CMoy pocket Headphone amplifier. Input coupling capacitor tests

My son recently became the proud owner of a pair of audiophile headphones. A pair of Beyerdynamic DT880 premiums. See the specs http://zs6rsh.blogspot.com/2015/07/headset-technical-specification.html

 Motivated by my friend Chris KD4PBJ, I built a prototype of the CMoy pocket headphone amplifier.  As specified, I used an OPA 2132PA Burr-Brown Op Amp. The rest of the components came out of my junkbox.

In reading the section on Input Capacitors for Headphone Amps, the author recommends the use of Polypropylene Film Capacitors as the best choice followed by Polyester Capacitors and to stay well away from Ceramic Capacitors. A very interesting read. I understand that a major issue in audiophile amps with AC coupling is Phase Distortion caused by the low frequency roll-off due to the RC high pass filter formed by the coupling capacitor and the amplifier input impedance. This causes a smearing of the Bass notes due to Phase delay from about 10Hz through 100Hz for a 0.1uF input coupling capacitor looking into an input impedance of 100Kohms.

Curious about this, I set up a test fixture to measure the phase distortion differences between what looked like a film poly capacitor in my junkbox (and which I deployed in my prototype) and a ceramic capacitor. I also ran an ideal model using LT spice. 

The high pass RC filter measured, consisted of the 0.1uF test capacitor and a 100Kohm resistor. As input I used a function generator set to approx 20mVp-p sinewave and two 10X probes connected to an oscilloscope to measure the phase delay between the input and output of the circuit. Lissajous figures were used to measure the phase angle to approximately a plus/minus 2 degree accuracy.

Attached find a graph of the plotted result. I noted that the measured -3dB filter knee for the poly capacitor was 17Hz plus/minus 2 Hz and the ceramic was 21Hz plus/minus 2Hz.

As can be seen, I was unable to measure a big difference in the phase delay of these capacitors using my 1970's era home lab equipment. The measured results compare favorably with the ideal capacitor phase delay calculated results.

In delving into the fascinating world of the audiophile at http://www.head-fi.org/ I came to realize that what constitutes a 'good sonic' is something highly complex and thus, highly subjective. So what is it that makes a polypropylene film capacitor 'sound' better than a ceramic capacitor? It is not revealed in my simplistic measurements. Perhaps driving the amp with a step signal (squarewave) would reveal much more about the differences in these capacitors?

The amp works very well with the DT880's to my ear although clipping of Bass notes was noticed when driven with 9volts. No clipping when using an 18volt source. Leading to an idea to build a buffer type virtual ground for the next version.

Thanks to the excellent website http://www.tangentsoft.net/, I have experienced a glimpse into the  complex world of the audiophile.

Now can I use one of these amplifiers to replace an LM386 in my regen receiver? :)

CMoy Prototype 1. The input and outputs should be isolated from the chassis!

The junkbox Wesco input capacitors can be seen. Next version needs 18volts to drive the DT880's. The Op Amp is a Burr-Brown OPA2132PA. Amazing sound! Zero hiss.

RC Network High Pass Filter test fixture. Input = 200mVp-p sinewave from a Function Generator (50ohm source impedance). Phase angles were measured using Lissajous figures. 10X scope probes used.