New Posts  All Forums:Forum Nav:

Home-Made IEMs - Page 126

post #1876 of 2081

Interestingly, after I moved the drivers back to an older housing, it seems to have dissipated. I think the rubber plug i printed to help seal the channel was flapping/covering the outlet and causing the distortion.

 

I might just make the channel to fit the nozzle in a single material next week instead of having multiple materials (rigid and rubber) in the same print.

 

False alarm!

post #1877 of 2081

Can you post pics of new shells with different material. What printer are you using?

post #1878 of 2081

Vector1 I didn't notice, but you can print rubber as well?

post #1879 of 2081
Quote:
Originally Posted by tranhieu View Post

Vector1 I didn't notice, but you can print rubber as well?
And mixed materials in the SAME PRINT!?
Wow, Didn't know that could be done, cool.
post #1880 of 2081

Google Objet 260

 

It's the cheapest of the high end polymeric material printers, and I can print in either veroclear (transparent rigid plastic) or tangoblack plus (rubber) all in the same print. I am however printing the rubber plugs separately and squeezing them in for better control.

 

The advantage is I can also print a mixture of both materials, which yields an in-between material of slightly more rigid rubber, or more flexible plastic.

 

I'll try to get more photos of the prints next week with more improvements.

 

I'm waiting on my WBFK delivery before I continue testing so it's a break for the weekend!

post #1881 of 2081

I have resolved a few comfort issues I was having with my CIEM's and I can pass along the following advice.

 

First, don't make the canal section to long, you only want them to go slightly past the 2nd bend. Beyond the 2nd bend, the ear canal is very delicate and easy to traumatize with even small amounts of pressure or movement and in addition, going deeper than that wont produce a better sound.

 

Second, the cable connection should come out of the shell at a 45° angle and have plenty of clearance. Most peoples outer ears are not flush against their heads and the surface of the CIEM will be at an angle with the front pointed into the head, if the cable connection is parallel with the surface of the CIEM, than the cable will dig into your temple and give you a headache like you wouldn't believe. This can also prevent the CIEM from sitting properly in the ear when the cable is plugged in resulting in a poor seal and an uncomfortable fit.

post #1882 of 2081
Quote:
Originally Posted by CMOS1138 View Post
 

I have resolved a few comfort issues I was having with my CIEM's and I can pass along the following advice.

 

First, don't make the canal section to long, you only want them to go slightly past the 2nd bend. Beyond the 2nd bend, the ear canal is very delicate and easy to traumatize with even small amounts of pressure or movement and in addition, going deeper than that wont produce a better sound.

 

Second, the cable connection should come out of the shell at a 45° angle and have plenty of clearance. Most peoples outer ears are not flush against their heads and the surface of the CIEM will be at an angle with the front pointed into the head, if the cable connection is parallel with the surface of the CIEM, than the cable will dig into your temple and give you a headache like you wouldn't believe. This can also prevent the CIEM from sitting properly in the ear when the cable is plugged in resulting in a poor seal and an uncomfortable fit.

 

 

If you had a chance to look at my designs, should I shift my mmcx connector outwards/rotate it outwards to avoid this? I have tried to follow the shure 846 as a design idea, but I'm not sure if my slightly smaller one will encounter the same problem till I test it next week.

 

Also, is it right to say that increasing the acoustic volume inside the earphone by extending length of tube/increasing cross sectional area can help tune the earphone? I can increase the volume of the cavity that couples all the input nozzles to the output nozzles, would that help give a peak at the 1kHz to 4kHz range?

 

However, would a high pass on the CI make any sense? I'm getting a pretty strong hump around the 250Hz to 500Hz range, which tends to make instruments sound just that little bit too loud. The obvious answer would be to just damp down the low end, but then I lose my subbass response. I have tested a highpass on DTEC / low pass on CI before but not reversed it yet (running CI on resistor only so far)

 

Alternatively, boosting up the 1k-4kHz range for clearer vocals acoustically would be good if I had any idea how to do it.

 

Currently, the CI on resistor, DTEC on lowpass setup without WBFK sounds pretty good, but lacking for vocals. I have EQed a 1-4dB boost around from 1kHz onwards to 16kHz (parametric equaliser on sansa clip zip/media monkey electri-q plugin) and it pretty much what I want from the earphones! I expect the addition of the WBFK on monday with a small cap for highpass ought to help reduce the need for EQ to some extent by putting in my high frequencies!

 

In hindsight, my ideal sound signature might not necessarily be a truly flat response, but more towards a sub-bass boosted/ gentle V response because I love having a bit of sub 100Hz rumble and detailed, forward vocals.Currently, hearing all the instruments so clearly is kinda detracting from vocals IMO but YMMV.

post #1883 of 2081
Quote:
Originally Posted by vector1 View Post
 

 

If you had a chance to look at my designs, should I shift my mmcx connector outwards/rotate it outwards to avoid this? I have tried to follow the shure 846 as a design idea, but I'm not sure if my slightly smaller one will encounter the same problem till I test it next week.

 

...

 

In hindsight, my ideal sound signature might not necessarily be a truly flat response, but more towards a sub-bass boosted/ gentle V response because I love having a bit of sub 100Hz rumble and detailed, forward vocals.Currently, hearing all the instruments so clearly is kinda detracting from vocals IMO but YMMV.

 

 

The advantage of your design is that the connection point is shifted back away from the front of the ear (I am referring to the front as the part of the ear closest to your face,) that should give you enough room to move the cable around and not have it push into the side of your head but it will depend on the shape of your ear as well as the shape of your cable. 

 

 

I can't offer much help in sound tuning as I am working by trial and error as well as the helpful suggestions of piotrus-g and others on this forum. I do agree that a flat sound is not very desirable, My GQ design sounds terribly boring and lifeless because the sound is so flat using white dampers, I am eagerly waiting for my damper order to arrive, right now even my SR driver design sounds much better (knowles SR6438NWS-000 driver with a short sound tube and green dampers, very nice sound considering the low price.)

post #1884 of 2081
Quote:
Originally Posted by CMOS1138 View Post
 

 

 

The advantage of your design is that the connection point is shifted back away from the front of the ear (I am referring to the front as the part of the ear closest to your face,) that should give you enough room to move the cable around and not have it push into the side of your head but it will depend on the shape of your ear as well as the shape of your cable. 

 

 

I can't offer much help in sound tuning as I am working by trial and error as well as the helpful suggestions of piotrus-g and others on this forum. I do agree that a flat sound is not very desirable, My GQ design sounds terribly boring and lifeless because the sound is so flat using white dampers, I am eagerly waiting for my damper order to arrive, right now even my SR driver design sounds much better (knowles SR6438NWS-000 driver with a short sound tube and green dampers, very nice sound considering the low price.)

 

Thanks, I think I understand what you mean, if you're designing fixed cables into CIEMs. I guess my style will allow for adjustment, considering that the wires I have now also have memory plastic around the ear to hold a rigid shape around the ear, so I don't expect any problems with that.

 

For all that talk about flat response, I think it's more about what sounds good to us :).

post #1885 of 2081

I've just soldered everything together with the WBFKs in place!

 

I'm running a 15ohm/100uF lowpass filter on the DTECs, high pass of ~15ohm (still tweaking) and 100uF on the CIs, and WBFK on full range.

 

It seems to work pretty well for me so far, but it could stand to use a slight boost to the 1kHz to 4kHz range, of about 1-3dB from my EQing to taste.

 

The bass is pretty impressive if I say so myself, and the mids are current not too recessed/forward, but I prefer them brighter. Highs are all there and not too fatiguing with the WBFK giving the air/sparkle at 4kHz+ range.

 

Once I settle on a suitable highpass for the CIs, I'll be soldering everything with 0.2mm wire and ordering the actual SMD caps/resistors!

 

Does anyone have more suggestions on boosting the 1kHz - 4kHz range acoustically or without adding too many components?

post #1886 of 2081
Quote:
Originally Posted by vector1 View Post

I've just soldered everything together with the WBFKs in place!

I'm running a 15ohm/100uF lowpass filter on the DTECs, high pass of ~15ohm (still tweaking) and 100uF on the CIs, and WBFK on full range.

It seems to work pretty well for me so far, but it could stand to use a slight boost to the 1kHz to 4kHz range, of about 1-3dB from my EQing to taste.

The bass is pretty impressive if I say so myself, and the mids are current not too recessed/forward, but I prefer them brighter. Highs are all there and not too fatiguing with the WBFK giving the air/sparkle at 4kHz+ range.

Once I settle on a suitable highpass for the CIs, I'll be soldering everything with 0.2mm wire and ordering the actual SMD caps/resistors!

Does anyone have more suggestions on boosting the 1kHz - 4kHz range acoustically or without adding too many components?
Yes, loose the high pass on ci and use resistor instead.
Else you may try adding resistor after your low pass or connect your wbfk by center tap of CI
post #1887 of 2081
Quote:
Originally Posted by piotrus-g View Post



Yes, loose the high pass on ci and use resistor instead.
Else you may try adding resistor after your low pass or connect your wbfk by center tap of CI

 

I tried that first. The resistor doesn't give anything to the sound, only taking away the high response as I tune it up.

 

I could run CI/WBFK without the highpass on full range, but having the high pass helps me to tune the balance between the different drivers as I can balance left right channels using the resistors.

 

TBH, very little difference between highpass and full range for CI. I just need to balance out the bass since the resistance changes when I remove the high pass circuit.

 

Edit: I'm now testing CI/WBFK on full range, DTEC on lowpass (tuning for proper bass, maybe around 7.5-10ohms with 100uF cap.). This still sounds pretty good.

 

Whatever I change, I'm only going to get very minor adjustments since I already know what sound I'm going for. So far, the mids are a just little too laid back for my taste, but forward enough in most songs.

 

The problem is, how do I maintain the subbass, while boosting mids? I can tune bass down to emphasis mids more, but then I lose that kick from the bass. I think the bloat into the mids from the DTECs is the main problem here.


Edited by vector1 - 6/30/14 at 8:35am
post #1888 of 2081

In hindsight, I already know the solution. I just put a 2nd order lowpass filter on the DTEC. That has nicely tamed the bass bloat at the lower mid range!

 

Using 2x 100uF filters, 2x 5ohm resistors, gives me all the kick around 100Hz, without bleeding into the mid range.

post #1889 of 2081

So in conclusion, the problem is not a lack of mid, but the bloat from the DTEC on bass to mids that made it feel lacking. A 2nd order lowpass on the DTEC has helped to tame that very nicely, and even without any filtering on WBFK or CI, the sound is already all that I want from this!

 

Tuning can be achieved by putting a 10ohm variable resistor on the first resistor position, and that controls the bass relative to the rest. I'm hovering around 7.5ohms for extra clarity in the mids while losing a little bass. The other is a fixed resistor 5ohm.

 

For those electrically inclined, here are some simulations that confirm my tests.

 

This is the electrical diagram used to simulate in orcad cadance (pspice)

 

Green = 2nd order lowpass, 10 ohm first,followed by 5ohm

Purple = 2nd order lowpass, 5ohm first followed by 10 ohm

Red = 1st order 5ohm lowpass

Violet = 2nd order, 5ohm x2.

 

 

 

 

For the results

 

Red, the first order is quite inadequate to deal with mid-range bass, with too slow roll off and causing lots of bloat.

 

Violet, 2nd order with 2 5ohm resistors seems to have much better roll-off at the expense of some signal.

 

Purple and green are 2nd order with the 10ohm resistor varying in position. While band-pass range remains the same, the roll-off is faster/ cut off frequency is lower if 10ohm is first rather than 5ohm.

 

Edit: wrong snip


Edited by vector1 - 6/30/14 at 9:47am
post #1890 of 2081

For some reason, putting a cap on the WBFK (1uF) makes the sound much "fuller". It seems to interact with the rest of the circuit strangely, but is minimised behind a capacitor.

 

Current setup, 2nd order lowpass (5ohmx2, 100uFx2) on DTEC, CI full range, 1uF cap on WBFK. Each driver is in parallel w.r.t the +ve and -ve wires.

New Posts  All Forums:Forum Nav:
  Return Home