• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 29 subscribers
  • Views 296 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPA6120A2: Rfeedback vs phase margin

Andrew Russell
Prodigy 10 points
Part Number: TPA6120A2

Tool/software:

The TPA6120 oscillates easily if trying to drive headphones directly. The fix per the app note is to insert a resistor in series with the output of not less than 10 Ohms but in the data sheet reference circuits shown as 32 Ohms.  On large diaphragm dynamic cans, this is not good because you really want to drive them with a low source resistance.  Since this is a CFA, can Rfeedback not be set to a much higher value than the 1.5k shown in the data sheet with a concomitant change in Rgain to get more phase margin, and then allow (nearly) direct drive of the headphones using say a small series output inductor + parallel damping resistor? The trade-off would be lower closed loop bandwidth and slower rise/fall times but its worsth it to get the low output impedance drive. A simple formula for this, or some phase margin plots vs Rfeedback would be a welcome addition to the data sheet, or an app note. Assume worst case capacitive loads of <2 nF, but more typically <500pF.

  • 0
    TI__Prodigy 380 points

    Hi Andrew,

    That's a great question. There is not much downside to increasing Rf besides the decrease in bandwidth that you mentioned. There are some increases in thermal noise, but I don't think they would compare to other noise sources. If you can increase Rf while meeting your BW requirements, I would do that. Just don't make it crazy big (like 1MOhm) because then you can go back into instability, I think because it creates a pole with the input capacitance.

    I ran a simulation with a guess of what your system may look like. With a gain of 3, I was able to achieve stability at Rf = 6k.

    If your system doesn't allow you to increase Rf by much, your idea of the parallel resistor and series inductor is good too! I simulated that, and reached stability with Rf = 3k, L = 10u, and R = 50. That's just an example, I didn't do enough testing to get a threshold.

    I wanted to mention that when you choose a resistor for that, make sure to consider this in your total load resistance. Reference the data sheet to make sure your output voltage and distortion are still where you want them to be. 

    I'm not sure what your requirements are so I can't give you numerical advice, but hopefully that helps

    As for an easy way to see the phase margin vs Rf, I don't know of a really simple graph. You can however estimate it from looking at the peaking of the frequency response over different Rfs. TI precision labs has a video about this under Stability --> Measuring System Stability.

    If you want to simulate it the long way, it's not too hard. There's another video about this under Current-feedback amplifiers --> Current feedback amplifiers - Spice simulation

    Best,

    Hypatia