[BadgerMcBadger]

are my ear canals too skinny for CIEMs? i can post more pictures if neccesery

 

[piotrus-g]

are my ear canals too skinny for CIEMs? i can post more pictures if neccesery

This is a good impression

 

[stegeoc]

Yes you can use STEPs but you can just use ARTA - it has an option to measure distortion inside the program. You have to have 64k FFT and sweep sequence length 64k with selected "center peak of impulse response" once you run measurement go to analysis and go to frequency response and distortion. Be advised - THDs rise with volume. It's generally a rule of thumb to set the 1kHz response of the IEM at 94dB SPL for measurement of THD .

Excellent, will give that a try soon. Do you have any rules of thumb for what is considered acceptable/good/bad? Or is that likely to vary based on in the individual setup?

Though I'll probably measure a selection of known iems anyway to give a baseline I can relate to

 

[BadgerMcBadger]

This is a good impression

good to know. i planned using the rab-p build but the 3mm at the end it bigger than my canal so ill have to drop it

 

[piotrus-g]

Excellent, will give that a try soon. Do you have any rules of thumb for what is considered acceptable/good/bad? Or is that likely to vary based on in the individual setup?

Though I'll probably measure a selection of known iems anyway to give a baseline I can relate to

<2% at anything below 200Hz is acceptable
<1% at 1kHz and above is ok
>3% at any range becomes very noticeable
Overall you are shooting at lowest possible - always, but some distortion can give a warmer tone to the sound.
Sonion BAs distort far far more than Knowles' - not saying this to bash any commercial designs, this is just a fact easily confirmable by checking drivers' datasheets

 

[swtnate]

Excellent, will give that a try soon. Do you have any rules of thumb for what is considered acceptable/good/bad? Or is that likely to vary based on in the individual setup?

Though I'll probably measure a selection of known iems anyway to give a baseline I can relate to

My focus has always been on stability of the distortion. Such as where are you seeing the spike in the frequency graph? Some distortion can be easily mitigated by tube length and damper placement. Electronic induced distortion is only mitigated by components. So, if you have excessive “noise” in the circuit that gets accentuated by wrong tube length and improper damper placement, your tests will show that. I will try to find the MELF resistor data sheets I use.

 

[kha1id]

Have been testing several RC Low Pass on Low BA Drivers. It sound fatigue and bored to my ears somehow.

There are some variations:-
RC 30R+2.2 or 3.3 uf on 33AJ007i/9 + 2389 hi pass at 7k Sharing White Damper + 2mm ID
RC 30R+6.8uf on CI 22955 + 2389 hi pass at 7k Sharing White Damper + 2mm ID

Like what it does on UERM

 

[stegeoc]

<2% at anything below 200Hz is acceptable
<1% at 1kHz and above is ok
>3% at any range becomes very noticeable
Overall you are shooting at lowest possible - always, but some distortion can give a warmer tone to the sound.
Sonion BAs distort far far more than Knowles' - not saying this to bash any commercial designs, this is just a fact easily confirmable by checking drivers' datasheets

Thanks, will see how things stack up. I have realised that I have no idea at the moment what loudness I'm measuring at. For my normal measurements I have it set so that it is at a fairly loud listening volume, which seems reasonable. Will try to calibrate that though (I do have a calibrated USB mic, so should be able to use that somehow.

I did a couple of very quick tests and one my in-progress build I was surprised that the 3rd order (I think) was quite high on a Sonion 38xj bass. Still less than 1% but it was much higher than the distortion on the Savant Pro that I also had to hand and has tiny drivers for bass. Nothing calibrated and sensitivities different, but it is interesting you mention Sonions' distoring more. I had always imagined the 38xj to be very clean given its size and power. Sounds good anyway though!

Like you say, harmonic distortion of a certain kind is ok anyway and even strived for sometimes. In some ways it is distortion that gives the feeling of loudness so without it it is more difficult to feel satisfied. But the wrong kind in the wrong place is of course very bad. Lot's to learn. Cheers!

 

[stegeoc]

Quick question on the GQ crossover... Does anyone have a schematic for it?

I built a GQ + RAB + 38XJ and did my experimenting on one side, but now I've built the 'same' on the other side, the treble is a different shape and I think I have accidentally modified/broken the GQ crossover on one (I don't know which) or maybe even killed one of the drivers somehow. I can see a difference in the Impedance curve for high frequencies too, so that all points to the GQ.

I've actually broken a number of soldering iron tips (very cheap ones) so am now using a huge tip which itself is miss-shaped already and soldering small components can be quite challenging so I am hoping to remove the existing SMDs and add some wires to some off-board components.

I should also order some more tips (and probably a new iron).

Update: Below is a pic of the GQ. Surprisingly I found that the datasheet for the driver actually shows the PCB traces, so that helps a lot already. I have a strong feeling that the WBFK is being left out on one side and I'm going to have to rebuild these



Thanks!

 

[BadgerMcBadger]

can someone please help me where to cut?

 

[GeneBush]

can someone please help me where to cut?

You need some new impressions my friend, these are poorly made but before you should consider a proper clean of your ear canals by a professional and use an otoblocker

 

[Wgibson]

$75 shipped in the US for 500ml of ceramic resin, I'll risk it. Planning to try dip/paint and cure by hand a single layer at a time, see if I can make small chambers and nozzles from a wax mold or something like that. I don't know the heat curing process details yet.

https://www.alphasense.net/product-page/3d-ceramic-resin-for-sla-lcd-and-dlp-printers

 

[BadgerMcBadger]

You need some new impressions my friend, these are poorly made but before you should consider a proper clean of your ear canals by a professional and use an otoblocker

they had an otoblock but i removed it, didnt think i had to keep it in there
why do you think they are not good?

 

[slakoth453]

$75 shipped in the US for 500ml of ceramic resin, I'll risk it. Planning to try dip/paint and cure by hand a single layer at a time, see if I can make small chambers and nozzles from a wax mold or something like that. I don't know the heat curing process details yet.

https://www.alphasense.net/product-page/3d-ceramic-resin-for-sla-lcd-and-dlp-printers

I think that stuff is for high temperature applications. You burn the resin out and leave ceramic behind. I imagine that it will be porous though. Not sure why you would want to use that for iems.

 

[citiesandshapes]

they had an otoblock but i removed it, didnt think i had to keep it in there
why do you think they are not good?

It would help us to have pictures from these 4 angles so we can see all the important parts of the impression. But based on your pictures, looks like primary issue is the canal area. They look quite skinny and there’s a good amount of wax on the impression. Get your ears as clean as possible then do another set