[carmatic]

A year or so ago, I made an impedance adaptor for my IEM's according to this https://diyaudioheaven.wordpress.com/headphones/headphone-attenuation-adapter/ , so that I can eliminate the hiss from the amplifier and also I can use more of the volume knob, plus it makes the audio coming from 'low end' devices like computer monitors and sound cards sound better due to driving their amps at natural levels

But I have always had a problem with the bass response mismatch in my IEM's, so I have decided to modify the headphone attenuator from just being a simple resistor circuit to incorporate a separate lowpass filter for each channel
I used the Falstad simulator and I found that an inductor of around 50 miliHenry's together with a low resistance would act like a deep bass frequency filter, with gradually decreasing attenuation from 100hz downwards

The closest thing I could find are these toroidal transformers :

They are specified as '2x47mH' , so I merely used one side to act as a 47mH inductor

The other thing I needed to change was the low-resistance output impedance, it needed a 470microfarad capacitor in series to stop the low frequencies being bypassed by the inductor from simply also bypassing the driver and presenting a low impedance load to the amplifier
I also couldn't find a non-polarized 470mF capacitor, so I had to improvise with back-to-back electrolytic capacitors in parallel with back-to-back diodes to prevent the capacitors from being charged the wrong way
I added a potentiometer so that I can adjust the level of bass bypass so that I can vary the level of bass in each channel
Here is the schematic for one channel:

the 45 ohm resistor represents my IEM driver

this is the link to the Falstad simulation http://www.falstad.com/circuit/circ...YMgGBBd1WmAwGBC7SiYukVEYAHsmW8VJAFt1DnIBdEGEA

And here is the actual physical circuit I made over the weekend:

It has a bunch of long wires running the length of the board since the physical layout doesn't match the circuit layout, and I did it that way because I didn't think of adding the capacitors initially, but actually it keeps the tall components away from the headphone socket so it's easier to hold while you are plugging or unplugging the headphones
I used higher valued resistors in parallel to get the resistance values I need, I figured that since none of the resistors have the exact resistance they are labelled with, I would use them in parallel so that their variances are averaged out and I get a resistance value which is closer to what I need

From actualy using it with my ER4-SR's, I find that I needed to increase the volume on my amplifier to get the same loudness, but the bass is also increased which I like, and most importantly I could correct for the bass imbalance of my drivers...
However, the sound quality has changed, it is no longer as 'pure' as when I plug the IEM's directly into the amplifier, or if I was using my old simple resistance attenuator... I mean all the frequencies are still there, but they sound more 'distant' now, this is especially apparent in complex sounds like vocals and good recordings of musical instruments... perhaps it is to do with my capacitor and diode arrangement?
Also, one side of each of the toroidal transformers is not connected to anything... according to the simulation, if I were to short out this unused side, it removes the frequency-dependent attenuation, as if the entire inductor is shorted... maybe in the future, I can add some kind of bandpass circuit to this side to manipulate the sound more?

 

[jagwap]

An interesting idea. However I would not assume it is transparent. The higher output impedance will affect the frequency response unless you headphones are purely resistive in impedance, which they are not >500Hz.

Also I think you can lose the diodes. At best they are doing nothing, at worst they will start to conduct a little at low frequencies, causing THD. Back to back electrolytics sort themselves out through their leakage current at these voltages.

 

[carmatic]

An interesting idea. However I would not assume it is transparent. The higher output impedance will affect the frequency response unless you headphones are purely resistive in impedance, which they are not >500Hz.

Also I think you can lose the diodes. At best they are doing nothing, at worst they will start to conduct a little at low frequencies, causing THD. Back to back electrolytics sort themselves out through their leakage current at these voltages.

hey, thanks for your feedback
I was worried about reverse polarizing the capacitors, but it turns out you are right, deleting the diodes from the simulation seems to make it still run properly! I will remove them from my board and see how the sound improves

 

[carmatic]

The higher output impedance will affect the frequency response unless you headphones are purely resistive in impedance, which they are not >500Hz.

Regarding this.... the output impedance should be 5 ohms at high frequencies, since that is the value of the resistance in parallel with the driver... at low frequencies, the capacitors which are in series with the resistor (and therefore also in parallel with the driver) will increase the impedance to force the low frequencies towards the driver

 

[carmatic]

.

Also I think you can lose the diodes. At best they are doing nothing, at worst they will start to conduct a little at low frequencies, causing THD. Back to back electrolytics sort themselves out through their leakage current at these voltages.

So i've removed the diodes, and now i am able to push even more bass out of my IEM's without distortion! it's great! and the potentiometer adjustments are much more tangible now , i can actually hear the bass getting stronger and stronger as i turn on the knob, instead of the sound suddenly distorting past a certain treshold