[vector1]

  Are you measuring it on open shell or bare drivers? I measured HODVTEC yesterday and while bare driver measured quite poorly it improved dramatically when I put my fingers near the vents almost closing them, bass improved and response was really good.
It's because HODVTEC having 4 quite large vents makes it seem as if the sound was leaking out.

 
It was open shell, considering only the front of the driver is embedded in the boot that I printed for testing. I'll certainly try blocking the vents tonight with some pieces of tape to see if that helps.
 
Also, the HODVTEC wiring diagram is quite confusing until I realized that the drivers were connected in parallel not series... Couldn't decide which side to solder on until I put some thought into it.
 
I'd love to test out sonion drivers, but I certainly can't afford to buy a tray of 25 at once... HODVTEC should do for now!
 
Xymordos
 
I can also hear that rolloff between 50-20Hz, which is why I ask. On the whole, having both CI and HODVTEC in the same shell might not do much for the overall sound. Maybe dual HODVTEC would be easier to configure than CI and HODVTEC, which are subtly different when I listen to then individually. I prefer HODVTEC at the moment though.

 

[tranhieu]

  I didn't listen to it but if I was to pick ultimate woofer it would be 38DJP007Mi/8a - this is one is unbeatable right now. From what I know no other woofer performs like this

Do you have a zoomed in impulse response graph of the 38DJP007Mi/8a? 

 

[Xymordos]

Wow the response of that driver is amazing o.o

 

[MuZo2]

There is no non-vented 3800
@muzo2 - as described it's Sonion 38DJP007Mi/8a
You can find this woofer in Roxanne, but Roxanne has two of those in each earpiece

Sorry I wanted to say dd drivers can be better alternative if tuned correctly. And with all those small drivers available its feasible now.

 

[TwinQY]

Would you folks reckon that a 38AJ007Mi/8a + 33A007 + 2 x 2354 setup be viable? Before I had been thinking to go HODVTEC + 33A007 + SWFK, but now I'm thinking that I'd like the 2354's characteristics a bit more. Not to mention that the THD on the 38AJ with the vents closed looks nice. Just occured to me to ask here since the 3800 came up anyways.
 
Comparatively, the CW-L51a has a similar setup but with a 2800 (dual 2600) instead of the 2354s. The CW-L71a I've heard uses two 2800s.
 
As an addendum, how is Colsan Microelectronics for shipping to NA in general? They're the only place I know of that will take orders lower than whatever Sonion's MOQ was. I could be mistaken.

 

[tranhieu]

  Would you folks reckon that a 38AJ007Mi/8a + 33A007 + 2 x 2354 setup be viable? Before I had been thinking to go HODVTEC + 33A007 + SWFK, but now I'm thinking that I'd like the 2354's characteristics a bit more. Not to mention that the THD on the 38AJ with the vents closed looks nice. Just occured to me to ask here since the 3800 came up anyways.
 
Comparatively, the CW-L51a has a similar setup but with a 2800 (dual 2600) instead of the 2354s. The CW-L71a I've heard uses two 2800s.
 
As an addendum, how is Colsan Microelectronics for shipping to NA in general? They're the only place I know of that will take orders lower than whatever Sonion's MOQ was. I could be mistaken.

1x2354 would be plenty enough if the low pass on the 3800 is done correctly.
 
By NA you mean North America? As far as I know Colsan is only responsible for UK and EU market and they will refer to your nearest distributor (in this case Sonion's US distributor) if you are not located in the region they're in charged of.

 

[shakur1996]

I didn't listen to it but if I was to pick ultimate woofer it would be 38DJP007Mi/8a - this is one is unbeatable right now. From what I know no other woofer performs like this

Peter, are you talking about just SPL level of 3800 or about its technical performance. In other words, 3800 is special because it's just very loud in the bass region or its technical performance(=quality) in the bass region is superb?

 

[piotrus-g]

  Do you have a zoomed in impulse response graph of the 38DJP007Mi/8a? 

It was measured on higher impedance set-up so with low impedance it can be slightly different
 

  Would you folks reckon that a 38AJ007Mi/8a + 33A007 + 2 x 2354 setup be viable? Before I had been thinking to go HODVTEC + 33A007 + SWFK, but now I'm thinking that I'd like the 2354's characteristics a bit more. Not to mention that the THD on the 38AJ with the vents closed looks nice. Just occured to me to ask here since the 3800 came up anyways.
 
Comparatively, the CW-L51a has a similar setup but with a 2800 (dual 2600) instead of the 2354s. The CW-L71a I've heard uses two 2800s.
 
As an addendum, how is Colsan Microelectronics for shipping to NA in general? They're the only place I know of that will take orders lower than whatever Sonion's MOQ was. I could be mistaken.

I'd say 33A007 is not necessary and I'd personally go with 2354 and 2389 instead of dual 2354 - just for the sake of slightly different sound and the fact you'd be able to wire 2389 in half coil and get bigger output to match high SPL from 3800. Nevertheless it will be very bassy set up and you have to keep that in mind. If you go with 3800 expect really low lows.
Colsan's MOQ is 25 pcs
 

Peter, are you talking about just SPL level of 3800 or about its technical performance. In other words, 3800 is special because it's just very loud in the bass region or its technical performance(=quality) in the bass region is superb?

I think both... the unit I measured here is simply the only woofer I know that drops after 500Hz and doesn't go up again. CIs, BKs, 3300, 3700, (HO)D(V)TECs even the regular 38AJ always has some sort of bump around 1-2kHz that is higher than actual bass response, which means it has to be attenuated with dampers.
From my perspective (as a manufacturer of silicone IEMs) the best woofer would be smaller sized woofer with the same type of response and much lower SPL. 110-115dB is plenty, who needs 125dB? With 125dB you either have to go with a lot of resistance or very low impedance for tweeters - this is not something that is DAP-friendly solution

 

[tranhieu]

 

It was measured on higher impedance set-up so with low impedance it can be slightly different
 

Cheers. Decay time looks similar to that of the CI. 125dB doesn't seem too much considering it's a dual receiver.

 

[piotrus-g]

  Cheers. Decay time looks similar to that of the CI. 125dB doesn't seem too much considering it's a dual receiver.

Well. 2389 have 104dB at 1kHz. With two 2389 you get 110dB so you still have 15dB difference between 200 and 1000 Hz. So you need at least 3 2300 drivers to make it sound around natural. or 4-5 TWFK-30017 If you're working in acrylics - sure it's quite possible set up; in silicone it's impossible though

 

[tranhieu]

  Well. 2389 have 104dB at 1kHz. With two 2389 you get 110dB so you still have 15dB difference between 200 and 1000 Hz. So you need at least 3 2300 drivers to make it sound around natural. or 4-5 TWFK-30017 If you're working in acrylics - sure it's quite possible set up; in silicone it's impossible though

Well SPL is one thing, but what matters more is the low pass's cutoff. I'm pretty sure there'd be no problem with a 15dB difference if one can move the cutoff down to 100hz or below. I've tried 120dB with a dual CI before and they sounded pretty decent with a ~600hz cutoff already.
 
And personally I don't like a dual 2300 setup, unless they are half-wired.

 

[Xymordos]

What's a target response I should aim for if I'm measuring on the ARTA? I did a lot of testing and matching on my 5 driver. I replaced the ED23147 with a BK28507 as full range. After a lot of trials and errors, I got a response like this. I have the CI and the bk as full range, but with them wired normally, there is a 10db dip at 3kHz. With both of them wired reversely, there is a 10db dip at 2kHz. This graph has the CI wired normally, and the BK reverse wired. In every configuration I tried there seems to be a high-mid range phase cancellation of some sort (not an enormous dip of 30db though). The response above 10kHz and below 20Hz seems very inaccurate as they're quite erratic every time I measure them. 
te

 

[MuZo2]

If you measure your favorite IEM(not made by you) and overlay on it it should give you target response.

 

[Xymordos]

When I first got my 'DIY' 711, I measured my own UERMs. This is the frequency response of my UERM left earpiece (right earpiece bass driver is wonky). I wonder if this 711 is accurate based on this graph?
 

 

[tranhieu]

No it's not...