf R=10ohms (need to test this), white filter [1mm x 10mm PVC]
SWFK, C=0.13uf

Should get you a pretty flat response.
Click to expand...
White dampers are weak for ED29689, from my old experience.


Final 2 mk II(zero damper edition)
The original Final 2 has its own advantages, this has its own


Knowles FED30048
Knowles FED30027

Same tube configuration as Final 2 but with a twist

FED30048 (2mm ID/ 7.5mm~8mm length tube, No damper) + 25ohm resistor

Knowles FED30027( 2mm ID/ 9.5mm~10mm length tube, No damper)


This is full parallel mode



I will be introducing a crossover edition of Final 2, which will have 48ohm impedance.

And FED30048 zobel will also be introduced

I am working on that


And a ferrofluid driver zobel may be the next big thing as zobel equalizes voltage(impedance spikes), this may lead to less heat generation and more stability.

@eunice @Xymordos
 
Last edited:
Jul 11, 2019 at 7:25 PM Post #9,563 of 15,996
In my opinion, the best summary is the fact that big companies (and CF can afford it) can introduce their own custom driver versions, so Hodvtec is not stock Hodvtec etc.... that's why you can't always model on tubes and driver appearance. Because there is something completely different inside metal housing.
 
Jul 11, 2019 at 10:33 PM Post #9,564 of 15,996
SupremusDoofus said:
What do you guys think about this : https://www.reddit.com/r/DIEMs/comments/cbng8t/a_4_page_paper_of_how_sound_tubes_affect_iems_in/
Click to expand...
Well 1 point he missed

The closer the damper is to the spout, it dampens the spout resonance or peaks more(rolls off less). This area is has highest escape velocity and pressure so lack of dynamics or slowdown effect is least.
The damper in center of tube dampens {spout + 1/2 tube} resonance but rolls of more(peaks are not damped properly or damped too much(depend how you design). Slow down is noticeable(dont confuse with phase and all, it is the pressure sense of ears).

On tube ends it dampens {Tube + earcanal effective reflection back} and slow down is noticeable. It has highest roll off. Well why didn't I mention spout here, as the effective resonance of spout is changed by the whole tube after which it is damped..


Well, you can go with his guide, its easier to follow

I am still reading it
 
Last edited:
Jul 12, 2019 at 12:27 PM Post #9,565 of 15,996
Xymordos said:
I've never used a damper for the 31736 either actually. But I think the ED29689 will definitely need a damper.

Oh, I just thought of something. You could try this setup for CI+ED+SWFK:

CI22955, C=6.8uf R=20ohms, white filter [1mm x 10mm PVC]
ED29689, C=2.2uf R=10ohms (need to test this), white filter [1mm x 10mm PVC]
SWFK, C=0.13uf

Should get you a pretty flat response.
Click to expand...

Did it, pretty neutral(I mean its very neutral but extended)

But white damper kinda gives it glare type of mids(to overpowering upper mids). Its like over sharp. I used conventional green and it was kinda fixed.

I was working with DTEC-30008. Nice driver after zobel. Zobelling it makes the whole impedance come down to ~31.5 ohms aggregate and the speed boost is significant . Other driver is DTEC-30265(not using for this build)
Now the question is I can cross them at 72Hz easily with the classic crossover Rc= 100ohm, Cc= 22uF.
Now the question is, how to cross the RAB32063(it will have the zobel). Capacitor selection is hard....and seriously I don't know how to really cross stuff with RC circuits.
I can also go full range on both, but I dont think its worth it for my knowledge as I want to learn some crossover stuff.

And one more thing, if we take a physical non ideal full range driver and take out the lower frequency, the transient increases significantly. The best point is 100Hz and bellow(coming from speaker building knowledge).

But I will post a full range model with HODVTEC, RAB and a SWFK(update)
 
Last edited:
Jul 13, 2019 at 7:51 AM Post #9,566 of 15,996
dhruvmeena96 said:
Damm that looks nice....

Is that the place you kept for Apex module(with white type paper)

And is the tube from dynamic driver a pressure release front or back
Click to expand...
back vent. if you look carefully you can see the front of the driver shooting inside the entire shell. it's probably low passed or/and has a very small slit at the nozzle or along the apex tubing just for it to act has a low pass. I have no idea though
 
Last edited:
Jul 13, 2019 at 10:22 AM Post #9,569 of 15,996
dhruvmeena96 said:
@Xymordos @piotrus-g @Kulgrinda @Choy Wei De
Can you guys fit a 12.5mm x 12.5mm x 6mm componemet inside a shell with HODTEC, RAB and SWFK inside.

Its a Bourns inductor of 390uH value
Click to expand...
look at Coilcraft. They probably have more values around 330 but there might be a small enough one at 390uH
also you need to take into consideration the DCR since it can play an ever bigger role than 330 vs 390 depending on what you try too achieve
 
Last edited:
Jul 13, 2019 at 10:56 AM Post #9,570 of 15,996
ForceMajeure said:
look at Coilcraft. They probably have more values around 330 but there might be a small enough one at 390uH
also you need to take into consideration the DCR since it can play an ever bigger role than 330 vs 390 depending on what you try too achieve
Click to expand...
The Bourns version is 0.7ohms

https://uk.farnell.com/bourns/srr1260-391k/inductor-shielded-390uh-1a-10/dp/2531339

I will check coilcraft also

According to some known people(cannot name here, the thread and forum), LC crossover seems more better due to low energy and resistor parasitic losses.

My idea is 0.5uF on SWFK 32254 with 390uH(0.39mH) on HODVTEC high impedance 31516 and RAB32063 parallel with their own zobel(mixed zobel and not discreet due to space limitations).

And best part is that Bourns inductor is sealed, so no more EMF variable to play with.


Design is simple

Subwoofer, Mids, upper mids
HODVTEC 31516(dual vents closed)

Zobel
Rz= 214ohms
Cz= 0.220uF

Tubing
1.0mm ID/11mm length

Damper
Yellow damper after 4mm from spout



RAB 32063

Zobel
Rz= 38ohms
Cz= 5uF or nearby

Tubing
1.2mm ID/7mm length

Damper
Brown damper after 3mm from spout


Now fusing these zobel together and making these two driver in parallel


Adding a 0.39mH inductor in series to both of these parallel driver.

Tweeter above 10kHz
SWFK 32254 with 5uF or 4.7uF capacitor


Voila

This is the setup


But if we can fit the inductor in shell, it will be amazing
 
Last edited:
You must log in or register to reply here.

Users who are viewing this thread

Total: 20 (members: 3, guests: 17)
Back
Top