[dhruvmeena96]

If we put sonion bass driver like 38D1XJ007Mi/8a in internal series(Sonion drivers comes with tabs only and no predefined connection)

We get a effective inductance of 31.6mH, DCR of 50ohms, which leads to internal low pass. Then its accupass structure and internal damper, low passes it good enough.

A L-pad of 4.7ohms on series and parallel is needed to flatten impedance while lowering it, and lower sensitivity. Adjust series resistor for taste, this is reference design

And a yellow damper(near spout) and orange damper(in between tube)
Best subwoofer setup acting as full range electrically.



Now comes the Midrange section, properly mastered
33AJ007i(non vented)

Series arrangement here, as we need high inductance value to make its peak smoother and make it natural low pass. Then we add a high pass to this setup, to limit itself to midrange.

And since we are not using zobel, we will add parallel resistor

10uF series capacitor
47ohms parallel resistor.

This will fix impedance peak and lower impedance for further usage ahead.
Use Yellow damper in between of tube



Best upper midrange and treble
Sonion 2389D(half coil parallel) + 2ohm resistor + 2.2uF capacitor.

Its way cleaner than single sonion 2389 and is more accurate to cross it up.

Brown damper in between the tube



Add 5ohm to whole circuit to make it decent resistant for proper loudness

Woofer tube lenght is 14mm(1.2mm ID)
Midrange tube lenght is 12mm(2mm ID)
Upper midrange and tweeter is 24mm(2mm ID)

@eunice @Jedrula1

Something of my simulation to freqPhase. Acoustically flat phase from tubing and linear phase in electric signal



True Harman kardon response

And you can add switch for bass tuning according to different resistor values. Increasing series resistor on L-pad tames the bass and lowers the volume(negligible but it does, as it increase circuit impedance.(recommended as it defines the taste of individual)

Add one more switch for treble tuning too, just increase or decrease the resistor and treble will increase or decrease. 2ohm is needed, extra can be added.(not recommended as treble is tuned perfectly....to response and ears).


No EST, its not needed.


Sonion 2389D is fastest BA on planet on half coil + 2ohm and a crossover.

 

[wolkegeist]

Soneeter's bard's headspace

What is this thing and how is affaect the sound? Has someone ever open it to see what's inside, if so i'd very much like to DIY it if possible.

 

[dhruvmeena96]

Best way to increase driver count

Method 1
TWFK23991(high impedance) and TWFK30017(low impedance

FK of both driver parallel = 20ohm DCR
WBFK of both driver parallel = 10ohm DCR



Just go on and on, and when you feel that impedance is too low, add resistor to compensate both FK and WBFK individually.

Keep 20ohm to FK and 10ohms to WBFK
Then make a crossover for FK and WBFK.







Method 2
Use TWFK 30017 and make a crossover as per Knowles guideline.

Take as many TWFK with crossover and slap them in parallel and compensate with resistor







Method 3
TWFK30017 x 4 drivers

FK x 4 (parallel)+ 20ohms (19ohms for exact)
WBFK x 4(parallel) + 10ohms(9ohms for exact)





Method 4
DWFK-31785 x 4 parallel + 20ohm
This method easily increase drive count.



All these method easily increase driver count for driver wars. Best method is TWFK23991 and TWFK 30017.

It is because in 2x23991 and 2x30017 (4driver set, 8 drivers per side), with 10ohms on FK parallel and 5 ohms on WBFK parallel, we get 4 mids, 4 treble and flatter yet controlled impedance(not too low).

@piotrus-g SCARY (2xCI and TWFK) but with 10drivers(8 from 4 TWFK and 2xCI) and some crossover tweaks.

Let's see what happens

 

[SupremusDoofus]

From where can we source sonion drivers from? I know taobao has them but taobao is all Chinese and google translate isnt helping very much. Besides the prices shown on taobao, is it just for one unit or for a pair like how soundlink sells them?

 

[eunice]

Soundlink has the Sonion 33AJ007i/9
https://s.click.aliexpress.com/e/1D1Wksg
Idk about the rest. You could check https://colsanmicro.com/

 

[piotrus-g]

The thing is, cable value are so negligible that, even if it affects the SQ, it is not heard by our ears. Our ears have a limit and so is brain

And if you wanna debate on cable
OFC and OCC copper have 2% of difference and the difference if calculated in decimal score, is so negligible that anybody would start laughing

Electrically and signal science point of view, cable has least effect.

The only positive effects are
1. Look and feel
2. Braiding to cancel any sort of inductance happening
3. More cable on braid means more parallel wire, which effect the resistance.
4. Better isolation of channels in hi fi cable(crosstalk), due to higher quality(only for balanced cables)

If you want to have signature change, that won't happen. Everything is going inside the brain bro


Yah, I forgot

Better microphonics as you up the prices as concept of litz cable comes into play. They are able to pack more cores of cable inside single one which helps in reducing bulkyness while going parallel internally.

I know what I am talking in cables. But the manufacturing cost and the hype just makes a false advertisements.

Sony EX1000 cable cost 99% and is 7N OCC copper which sound as good as Plusssounds tri copper

The best cable for me is Penon Draco, as I like looks and feel and given advantages.. But shelling tonne on a piece of wire is something I cannot do(penon draco was gifted to me)


There has been long debate, and we strive for difference that makes sense and not on materialistic approach of having being and shing. That's a rule a engineer has to follow, like doctors oath on protecting patient, we oath on improving technology and make it affordable

See Ferrari debut SF90 which outclasses Ferrari LA Ferrari at less that 1/2 cost.

This is what I do. I dont want my friends to shell out cash for that pseudo 10% improvement.

And If you have flat impedance and phase, cable effects are 0 then. And we have been playing here with flat impedance for a nice period of time.


Well, it is a good curiosity... Hope this answer your question well

Thankyou

Unfortunately you are not entirely right.
I was big cable skeptic until I heard FIBAE1 with higher end cable few years ago. That changed my perspective a bit, later on we found out what is happening and it was all confirmed with measurements.
Basically the biggest difference a good cable makes is in crosstalk, and it will play big role in stuff like amount of bass the IEM can produce. Especially when you have flat and low impedance cross talk starts to play huge role. A 4 wire single-ended cable improves crosstalk by about 15dB (as far as I recall) compared to 3-wire single ended cable. A higher gouge pure copper 4-wire SE improved cross talk number by something close to 30dB. Then there are some matters that play smaller role but still do - like impedance of the cable.

Since I think you have FIBAE1 - try it for yourself - listen to it with stock cable and then switch to OCC Copper with 26-24AWG wire.

Soundlink has the Sonion 33AJ007i/9
https://s.click.aliexpress.com/e/1D1Wksg
Idk about the rest. You could check https://colsanmicro.com/

Colsan will no longer be Sonion distributor as of October 2019.

 

[dhruvmeena96]

Unfortunately you are not entirely right.
I was big cable skeptic until I heard FIBAE1 with higher end cable few years ago. That changed my perspective a bit, later on we found out what is happening and it was all confirmed with measurements.
Basically the biggest difference a good cable makes is in crosstalk, and it will play big role in stuff like amount of bass the IEM can produce. Especially when you have flat and low impedance cross talk starts to play huge role. A 4 wire single-ended cable improves crosstalk by about 15dB (as far as I recall) compared to 3-wire single ended cable. A higher gouge pure copper 4-wire SE improved cross talk number by something close to 30dB. Then there are some matters that play smaller role but still do - like impedance of the cable.

Since I think you have FIBAE1 - try it for yourself - listen to it with stock cable and then switch to OCC Copper with 26-24AWG wire.


Colsan will no longer be Sonion distributor as of October 2019.

Thanks for your reply

I did mention what a cable improves.
And I mentioned crosstalk.

I don't think it changes the signature as people describe.
And I think channel crosstalk improves more in balanced topology rather than single ended....but I might be wrong.

Single ended or single ground are more dependent on amp crosstalk.

I can imagine why 4 cable single ended can improve crosstalks.

The channels are mostly isolated in cable, cancelling the EMF or the thumb ruled rotatory magnetic field. This leads better result than three cable design as the ground signal is often very different (L+R ) which leads to occurrence of issue rather than solving it. This causes the 15dB of crosstalk elimination.

On my simulation, it shows that single ended 3 cable actually adds crosstalk, while 4 cable fixes magnetic influence on other channel. On isolated field, the difference from isolated cable design to 4 cable(very bad, cheap metal) design leads microphonicsprovement on crosstalk whereas 3 cable(very bad metal) design leads to 6dB of added crosstalk.

The difference is 10dB. Improve material to economical standards and increase cable count and you will get 15dB from cheapo metal.

Its all about cable topology.

Litz configuration effect the total packaging plus microphonics. Due to its structure, they are more rigid and more immune to microphonics. These structure, if used properly can eliminate crosstalk issues too.


But, what we talked was signature change or overall change in sound.



But expensive cables with snake oil in it(Dita) and cyro silver is pointless.



I know the bad crosstalk from bad cable in RAB-p (relatively flat impedance)

Crosstalk just creeps up in bad cable(I have some old generation mmcx cable, the old cheap cable for 3$)

But I think most of normal cables are OK.

Well FIBAE 1 is even more flat than the RAB-p, so the issue might be even more then.






Targeting Area 51 for sonion drivers then....who all are in

 

[piotrus-g]

You did mention crosstalk but your first sentence in the post is

The thing is, cable value are so negligible that, even if it affects the SQ, it is not heard by our ears. Our ears have a limit and so is brain

and it's not true, the cross talk impacts bass to such an extent that you can hear (and measure) the roll-off with bad cable compared to higher end cable

 

[dhruvmeena96]

You did mention crosstalk but your first sentence in the post is

and it's not true, the cross talk impacts bass to such an extent that you can hear (and measure) the roll-off with bad cable compared to higher end cable

Sorry for mistake....
Well you are right

 

[wolkegeist]

@dhruvmeena96

I find the MASM 3 a bit lack of bass, should I add 1 more bass driver to the design? If so I think i need to move them into new shells.

I also like the MASM7 design tho, BS6+RAB. Which one do you think I should go for? I think I’ll order premade shells from taobao and redo the nozzle so it could fit more tubes.

 

[dhruvmeena96]

@dhruvmeena96

I find the MASM 3 a bit lack of bass, should I add 1 more bass driver to the design? If so I think i need to move them into new shells.

I also like the MASM7 design tho, BS6+RAB. Which one do you think I should go for? I think I’ll order premade shells from taobao and redo the nozzle so it could fit more tubes.

MASM 3 ENHANCED bass

Knowles HE instead of GQ
Same tubing lenght(shared between ED and CI)
Brown near spout, white in center

RAB32033
Same tube
No damper



MASM 7

Bellsing 6
Using 3mm ID, you stretch and create a adapter around the three bores. Then you take 2mm ID/3mm OD tube and insert inside the 3mm ID tube.

Same lenght as orignal MASM GQ
Damper
Green near spout
White in center


Same goes with RAB32033

Either make the zobel of two discrete( better error control due to parallel) or combine the zobel of these two. BS6 + zobel and RAB + zobel.

How is the bass lacking, you mean the subbass right.
Well on FR, the bass extension is fine. Issue is very tight decay and small impulse response making it's bass too fast.
I was thinking of adding bass driver also

Knowles HODTEC 31268
Zobel(1.428uF and 140ohms)
L-pad(4.7ohms series and 4.7ohms parallel)
18mm lenght tubing
1mm ID

Yellow damper near spout
Yellow damper in center of tube



For tweeter
SWFK-32255(shared with RAB tube)
1uF capacitor




See, you can modify the design any way you like.

Just make sure that all drivers except for tweeters must be full range as that is the trick behind the cohesiveness.
But I think, MASM 3 bass is more than enough if you listen to bassy tracks

 

[wolkegeist]

MASM 3 ENHANCED bass

Knowles HE instead of GQ
Same tubing lenght(shared between ED and CI)
Brown near spout, white in center

RAB32033
Same tube
No damper



MASM 7

Bellsing 6
Using 3mm ID, you stretch and create a adapter around the three bores. Then you take 2mm ID/3mm OD tube and insert inside the 3mm ID tube.

Same lenght as orignal MASM GQ
Damper
Green near spout
White in center


Same goes with RAB32033

Either make the zobel of two discrete( better error control due to parallel) or combine the zobel of these two. BS6 + zobel and RAB + zobel.

How is the bass lacking, you mean the subbass right.
Well on FR, the bass extension is fine. Issue is very tight decay and small impulse response making it's bass too fast.
I was thinking of adding bass driver also

Knowles HODTEC 31268
Zobel(1.428uF and 140ohms)
L-pad(4.7ohms series and 4.7ohms parallel)
18mm lenght tubing
1mm ID

Yellow damper near spout
Yellow damper in center of tube



For tweeter
SWFK-32255(shared with RAB tube)
1uF capacitor




See, you can modify the design any way you like.

Just make sure that all drivers except for tweeters must be full range as that is the trick behind the cohesiveness.
But I think, MASM 3 bass is more than enough if you listen to bassy tracks

Yeah it’s lacking in sub-bass. I can’t hear bass guitar at all, or it sound very fainted.
I’m thinking about dual RAF + GQ, and save the RAB for BS6 later. The shell I’m using is really tight tho, I don’t think I can fit HODVTEC, but if I remake the zobel with smaller components (I use big metal film resistors here, will switch to SMD ones) I can fit dual RAF in. I found the old design on this thread claimed that dual RAF has dynamic level bass.
RAF in series, 22Ohm in series, zobel 41Ohm, 4.7 uF

 

[dhruvmeena96]

Yeah it’s lacking in sub-bass. I can’t hear bass guitar at all, or it sound very fainted.
I’m thinking about dual RAF + GQ, and save the RAB for BS6 later. The shell I’m using is really tight tho, I don’t think I can fit HODVTEC, but if I remake the zobel with smaller components (I use big metal film resistors here, will switch to SMD ones) I can fit dual RAF in. I found the old design on this thread claimed that dual RAF has dynamic level bass.
RAF in series, 22Ohm in series, zobel 41Ohm, 4.7 uF

On dual RAF series, you might not get the treble of single RAF zobel. Done and tried. The bass increases due to increase in inductance which may lead to treble limiting .

First of all, I mentioned a HODTEC and not HODVTEC. But if space is your concern

Go for BK or CI. I prefer BK as it is more agile and will suit MASM 3 signature.

BK-21610 -> 30ohm resistor -> L-pad (4.7 parallel and 4.7ohm series)

Yellow damper x 2 (near spout and at center)

Small ID tube like 1mm(18mm length)

 

[wolkegeist]

On dual RAF series, you might not get the treble of single RAF zobel. Done and tried. The bass increases due to increase in inductance which may lead to treble limiting .

First of all, I mentioned a HODTEC and not HODVTEC. But if space is your concern

Go for BK or CI. I prefer BK as it is more agile and will suit MASM 3 signature.

BK-21610 -> 30ohm resistor -> L-pad (4.7 parallel and 4.7ohm series)

Yellow damper x 2 (near spout and at center)

Small ID tube like 1mm(18mm length)

How about CI tho, I see that they are more available than BK. I can’t find them on taobao.

 

[dhruvmeena96]

How about CI tho, I see that they are more available than BK. I can’t find them on taobao.

OK, then go with

[CI-22955 -> 26ohm resistor (series) ]-> L-pad(4.7ohm both parallel and series)

We need a CI to be 48ohm DCR to 50ohm DCR to apply L-pad. So we added CI with 26ohm resistor. 30ohms and others are fine till the desired DCR is between 42ohms to 55ohms. Ideal is 48ohms

Twin yellow stacked near the spout(the smaller Diameter yellow damper)

18mm length, 1mm ID.
Smaller ID, the better low pass