[Xymordos]

That's what the shop claims...but then from the photos you can't see the WBFK drivers at all! I think it is just a driver the shop itself glued together to sell.
 
 

  Quote:
 
Around 2,1-2,2kHz

 
How did you get that calculation? I use the 1/2piRC formula with R=3.65 and C=2.2uf and that doesn't work

 

[MuZo2]

Doesn't look like Sony drivers.

 

[piotrus-g]

  That's what the shop claims...but then from the photos you can't see the WBFK drivers at all! I think it is just a driver the shop itself glued together to sell.
 
 
 
How did you get that calculation? I use the 1/2piRC formula with R=3.65 and C=2.2uf and that doesn't work

Impedance of this driver at around 2,2kHz is ~32 Ohm - you do the math
 
You're using Rdc for your calculations.

 

[wjp007]

Had some time and started experimenting with different tube lengths and tube styles with a GQ driver.  My goal was to try and improve the high frequency extension.  The results are promising with my crude measurement setup. The green trace that was from a stepped design shows considerable extension above 10K.  I'm going to try cone shaped horn designs next.
 
I tried the following:
1) White - Stock GQ driver no tube
2) Green - Stepped tube 1.3mm ID 7mm long followed by 2.4mm ID 12mm long
3) Blue - Single tube 1.3mm ID 25mm long
4) Purple - Stepped 1.3mm ID 12mm long followed by 2.4mm ID 12mm  long
5) Yellow - Single tube 2.4mm ID 25mm long
6) Red - Single tube 2.4mm ID 18mm long
 

 
High Frequency Close Up

 

[CMOS1138]

I am curious how people are able to control the tube length so precisely in a custom shell? when I made my GQ + DTEC design I simply pulled the driver as far down the canal section as it would go and then cut the extra tubing sticking out of the end of the shell. I know that isn't very accurate but everyone's ears are shaped differently and will require different tubing lengths assuming you use as little as possible for the high driver.
 
Speaking of that particular design, I am amazed at the rich sound I am getting out of it, it is so natural sounding with nothing artificial or distorted. Letting DETEC influence the mids a little bit by using a green filter is the right choice for my personal enjoyment.

 

[Xymordos]

  Impedance of this driver at around 2,2kHz is ~32 Ohm - you do the math
 
You're using Rdc for your calculations.

Ah, so that's what you use. Thanks! 

 

[vector1]

Looking at all the positive reviews of earphones using the ED/WBFK combo or GQ, I'm starting to look forward to trying it out myself soon! Great frequency graphs and very nice comparison! I'll be sure to try to simulate some stacked tube using the 3D printed channel.
 
Also, I'm wondering if it's worth while to try something like the shure 846 microchannels that make up their "lowpass" filter... I could print something like that out pretty easily, or a nice winding channel to increase length.

 

[Xymordos]

Using the WBFK is pure annoyance -.- Probably will never want to use it again...

 

[vector1]

  Using the WBFK is pure annoyance -.- Probably will never want to use it again...

 
First time I used it I destroyed the first two. That made me rethink my soldering technique and now the ones I have have survived testing for a couple of weeks already.
 
It's not that much more difficult to use, just need a bit more care.

 

[Xymordos]

I never destroyed any, but getting rid of the dips caused by using that driver takes a lot of patience...I can't even tell if it brings out extra details in the high frequency either LOL

 

[piotrus-g]

Well I'm playing around with WBFK right now and it does no good with combination with Sonion 2389 which means it won't work with ED either (those drivers are practically the same). I'd say if you plan on using WBFK get different midrange. BK or EF maybe or 2600. I think working out proper crossover may not be worth the hassle for DIY unless you really have to use those drivers or you know.. have too much spare time
 
I mean I was able to get quite good response with 2389 - but with shifted all peaks missing points needed to compensate ear resonances.

 

[vector1]

What do you all think of the GQ graphs posted above then? They look pretty decent to me, especially if used with a relatively long/stepped tube. I'll experiment with printing a winding channel rather than a straight channel to increase the length in a small area.
 
Is it okay to have that ~6db drop past 10kHz, or does it necessarily have to be straight? I don't know if that'll make the high end too hot. Perhaps doubling down on the WBFKs might work?

 

[Xymordos]

WBFK can be fitted properly I think by changing the distance the driver is from your ear But matching this between left and right ear is difficult

 

[vector1]

  WBFK can be fitted properly I think by changing the distance the driver is from your ear But matching this between left and right ear is difficult

 
Perhaps I never have this problem because my channels are 3D printed and should be accurate to less than 0.5mm depending on how deep I push the WBFKs into the slot.

 

[Xymordos]

But your design didn't use the ED29689. WBFK always create a dip in the frequency response when used together with that driver as Peter said.