How in the world did I miss this chip for headphone amp applications?

JTEX

JTEX

Well-known member
Just discovered OPA1688. A great opamp that happily drives 16 Ohm loads at super low distortion and low quiescent current. I can't believe I wasn't aware of this not-so-new chip, being a TI fan and having spent quite a bit of time researching headphone amp options. Damn me!

In my own circuits I had settled on using both halves of an OPA1692 in parallel per channel. So, two dual opamps for stereo. Twice the real estate, twice the price of an OPA1688, and no advantages. Great. Now i have to redesign everything I make that has a headphone amp.
 
Nevermind. I am getting senile, it seems. Looking back through my docs, I think I was right about using 2xOPA1692. It puts out more power than a single OPA1688, for less total idle current (2x650uA = 1.3mA, vs 1.6mA).

Here's what I had measured a couple of years back:

OPA1692 with paralleled halves:
80mW into 16Ohms @ 0.006% THD, or 140mW/32Ohms @ 0.004% THD
idle current: 1.3mA

OPA1688 does get even lower distortion, but you can only get about 50mW out of it into 32 ohms, or 30mW/16Ohm. Not enough to blow your head off playing bass.
 
I would have never thought two would be better than one for low power. I really enjoy reading about your ultra low power designs, as it's something I never do anything with or think about for my own designs.
 
I did get even lower idle current and higher output power using CMOS op amps such as OPA2196 or OPA2990, but they just don't sound good enough to me with headphones. They get crossover distortion going into low loads. That sounds pretty nasty even at 0.1%. It's not harmonically related to the signal, so it realy sticks out as a sort of fizz or grit.

I have it on my agenda to try all 4 units of an OPA4990 in parallel. Still just under 0.5mA total idle current. Maybe 4 of them sharing the load would reduce crossover distortion to an acceptable level.
 
You must log in or register to reply here.
Back
Top