Home-Made IEMs
Jan 9, 2021 at 7:05 PM Post #12,122 of 12,426

Wgibson

100+ Head-Fier
Joined
Aug 18, 2020
Posts
140
Likes
109
Location
Tulsa, OK
Driver inside a zobel update, took some impedance measurements with a CI-30120 and Raf-32873 because the numbers worked out pretty well on paper. CI alone, then with regular zobel, then with Raf as the Rz, then zobeled the Raf.

Have not measured frequency response yet, but this should be a functional 2 driver design (or 5 driver, if you did 4 Rafs in 2x2 series/parallel) as the cap inside the CI's zobel would put the Raf at about 330hz high pass filter. And you could mechanically low pass the CI, with small ID/long tube, and/or dampers.

Next step is to get individual driver frequency measurements to see if they are in the right ballpark to play together as far as volume goes.

Zobel impedance - RAF inside.jpg
 
Jan 9, 2021 at 7:30 PM Post #12,123 of 12,426

dhruvmeena96

Headphoneus Supremus
Joined
Dec 13, 2014
Posts
2,599
Likes
1,291
Location
Gurgaon, India
Driver inside a zobel update, took some impedance measurements with a CI-30120 and Raf-32873 because the numbers worked out pretty well on paper. CI alone, then with regular zobel, then with Raf as the Rz, then zobeled the Raf.

Have not measured frequency response yet, but this should be a functional 2 driver design (or 5 driver, if you did 4 Rafs in 2x2 series/parallel) as the cap inside the CI's zobel would put the Raf at about 330hz high pass filter. And you could mechanically low pass the CI, with small ID/long tube, and/or dampers.

Next step is to get individual driver frequency measurements to see if they are in the right ballpark to play together as far as volume goes.

Next step is to see frequency response and if RAF is causing a dip
Then inverting it's phase or the whole zobel (+ to - side and - to + side).. if it is causing a dip in normal setup
 
Last edited:
Jan 9, 2021 at 9:02 PM Post #12,124 of 12,426

Wgibson

100+ Head-Fier
Joined
Aug 18, 2020
Posts
140
Likes
109
Location
Tulsa, OK
It is workable, could use some help on the high end. The 0.2mm ID brass tube is almost impractically small, but clearly functions as a low pass filter. Maybe just use a short section of it as an orifice, and maintain more of the subbass volume. Lots of physical tweaking to do if I decide to finalize it. This was mostly just for proof of concept, and electrical practice.

The Raf 330hz high pass showed up as pretty much as expected. No cancelation issues either.

Edit: in theory, I could close the gap on low/mid range between the blue and yellow curves, by adjusting the CI tubing.

CI-RAF inside zobel.jpg
 
Last edited:
Jan 9, 2021 at 9:50 PM Post #12,125 of 12,426

dhruvmeena96

Headphoneus Supremus
Joined
Dec 13, 2014
Posts
2,599
Likes
1,291
Location
Gurgaon, India
It is workable, could use some help on the high end. The 0.2mm ID brass tube is almost impractically small, but clearly functions as a low pass filter. Maybe just use a short section of it as an orifice, and maintain more of the subbass volume. Lots of physical tweaking to do if I decide to finalize it. This was mostly just for proof of concept, and electrical practice.

The Raf 330hz high pass showed up as pretty much as expected. No cancelation issues either.

Edit: in theory, I could close the gap on low/mid range between the blue and yellow curves, by adjusting the CI tubing.

What I will do is heavily damp CI with two yellow first
 
Last edited:
Jan 9, 2021 at 11:34 PM Post #12,126 of 12,426

Wgibson

100+ Head-Fier
Joined
Aug 18, 2020
Posts
140
Likes
109
Location
Tulsa, OK
I have some dampers in the mail, the benefit of smaller brass tubes is easy fit into the nozzle. And cost, 1 foot of small brass tube costs less than a pair of dampers. But I haven't listened yet, could be other issues.

Could be other advantages too, regarding previous discussion in the thread about 3d printing a spiral tube shape, similar to the se846 manifold, maybe you could coil the brass on the outside of a PVC tube to get extra length.

Edit: I may put a SR-6438 in there instead of the Raf, that could be an interesting cheap-ish basshead set up. I have 4 pairs of the Raf and want to do something a little more reasonable with them. Not worth using up the Rafs unless I add another driver(s) for a boost on the top end, and I need to get more finished projects under my belt before I go that complex.

The SR-6438 is a little more sensitive than the Raf, which will be good because the CI is relatively loud. 25 ohm SR vs 22 Raf, may mess up the impedance a bit but it should still be pretty flat.
 
Last edited:
Jan 10, 2021 at 2:09 AM Post #12,127 of 12,426

dhruvmeena96

Headphoneus Supremus
Joined
Dec 13, 2014
Posts
2,599
Likes
1,291
Location
Gurgaon, India
I have some dampers in the mail, the benefit of smaller brass tubes is easy fit into the nozzle. And cost, 1 foot of small brass tube costs less than a pair of dampers. But I haven't listened yet, could be other issues.

Could be other advantages too, regarding previous discussion in the thread about 3d printing a spiral tube shape, similar to the se846 manifold, maybe you could coil the brass on the outside of a PVC tube to get extra length.

Edit: I may put a SR-6438 in there instead of the Raf, that could be an interesting cheap-ish basshead set up. I have 4 pairs of the Raf and want to do something a little more reasonable with them. Not worth using up the Rafs unless I add another driver(s) for a boost on the top end, and I need to get more finished projects under my belt before I go that complex.

The SR-6438 is a little more sensitive than the Raf, which will be good because the CI is relatively loud. 25 ohm SR vs 22 Raf, may mess up the impedance a bit but it should still be pretty flat.
i said to damp CI, so you can get harman style bass shelf and then some optimization afterwards will make it really sound flagship level nice
 
Jan 10, 2021 at 9:41 AM Post #12,128 of 12,426

HenacanPodwell

New Head-Fier
Joined
Oct 9, 2018
Posts
35
Likes
5
Location
Malaysia
Hello everyone, I am trying to cross 1audio 31736 at 10khz, i have tried many series capacitor and the best I could get is with 0.11uf which gives me this graph, and the SPL is crazy low. However, what I would like is to only have frequency above 10k,, is there anyway I could do this? Would love it if someone can guide me to the right path. Thanks ! :)
Edit: Picture added.
1610293347681.png
 
Last edited:
Jan 10, 2021 at 10:36 AM Post #12,129 of 12,426

dhruvmeena96

Headphoneus Supremus
Joined
Dec 13, 2014
Posts
2,599
Likes
1,291
Location
Gurgaon, India
Hello everyone, I am trying to cross 1audio 31736 at 10khz, i have tried many series capacitor and the best I could get is with 0.11uf which gives me this graph, and the SPL is crazy low. However, what I would like is to only have frequency above 10k,, is there anyway I could do this? Would love it if someone can guide me to the right path. Thanks ! :)
Add a bigger cap and control resist it.
There would be many combos but you have to do it

Smaller single caps also limits power transfer, so they might have lower distortion, but they also get a load of dB loss.

If it was sharp cutoff, then you might have got some efficiency, but it's hard to implement in iem
 
Jan 10, 2021 at 10:42 AM Post #12,130 of 12,426

HenacanPodwell

New Head-Fier
Joined
Oct 9, 2018
Posts
35
Likes
5
Location
Malaysia
Add a bigger cap and control resist it.
There would be many combos but you have to do it

Smaller single caps also limits power transfer, so they might have lower distortion, but they also get a load of dB loss.

If it was sharp cutoff, then you might have got some efficiency, but it's hard to implement in iem
Thank you for the reply!
What do you mean by control resist it? I tried bigger series capacitor, but it has a lot of 7 to 8k, and it makes it really sharp, i only wants the treble after 10k.
Also I just realized that i forgot to include the picture of the graph! Silly me.
1610293333461.png
 
Jan 10, 2021 at 11:13 AM Post #12,131 of 12,426

dhruvmeena96

Headphoneus Supremus
Joined
Dec 13, 2014
Posts
2,599
Likes
1,291
Location
Gurgaon, India
Thank you for the reply!
What do you mean by control resist it? I tried bigger series capacitor, but it has a lot of 7 to 8k, and it makes it really sharp, i only wants the treble after 10k.
Also I just realized that i forgot to include the picture of the graph! Silly me.
Hmmmm
Having bigger cap will cause you to have those peaks.

Ok here is my suggestion

Run swfk on fullrange and first adjust tube lenght and ID in such a way, that most of peaks and dip are mostly sorted(shifting peaks to comfortable region).

Bigger lengh tube shifts the peaks to right
Smaller lenght tube shifts peak to left.
Also, there is tube resonances, so as you increase tube lenght, the tube resonances make the peak shifting to right of frequency response graph starts to rise and maximum gain happens when BA natural peak coincides with resonance shift and tube resonance.

So here is my suggestion
Which might work perfectly for you

First adjust the tweeter tube lenght on fullrange. (2mm ID)
Then turn this tube into stepped horn consisting 1mm, 1.5mm and 2mm

Then try crossing it

Try to shift resonances to higher octaves
 
Jan 10, 2021 at 11:24 AM Post #12,132 of 12,426

HenacanPodwell

New Head-Fier
Joined
Oct 9, 2018
Posts
35
Likes
5
Location
Malaysia
Hmmmm
Having bigger cap will cause you to have those peaks.

Ok here is my suggestion

Run swfk on fullrange and first adjust tube lenght and ID in such a way, that most of peaks and dip are mostly sorted(shifting peaks to comfortable region).

Bigger lengh tube shifts the peaks to right
Smaller lenght tube shifts peak to left.
Also, there is tube resonances, so as you increase tube lenght, the tube resonances make the peak shifting to right of frequency response graph starts to rise and maximum gain happens when BA natural peak coincides with resonance shift and tube resonance.

So here is my suggestion
Which might work perfectly for you

First adjust the tweeter tube lenght on fullrange. (2mm ID)
Then turn this tube into stepped horn consisting 1mm, 1.5mm and 2mm

Then try crossing it

Try to shift resonances to higher octaves
Thank you for the swift reply! I see, I will try it out tomorrow as its getting late here. Another question, for dual drivers like 2389d, do i only need to solder one terminal of a ba or do i have to solder both of them? What about drivers like 30017? Thank you!
 
Jan 11, 2021 at 9:05 PM Post #12,133 of 12,426

dhruvmeena96

Headphoneus Supremus
Joined
Dec 13, 2014
Posts
2,599
Likes
1,291
Location
Gurgaon, India
Thank you for the swift reply! I see, I will try it out tomorrow as its getting late here. Another question, for dual drivers like 2389d, do i only need to solder one terminal of a ba or do i have to solder both of them? What about drivers like 30017? Thank you!
It's a choice..

I always solder wire on both BA of 2389 and then do any kind of series or parallel

For twfk 30017
I snip of the pre done wire

And seperate wbfk and FK

And then use it

Makes life easier as there are some situation where you need to reverse only one driver phase or you need very complex crossover on one specific driver
 

Users who are viewing this thread

Top