• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 62 subscribers
  • Views 329 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.

TPS62243: Output Noise

Nikos Antoniou
Intellectual 820 points
Part Number: TPS62243
Other Parts Discussed in Thread: OPA316, TPS62740, TPS62840

Dear Sir/Madame,

I am looking for a dc/dc buck converter for my application. The converter should have the following features:

- Output current = 300mA

- Fixed output voltage = 1.8V

- Low power consumption

- Low output noise

I've chosen the TPS62243 and paragraph 8.3.2 says that in order to minimize the output noise then you have to select the PWM mode correct? But, how much is the output noise in terms of uVrms or mVrms?

I am going to supply the OPA316 Op-amp. Op-amp are sensitive to noise? Do you think that they can both combined together without any malfunction?

Do you have any suggestions for a low noise buck converter based on the demands above? 

Thanks in advance

Nick

  • 0
    TI__Expert 3645 points

    Hello,
    the selected device should work fine with the OPA316, as a reference of the output ripple you can have a look at figure 18 in the d/s. We don't have any output noise measure in the d/s. You can also consider the tps62740 if low Iq is a feature that you may need. I notify this thread to a specialist that supports the OPA316, I wait for his feedback too.

  • Former Member
    0 Former Member in reply to Alfonso Furio

    Hello Nick and Alfonso,

    The OPA316 is a nice selection for a low noise op amp in the 10MHz bandwidth range.  Am I correct in understanding that you will take the output of the buck converter and use it to supply the op amp?

    If so, then I should offer you some words of advice.  Any changes to the supply of the amplifier, such as a ripple from the output of the buck converter, will show up as an offset voltage for the op amp.  This offset will then be gained up by any gain in the feedback loop and appear at the output.  To determine the magnitude of this effect, consider the power supply rejection ratio (PSRR) plot for the OPA316, as shown below.

    Notice that, as the frequency of oscillation on the supply pin rises, the effect on the offset of the amplifier will be greater.  At the switching frequency of this buck converter, 2.25MHz, the amplifier will have about 36/37dB of PSRR.  Whether or not this is acceptable is up to you and your system specifications.  However, it may make the output appear too "noisy."  I would also suggest you put some margin into your design.

    We have put together a Texas Instruments Precision Labs presentation on this subject to help customers understand this effect.  Please consider viewing it or looking through the slides.  Feel free to reach out with any further questions.

    Regards,

    Daniel

  • 0 in reply to Former Member
    Intellectual 820 points

    Thank you for your answers guys!!

    Daniel, yes I will take the output of the buck converter to supply (Vcc) a non-inverting op-amp. The non-inverted input will be connected on a 0.4V Voltage Reference and the inverted input will read the value of some sensors. A 12bit ADC will read the output of the op-amp.

    So, ideally, the maximum acceptable output error should not exceed ADC's resolution. I didn't know that buck converter's ripple is seen as offset voltage and amplified...

    Also, now I am really thinking about the PSRR you mentioned before. As I'm new in this concept I am not sure how should I handle it.

    I am going to supply the system from a Li-Ion 3.6V battery. Do you recommend LDO instead of switching regulators? I'm not sure, but after some calculations, in order to step down 4.1V (fully charged battery) to 1.8V using LDO the waste of energy through heat was huge (44% efficiency @ 200mA)

    Is there any reference design or design considerations for supplying op-amps through switching regulator in order to keep output error (noise) as low as possible?

    Thanks in advance

  • Former Member
    0 Former Member in reply to Nikos Antoniou

    Hello Nick,

    Happy to help.  The general rule of thumb is you'll want to keep all your possible error sources going into the ADC at less than 1/2 of the least significant bit.  The ripple will need to be contained to this magnitude and you should also consider any possible errors from op amp and component noise, op amp offset, bias current, etc.

    Once you have your design a little more finalized, you can open a new thread here on e2e and have one of the op amp or ADC applications engineers help you with your design.

    I'm not an expert on power design, but it does seem that going from 4.1V to 1.8V on an LDO would be energy inefficient.  I'll let Alfonso give his opinion on this matter.

    Regards,

    Daniel

  • 0 in reply to Former Member
    TI__Expert 3645 points

    Hello,
    the tps62243 should work fine with this OPA. The output ripple is less than 10mV and considering the PSRR of the op-amp, roughly you can consider that is attenuated at a value around 100uV. You should consider if this value is suitable for your final application. With the tps62243, you will obtain higher efficiency and if required from your application, you can add after the LDO. You can go in an example from 4.1V to 2.2V and then with an LDO go from 2.2V to 1.8V. Before trying this solution with the switching regulator+ LDO, I invite you to test before our EVM with the OPA316.

  • 0 in reply to Alfonso Furio
    Intellectual 820 points

    Thank you for your suggestion guys!! For sure I will proceed with Evaluation Boards first and I will consider the switching regulator+ LDO topology after that. I have one more question regarding the switching regulator.

    As I've mentioned, this is going to be a battery-powered application and Daniel suggested the tps62740 as well. This specific application will be designed to work continuously (it is required to read the sensors nonstop for around 3-4h) and therefore it will never go into shut down mode and the load will be always enabled.

    For the tps62243 I can see that Iq for PWM mode 3.8mA (quite high.. If I have a constant load it going to consumes 3.8mA constantly?) while for tps62740 I cannot find such information (I find only that Iq is 12.5uA for enabled load but I guess that this is for the PFM mode).

    Do you have any suggestions of choosing the less power-hungry converter regarding the above information?

    Thanks in advance

  • 0 in reply to Nikos Antoniou
    TI__Expert 3645 points

    Hello,

    in this case the Iq is important for your application and tps62740 should fit better your requirements. You can have also a look at tps62840, it has higher output current and it is the best in the market in terms of Iq.