[Farquarl]

The design will not be made by me. The guy who will scan the earmolds will also do the 3d file for the printer.
Hum, will try to find some local suppliers but i don't think i will be very lucky on this one

If anything fails send me a pm. I have a professional earmould software and Biocompatible resin.

 

[mattmatt]

If anything fails send me a pm. I have a professional earmould software and Biocompatible resin.

+1 for Farqual!

 

[musmecca]

So I have been working on the scary build as designed by piotrus-g. Built by pc27618349 as well. Basically 2 CI's and a Twfk with a pretty complex crossover. If you guys are on, i'd appreciate the feedback as my two screenshots are attached....noticeable dip. I was thinking phase issue but I checked the wiring and all seemed correct. The other issue is I DID use tantalum caps...I had the correct values and always put positive to positive. I remember a discussion a while back where it was suggested you had to put tantalums in series then someone else said it didn't matter. I guess my question is does it matter and is this possibly my issue?
https://imgur.com/gallery/MQSZDAW

 

[musmecca]

On a positive note, been making some great face plates lately. I've been cutting thin verners ( using a scroll saw) out of stabilized blocks then laying them on glass. I then coat with photoplast using a sponge applicator. I let them sit for 5 minutes to get the bubbles out of the wood then give a last smear of photoplast before sandwiching between glass...make sure to press the glass together and you will see the voids get forced out. I then cure 5 min both sides then release. Usually a razor breaks the seal but sometimes a rubber mallet is necessary. Final produt has a glass like finish with no greasy residue...I simply trim. glue to body, sand and laq.

https://imgur.com/gallery/9DqOPsA

 

[Farquarl]

So I have been working on the scary build as designed by piotrus-g. Built by pc27618349 as well. Basically 2 CI's and a Twfk with a pretty complex crossover. If you guys are on, i'd appreciate the feedback as my two screenshots are attached....noticeable dip. I was thinking phase issue but I checked the wiring and all seemed correct. The other issue is I DID use tantalum caps...I had the correct values and always put positive to positive. I remember a discussion a while back where it was suggested you had to put tantalums in series then someone else said it didn't matter. I guess my question is does it matter and is this possibly my issue?
https://imgur.com/gallery/MQSZDAW

Switch the polarity of the twfk of the one with the dip.
Tantalums in series with what? Polarity on Tantalums doesn't matter too much in this application. However I would recommend using ceramic ones as they are better and cheap.

 

[musmecca]

Switch the polarity of the twfk of the one with the dip.
Tantalums in series with what? Polarity on Tantalums doesn't matter too much in this application. However I would recommend using ceramic ones as they are better and cheap.

So this is what I was building from...Pitrous-g design on page 265:

Alright here's the second one more fancy, more neutral (still quite fun! but more on the sub-bass side)
It's 4 way design for those who don't fear soldering since it's more complicated.

Note that the twfk isn't denoted as positive/negative in the above design. In looking at the twfk spec sheet I assumed the left black was NEGATIVE and the side with the caps was POSITIVE, so thats how I wired it. One thing that may affect the dip is I did overflow my solder a bit for the negative onto the "no connection" terminal in the center. I checked the ohms of both my twfk and they were right at 28 ohms each so I assumed I didn't mess up anything. Your thoughts? I'll reverse them and see what happens...do you suggest I do the one with the dip or both twfks even though there isn't a pronounced dip in the other?

On the subject of tanatalums, series was with each other. For example having two 5 UF tanatlums with their negatives soldered together would yield a 10 UF cap with no polarity issues vs just using a 10 uf Tanatalum cap. I agree with your assessment that it shouldn't matter as the polarity makes a difference in higher voltage applications where the dielectric can be broken down over time if wired incorrectly...plus this is ac, not dc correct?

You have been a great help! I actually swapped the caps/resistors to the low pass on my purple build as you suggested. The shift didn't yield exactly what I was looking for so you have prompted me to complete my project board test rig! Any tips or tricks on how you set yours up (if you have one) would be appreciated. I'll post graphs once I have made progress.

 

[piotrus-g]

Note that the twfk isn't denoted as positive/negative in the above design. In looking at the twfk spec sheet I assumed the left black was NEGATIVE and the side with the caps was POSITIVE, so thats how I wired it.

The Red lines in the drawing are positive wires. Swapping polarity on TWFK is not that simple - see that the (now) negative pad has a wire connecting pads between both drivers.
I would say you either damaged one of the drivers or something is not right with your cross-over.
Start troubleshooting by measuring and comparing both CI set ups and then separately TWFKs. You should get a bit clearer idea where the problem is located.

On the subject of tanatalums, series was with each other. For example having two 5 UF tanatlums with their negatives soldered together would yield a 10 UF cap with no polarity issues vs just using a 10 uf Tanatalum cap.

Also caps capacity will be divided when wired in series and add up when wired in parallel. So the above is incorrect

 

[musmecca]

The Red lines in the drawing are positive wires. Swapping polarity on TWFK is not that simple - see that the (now) negative pad has a wire connecting pads between both drivers.
I would say you either damaged one of the drivers or something is not right with your cross-over.
Start troubleshooting by measuring and comparing both CI set ups and then separately TWFKs. You should get a bit clearer idea where the problem is located.

First let me say I'm honored you responded. You have contributed a ton on this site and it is much appreciated given you have a business to run. So I have attached 4 graphs: The first green is the left as wired...slight dip. The second red line is with the TWFK unhooked completely....you may be on to something as it looks flat. I thought this design would have a large bass boost....hmm. ( Are the vents for these ci's on the top and could I have gotten glue in them? Also...how much bass do you pick up when sealing the IEM?) Third graph in orange is the TWFK leads reversed. As you noted, it may not be as simple as swapping the leads. The last green is the right ear for reference. I'd appreciate your feedback...do you think the tantalum caps are to blame? Happy to draw out my schematic....If the pics don't load, the link is here:

https://imgur.com/gallery/iIxjpoh

right one:
https://imgur.com/gallery/h5oNOw9

 

[Farquarl]

I think there is most likely something wrong with your crossover/polarity. How you you measure your responses? All these graphs look kind of weird to me. The fact that the Dip dissapears ´when you measure without the TWFK looks even more like a phase issue. I forgot that the TWFK is wired that way, maybe change the polarity of both CI´s as this would be easier.
There are no vents on the CI so there is no way to block them.

 

[musmecca]

Also caps capacity will be divided when wired in series and add up when wired in parallel. So the above is incorrect
[/QUOTE]
Thanks for that. Is it a problem to use single Tanatalums (I put the + to +) or do you suggest always series/parallel?

 

[musmecca]

I think there is most likely something wrong with your crossover/polarity. How you you measure your responses? All these graphs look kind of weird to me. The fact that the Dip dissapears ´when you measure without the TWFK looks even more like a phase issue. I forgot that the TWFK is wired that way, maybe change the polarity of both CI´s as this would be easier.
There are no vents on the CI so there is no way to block them.

The dip disappearing was strange...I was thinking phase as well. Before I chase my tail some more I think your idea of examining my test rig is a good one. I'm using a umik-1 mic with the corresponding calibration file loaded into REW. My coupler is 15MM long. I test by wrapping the tubes in bluetack and making sure the opening has no holes. I have taken screenshots of my settings...you may see something amiss. One thing i have noticed is my scale....I need to condense it to the applicable ranges. Currently 1/6 smoothing, no phase, and 200 to 20K HZ. I have no idea what degrees are supposed to be set to. Thoughts?

 

[Farquarl]

The dip disappearing was strange...I was thinking phase as well. Before I chase my tail some more I think your idea of examining my test rig is a good one. I'm using a umik-1 mic with the corresponding calibration file loaded into REW. My coupler is 15MM long. I test by wrapping the tubes in bluetack and making sure the opening has no holes. I have taken screenshots of my settings...you may see something amiss. One thing i have noticed is my scale....I need to condense it to the applicable ranges. Currently 1/6 smoothing, no phase, and 200 to 20K HZ. I have no idea what degrees are supposed to be set to. Thoughts?

Are you sure the tube is attached airtight against you microphone? It kind of looks like it could be not. Otherwise no complaints.
On the software side you don't have to worrie about degrees since you are not measuring phase. I would recommended that you expand the lower limit from 200hz to 10Hz.
Please measure the twfk alone.

 

[cyph3r]

The design will not be made by me. The guy who will scan the earmolds will also do the 3d file for the printer.
Hum, will try to find some local suppliers but i don't think i will be very lucky on this one

If you can communicate with the service supplier, ask them to adhere to the proper post printing procedures (washing and UV curing). If not done correctly this can lead to discoloration, brittleness of the material. Also any unpolymerized resin (the goo that is poured in the printing vat) contains a number of components of which are many mildly to moderately toxic and or sensitizing (most notably the photosensizers compounds that catch the laser light quanta and start local polymerization) that can cause allergies. You don't want any of those residues in your ears or skin. Inspect the print after reception for any residues (likely clear highly viscous liquids, particularly as drawn by capillary forces into narrow openings, ducts etc); wash the printed parts with in water with mild detergent, postflush with water. Dont use isopropanol or other alcohols to clean the parts. That should do it; as additional you may want to coat any parts in skin contact with skin-compatible lacquer.

 

[musmecca]

Are you sure the tube is attached airtight against you microphone? It kind of looks like it could be not. Otherwise no complaints.
On the software side you don't have to worrie about degrees since you are not measuring phase. I would recommended that you expand the lower limit from 200hz to 10Hz.
Please measure the twfk alone.

You were right abouth the seal on the umik mic. I had used an industrial zip tie that was allowing a little air out just under the knot. I rolled the end with some pipe thread tape and used two zips....works great!

I did solve another problem fairly creatively that may be of use to others....the two ci's together apparently had the slightest of gaps between the two bodies and the tube that I never saw, and I had re-glued them numerous times. I was baffled I couldn't get the bass response of Piotrus-g's design, so after many failed screenshots I tried this....I got a long section of tubing, put one end into a foam earpiece and sealed with supertack and put it in my ear. I then simply ran music into the un mounted rig, passing the open end of my long tube over all junctions like a stethescope! As soon as I passed the open end over trouble spot, I could plainly hear the music and sound was indeed leaking there. I'd hit that spot with the bondic and problem solved. (BTW great suggestion on doing the tubes individually...it's what led me this way). With all my testing now lining up, I wanted to get these buttoned up. They were a BEAR to fit in the shells and I learned a lot about breaking wires and using strain relief in the future!....I quickly tested and sealed before something else decided to come loose. I need to do another touch up and smoothing coat as I was in a hurry so it isn't cosmetically perfect, but thanks Piotrus-G..it sounds AWESOME and I am thrilled. I apparently like bass...wondering what can top these!

 

[Farquarl]

If you can communicate with the service supplier, ask them to adhere to the proper post printing procedures (washing and UV curing). If not done correctly this can lead to discoloration, brittleness of the material. Also any unpolymerized resin (the goo that is poured in the printing vat) contains a number of components of which are many mildly to moderately toxic and or sensitizing (most notably the photosensizers compounds that catch the laser light quanta and start local polymerization) that can cause allergies. You don't want any of those residues in your ears or skin. Inspect the print after reception for any residues (likely clear highly viscous liquids, particularly as drawn by capillary forces into narrow openings, ducts etc); wash the printed parts with in water with mild detergent, postflush with water. Dont use isopropanol or other alcohols to clean the parts. That should do it; as additional you may want to coat any parts in skin contact with skin-compatible lacquer.

Great Job! I'm glad it worked out for you in the end.