• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 4 answers
  • Subscribers 67 subscribers
  • Views 463 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.

TPS62912: Filtering Noise and Isolating with Ground loop filter(TPS62912, TPS63710 and LDO there after)

Shyam Daram
Expert 1235 points
Part Number: TPS62912
Other Parts Discussed in Thread: OPA699

Hi,

I started my approach of Schemati design from below suggestions and arrived at a final schematic in slno

1. Ultra Low Noise Power Chain for Photo Sensitive applications - Power management forum - Power management - TI E2E support forums - Initial suggestions

2. Finalised Schematic Review  - Schematic Review of Power supply based on TPS62912RPUR,TPS7A9401DSC,TPS7A3301RGWT - Power management forum - Power management - TI E2E support forums

3. Finalised Negative Output Schematic - TPS63710: Ultra Low noise negative supply design - Power management forum - Power management - TI E2E support forums

Now that i have done my schemati, i want to take enough care in making sure the POWER GROUND Noise not entering in to opamp stages, i have maintained a seperate ground for Opamps called AGND

and its seperated with PGND palne through ground loop isolater as shown below

Now the BLM18PG121SN1 has a filtering curve of 

But effectively if see the curves of TPS62912 or 63710 both have noise dominantly in 100Hz to 100Khz

should i need to put beads which are having high impedance in the range of 100Hz to 100Khz, contarary to above bead ?

Kindly guide me in isolating the ground planes with proper ground loop filters to protect AGND

  • 0
    TI__Expert 7865 points

    Hello,

    An answer will be posted tomorrow

    Best,

    Calan

  • 0
    TI__Mastermind 18510 points

    I am not familiar your system design and not sure what's GND_CHASS in the schematic, but for 62912/63710 low noise device, actually what these device can do is reduce low frequency noise as you mentioned <100KHz, rather than  have noise dominantly in 100Hz to 100Khz

    For high-frequency noise immunity, ferrite bead still needed.

     Yuchang

  • 0 in reply to Yuchang Zhang
    Expert 1235 points

    , thanks for your response

    Above is how my system looks like, i have multiple sensor inputs coming from multiple photo diodes and they are passed through a TIA+Limiting amplifier+Comparator chain, to power above opamps i am using the power supply

    As shown in my initial post

    https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1230855/ultra-low-noise-power-chain-for-photo-sensitive-applications

    i face curves posted in my TIA and LA output which causes the degradation in SNR, for which its very much required to isolated DCDC from AGND of opamp, i am asking suggestions on how to decrease the noise which appears at my opamp output.

  • 0 in reply to Shyam Daram
    TI__Mastermind 18510 points

    Hi, first of all, since your complete system design is complicated, actually our suggestion is only for reference.

    so where is GND_CHASS used for? and what does TIA mean?

    The power converter on the same board with TIA/LA/Comparator, right? so your questions is the converter used for TIA/LA/Comparator but the TIA/LA may impact the quality for comparator, you just would make comparator a separated GND, right? if so, separate comparator GND with ferrite bead you posted should work, also suggest to parallel a 0ohm resistor(NC), just for backup, if no ground noise issue, stuff 0ohm, another concern is ferrite bead will cause I*R drop.

  • 0 in reply to Yuchang Zhang
    Expert 1235 points

    TIA is Transimpedance amplifier

    LA is limiting Amplifier

    The chasis ground is connected to mouting holes for external ground lug, AGND is TIA + LA + Comparator GND, and Power ground is for DCDC and LDO.

    I star connect these ground together.

    A 10nS pulse will come from photo diode, the current pulse will be converted to voltage via TIA and then OPA699 and comparator output is LVDS

    In this process, the Analog chain ground is completely maintained seprate from power ground, but still i see noise from power supply entering in to singal chain

    Yes the power converter is on same board.

    this time i have changed dcdc and ldo to reduce the noise, i want to ask help on same to , what care to be taken.

  • 0 in reply to Shyam Daram
    TI__Expert 8100 points

    Hi Shyam,

    My colleague will give you reply next week!

    BRs

    Zixu

  • 0 in reply to Shyam Daram
    TI__Mastermind 18510 points

    Shyam, our suggestion for your system GND connection is only for reference due to commonly we only common power converter related design rather complete system design.

    The old DCDC solution is not low noise/ripple solution, right?

    And 10ns pulse is the noise or signal? what's 10ns pulse frequency? Are you sure the noise is caused by power converter? if yes, low noise solution should be helpful, and separate GND as you mentioned will be better.

    My thought is below and just for reference

    PGND and AGND may be separated with Ferrite bead parallel 0ohm resistor firstly and then PGND connect Chasis GND(system case mounting hole) with Ferrite bead parallel high Frequency cap.

    Yuchang

  • 0 in reply to Yuchang Zhang
    Expert 1235 points

    Kindly accept my friend request, i shall share you my schematic.