LT1206 composite headphone amp

[hta95]

After reading a few topics here regarding the composite topology for headphone amp, I decided to build one based on LME49710 and LT1206 chips and I really want your recommendations and optimizations about it especially the layout.

It's the composite headphone amp with the LT1206 running at unity gain in the feedback loop of the LME49710. As my headphones have low impedance (32 and 50Ohm) I set the low gain at 2 and the high gain at ~7.7 (in case of buying a HD600 for example).

The power supply is a variation of the capacitance multiplier. It gives promising result in LTSPICE with 95dB PSRR at 100Hz and the output impedance of ~40mOhm flat out from 10Hz to 100kHz

Thanks for your time! 🙂

 

[DPH]

Any specific reason you chose not to push the LT1206's gain into the outer loop? I mean, in reality the distortion numbers are going to be rather infinitesimal, but even having a g = 2 to 5 from the 1206 takes pressure off the 49710 to swing large voltages.

 

[hta95]

It's stability issue. I just want to start with something safe. I often see the output chip in composite topology has g=1 to 2.

 

[DPH]

With the PM and outright speed of the 1206, you should be able to safely close around it. Especially with a bit of noise gain well above audio frequencies. Just because most composites *do* have the output circuit near unity, doesn't mean it's the best solution. Not that realistically you have much to gain distortion-wise. Worth having a read of jcx's topic on the matter. 😀

http://www.diyaudio.com/forums/solid-state/45794-high-loop-gain-composite-op-amp-circuits.html

 

[hta95]

Alright then, I'm reading it 😉 do you have any thought on the pcb layout?

 

[DPH]

I'm terrible at layout, so I don't want to screw things up for you. 🙂

 

[hta95]

After reading the jcx's topic I think I'm gonna change the lt1206 gain to 2. Because the overall gain of the circuit is 2, the gain of the output op amp can't go any further than that.