Homemade headphone amp sounds great, but...need some advice

[duesentrieb]

[size=x-small]Hi,[/size]
 
[size=x-small]Need your support![/size]
 
[size=x-small]Intro:[/size]
[size=x-small]When trying out my new custom-mold UE adaptors (2nd pic), I found they sound different on various eqt. (Iriver 340-HD-Music-Player, Samsung NC10-Netbook, N95-Multimedia Phone). [/size]
[size=x-small]The reason is the output impedance of the above devices in combination with the UE´s (Super.fi 5 Pro) impedance, which is really going rollercoaster (20Hz: 20Ohms, 1KHz:130Ohms, 11KHz:7Ohms, 20KHz: 12Ohms, calculated by measuring the voltage across the UE´s driven via a 240Ohms series resistor; due to the 2-way systems, first maximum probably caused by the inductance of the bass/middle tone system, the minimum probably caused by the series resonance of the coupling cap with the inductance of the tweeter).[/size]
 
[size=x-small]Here comes the question:[/size]
[size=x-small]Driven by a homemade low output impedance amp, as shown below, sound is really great, undistorted, linear and very private with the custom molds, even in loud trains, thanks to the great LME49720 dual amp in combination with 2 BUF634 video drivers.[/size]
[size=x-small]Please have a look at the design below, I invested a lot of thinking to drive <7Ohms loads with 250mA peak current drive capability, combined with lowest possible distortion and lowest noise (in the -6dB position, due to the extreme sensitivity of the UE´s and also in the +6dB positon to drive my Grado´s at home a little louder). [/size]
[size=x-small]Great ideas and circuit suggestions were taken from Jan Meier´s homepage (http://www.meier-audio.homepage.t-online.de/) and personal E-mail discussion.[/size]
 
[size=x-small]Now, I am using the amp also with the above mentioned Nokia N95 phone. [/size]
[size=x-small]When taking or making a call, there is this tack-tack-tack noise, caused by RF-rectification at the amp´s IC pn-junctions. [/size]
[size=x-small]I did my best to attenuate the 900MHz antenna signal from entering the circuitry by the shown ferrite beads, in combination with the shunting cap´s at the input, but the output leads (cable of the UE) still collects the signal quite strong. [/size]
[size=x-small]How are these Nokia mobile phone designers overcoming the problem, this device is perfectly unsusceptible for the max. of 2 Watts of RF power???[/size]
 
[size=x-small]Any ideas from the forum???[/size]
 
 
all the pics are also here...
http://www.head-fi.org/gallery/album/view/id/0/user_id/156470

 

[G1DRP]

I don't think you'll have much luck getting rid of that noise. The RF is pulsed and probably gets into the audio stages of the phone.
Cheers,
ian

 

[G1DRP]

Nice amp by the way!

 

[duesentrieb]

Hi G1DRP,
 
thanks for the comment.
Meanwhile I could achieve a reasonable improvement by inserting a heavy alumminum foil over the connector area. It is pushed elastically against the left and upper side of the enclosure due to a plastic insulator foil inside.
Well, the noise is still audible, but a phone call is no longer seriously disturbed...
(These Neutrik jacks really look like shielding the RF, but actually they are acting as antennas quite effectively, btw. the switch as well...)
 
Btw.: Most benefit in audio quality is gained with the N95 phone and the Samsung netbook (...really great portable sound now, even w 24bit resolution!).
The Iriver´s output amp has quite a low output impedance, but it lacks some current driving capability, therefore it sounds well also w/o amp at low volume...
 
Regarding the recurring questions in the forums about when an amp is useful, I can make one recommendation: Try connecting your source to a good HiFi amp, connect the headphone there and compare. If it gets better, then get a headphone amp...
 

 

[Ypoknons]

Have you tried the DIY forum? 'cuz I can't. 

 

[duesentrieb]

Good point!
 
Will try to continue to get the ultimate EMI answer in the DIY forum...

 

[MrProggie]

Have you tried ferrite clamps or clip-ons on the cables inside the box?
 
Edit: I see you use ferrite beads. Try some small clamps on the headphone wire...

 

[duesentrieb]

Thanks, good idea MrProggie,
 
Tried out, but did not have any effect, unfortunately...
 
In parallel I did modify the headphone cable from the shown relatively un-flexible thick unshielded one (which is molded to the sensor connector on the above pic.) to a flexible thinner shielded one (Why, this is a different story about the common ground wire of the UE up the cable branch, which impacts channel separation quite heavily).
Anyway, hat did help a lot!
When you hold the phone in active GSM mode to the earphone cable, noise is much less on the shielded part of the cable (no matter if the input cable is attached or not).
This means to me that most of the distortion is picked up and rectified by the output stage of the amp.
As a not so nice consequence (stability issues), I will try to shunt the output with some few pico farad caps.
Let´s see...