LME49600: OPA1622 headphone amplifier design review / TINA-TI instability at low rail voltages

Max Hunter

Part Number: LME49600
Other Parts Discussed in Thread: OPA1622, , TINA-TI, BUF634, BUF634A, OPA1633

Tool/software:

Hi,

I am designing a headphone amplifier for use in recording studio applications. It will need to drive high-impedance headphones to loud studio volumes.

I'm using the AK4499EX current-mode DAC. I have based a headphone amp design on the reference TI design https://www.ti.com/lit/ug/tidu672c/tidu672c.pdf?ts=1723909825481. The product is powered by a 12V supply which I'm regulating down to ±10V rails, and I'd like the output to swing as close to these rails as possible.

I modified the design to match the DAC; the transimpedance stage swings to ±9.1V and the OPA1622 has 0.5x gain to give a single-ended ±9.1V output. I would like more current than a single OPA1622 can provide so I have inserted an LME49600.

I am facing a simulation issue. The circuit behaves as designed if VCC / VEE are changed to 17V or greater. However, I'd like this circuit to operate on a lower voltage, ideally ±10V supply rails which can be cleanly derived from an unregulated 12V supply.

Could someone explain this behaviour? At ±16V the circuit should behave correctly but TINA-TI fails to run a transient simulation.

I would appreciate any suggestions or comments for a better headphone amplifier design. Is there a part like the LME49600 which will swing closer to the output rails? Could I instead parallel a second OPA1622? Or should I redesign the product with ±15V rails?

Attached a TINA-TI screenshot and a file. Running "Transient Analysis" it will fail, changing VCC and VEE to 17V will make the simulation work.

Max

AK4499-fail.TSC

5 months ago

0 Max Hunter 5 months ago

Prodigy 35 points

Hi, I have replaced the LME49600 with a BUF634 and the simulation appears stable. I suspect there might be a problem with the LME49600 (which appears as a LME49800) in TINA-TI.

However the BUF634 still swings nowhere near the output rails, which is a problem for this device. Is there another alternative suggestion so I can get 200+mA of current at 9V from a 10V rail?

0 Alec Saebeler 5 months ago in reply to Max Hunter

TI__Genius 17641 points

Hello Max,

I would recommend simulating with the BUF634A instead of the LME49600 or BUF634.

How much output swing do you require? The BUF634A is not a rail-to-rail output device, with headroom of 1.6-2.2V from supplies in the listed datasheet conditions.

Best,

Alec

0 Max Hunter 5 months ago in reply to Alec Saebeler

Prodigy 35 points

Hi Alec,

Thanks for your reply. I cannot find the BUF634A in the TINA-TI library, only the BUF634.

I would ideally like 9V peak (18V p-p), i.e. 1V from the voltage rails. I will settle for 16V p-p if needed.

It is a bit unclear how close the OPA1622 swings to the voltage rails. TIDU672C table 4 says it swings to 0.6V from the voltage rails (I have no idea where they found this figure). The datasheet suggests 0.9V at 600R load. However TINA-TI seems to simulate it as swinging 2.2V from each voltage rail maximum into the load shown.

Are there any downsides to paralleling two OPA1622 as below? What else would you recommend? I would love to drive 300mA at 18V p-p if possible.

Max

0 Alec Saebeler 5 months ago in reply to Max Hunter

TI__Genius 17641 points

Hello Max,

You can download the BUF634A spice model from the TI product page and use in TINA-TI. Typically the models on the TI website are more up-to-date than the models internal to TINA-TI, which could also explain some simulation issues.

I am looking into the OPA1622 model and its behavior for you. Have you considered the OPA1633 if you need to drive differential audio outputs?

I will also review your parallel OPA1622 as shared below your reply.

Thank you Max!

Best,

Alec

0 Max Hunter 5 months ago in reply to Alec Saebeler

Prodigy 35 points

Thank you Alec! I have come up with a new design using only OPA1622 op-amps as unity gain buffers for the output stage in a related question. I would love any feedback you have!

0 Alec Saebeler 5 months ago in reply to Max Hunter

TI__Genius 17641 points

Hello Max,

Great! I can review it or give another look if you would like.

Best,

Alec

0 Max Hunter 5 months ago in reply to Alec Saebeler

Prodigy 35 points

Sure thing - it's in a separate post for cleanliness! https://e2e.ti.com/support/audio-group/audio/f/audio-forum/1402723/opa1622-studio-headphone-amplifier-design-review

0 Alec Saebeler 5 months ago in reply to Max Hunter

TI__Genius 17641 points

Hello Max,

I think Chris covered the review side of things, I agree with his assessment.

I will mark this thread as resolved then, I appreciate your questions Max.

Best,

Alec

Audio

Audio forum

LME49600: OPA1622 headphone amplifier design review / TINA-TI instability at low rail voltages