• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 31 subscribers
  • Views 990 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

OPA1656: Datasheet Fig.7, THD+N not rising with frequencY (load=2k and Vout= 3V(rms) )

Frank Brenner
Prodigy 100 points
Part Number: OPA1656
Other Parts Discussed in Thread: OPA1612, OPA828

Dear all,

aiming to keep THD+N figures constant over the audio frequency range.

As per datasheet Fig 7, OPA1656 achieves this pretty well with loads of 2k (or more) @ output voltage 3V(rms)

OPA1612 achieves this flatness of THD+N curve , eg in inverting mode G=-1 (Fig.7 of OPA1612 datasheet)

Audio crossover network: Overall noise figures may benefit from higher internal signal voltages (Douglas Self: "The Design of active Crossovers").

Running higher signal voltages through OPA1656 or 1612, for eg. Vout= 8V(rms).

These 8V(rms) might result in higher THD+N @ 10-20kHz range, the THD+N curve then rises towards higher frequencies ?

At what V(out) Fig.7 will OPA1656 / OPA1612 leave the nearly flat THD+N characteristic?

Is the THD+N curve strictly related to output current, so 3V / 2k load  >> 1,5mA (rms) keeps the curve "flat" , as does 6V / 4k load?



Thanks

Frank

  • 0
    TI__Expert 6400 points

    Hi Frank,

    I have not found any exact characterization data you expected so far.
    While, I suppose the flatness of THD+N vs. frequency in 10 - 20kHz might not change in 8Vrms output condition for OPA1612/56, considering with its output impedance curve and full-power bandwidth.
    If you need the exact apples-to-apples comparison data, we have to acquire the THD+N vs. frequency data with your expected condition, but it take some time to implement the measurement.

    Thank you and Best regards,
    Iwata Etsuji

  • 0 in reply to Iwata Etsuji
    Prodigy 100 points

    Hello Iwata,

    thank you for your answer. I m not sure if I understood everything to full extend:

    Fig 7 of the datasheet shows the "rising" THD+N curve as soon as the load is lowered to circa one third (2k >> 600R) .

    With the lowering of the load resistance from 2k to 600R, the output current rises by ca 3x and then, the THD+N curve shows this (unwanted) rising towards higher frequencies.

    If this curve "rising" is because of higher output current , wouldn t you expect the same effect when the output current is increased because of a higher output voltage (x3)  ?

    Best Regards

    Frank

  • 0 in reply to Frank Brenner
    TI__Expert 6400 points

    Hi Frank,

    Yes, as you pointed out, there is a risk to degrade in THD+N with the higher loading current, since, as you know, the THD+N is mainly dominated with the open-loop gain of opamps (Aol), i.e. the output impedance and the full-power bandwidth is highly depending on Aol. On the other hand, the gain(gm) of the opamp output stage (i.e. Aol is affected via it) is depending on its output load conditions, in general, the collector follower topology(e.g. RRO opamps such as OPA1656, OPA1612, etc.) is more affected with the output load condition than emitter follower topology (e.g. not RRO opamps such as OPA828).

    If you expected the flatness of THD+N vs. not only frequency but also the output loading current, the solution such as Fig.51 could help it although the total THD+N might be inferior to the single opamp solution...
    While, OPA828 could be an alternative device(as a single opamp solution) with respect to your expectation, but I have not found the exact evidence to satisfy your expectation.

    Anyway, at this moment, it is necessary to measure the THD+N flatness with respect to your application condition...

    Thank you and Best regards,
    Iwata Etsuji

  • 0 in reply to Frank Brenner
    TI__Genius 10850 points

    Hi Frank,

    We haven't heard from you in a while. We hope this was resolved. Please reply to re-open if you have any additional questions.