• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 316 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.

Original question:

TPS2H160-Q1: Design questions

TPS2H160-Q1: Design Questions

d_zero
Genius 4735 points
Part Number: TPS2H160-Q1
Other Parts Discussed in Thread: TPS2H000-Q1, , TIDA-01323, LM74704-Q1, LM74701-Q1, LM7310, LM74700-Q1

Dear Team,

can you answer following questions:

1) Can we direct connect DIAG_EN to 3.3V without pull-up?

2) CL ( Current limit resistor) if i set it to 8.2k does it mean that CH1 has current limit of 245mA and CH2 has current limit of 245mA ? or is it 245mA for CH1+CH2

3) What error can i expect if i set the current limit to 245mA -> ±20%

4) Do i need to calculate RCS to take in account the upper error? 

Best Regards,

d.

  • 0
    TI__Mastermind 30255 points

    Hello D,

    1. It can- just keep in mind if you tie DIAG_EN high then the quiescent current will increase. 
    2. The current limit is set individually per channel. This means that 245mA can go through channel 1 and 245mA can go through channel 2 individually. 
    3. If your current limit resistor is 8.2k, assuming a 1% resistor tolerance:

    Technically 244mA is out of spec as we only spec the (nominal) current limit from 250mA, however it shouldn't be an issues as it is really close. Consider using the high-RON (and generally better priced) TPS2H000-Q1 if you need better accuracy at lower current limits.

    4) Yes the upper end of the accuracy should be accounted for- keep in mind during a fault condition (thermal, open load, etc.) the CS pin will be pulled high to VCS(H) which can be a max of 6.5V. If your ADC cannot handle this, it is recommended to have something something like a zener clamp or voltage divider to protect the MCU ADC.

    Best Regards,
    Tim

  • 0 in reply to Timothy Logan
    Genius 4735 points

    Hi Timothy,

    thank you for the replay.

    1) Regarding TPS2H000-Q1, it also has 15% current limit accuracy for 250mA if i'm not wrong?

    Also it 250mA if flowing through TPS2H000-Q1 Ch1 and Ch2 then max dissipation on TPS2H000-Q1 is 0.5A*1Ohm = 0.5W

    (while TPS2H160-Q1 has a dissipation of 80mW) , this is not small dissipation.

    Current sense accuracy TPS2H000-Q1 vs TPS2H160-Q1 is 4% vs 2.5% is that correct?

    2) My calculation for TPS2H160-Q1

    Current limit

    Vcl_th 0.8  
    Kcl 2500  
    Iout 0.25 A
    Rcl 8 kOmh
    Rcl_select 8.2 kOmh
    Rcl_s-1% 8.118 kOmh
    Rcl_s+1% 8.282 kOmh
    Iout_calc 243.9024 mA
    Iout_calc_+1% 241.4876 mA
    Iout_calc_-1% 246.3661 mA
    Iout_max 283.321 mA 15%
    Iout_min 205.2644 mA 15%

    Current sense

    Vcs_lin 4 V
    Vcs_high 6.5 V fault occurred
    Vcs 4 V
    Kcs 290  
    Iout 0.285 A
    Rcs 4.070175 kOhm
    Rcs_select 4.02 kOhm 1% tolerance
    Rcs_s-1% 3.9798 kOhm
    Rcs_s+1% 4.0602 kOhm
    Iout_calc 288.5572 mA
    Iout_calc_+1% 285.7002 mA
    Iout_calc_-1% 291.4719 mA
    Vcs_calc 4 V
    V_in_adc 1.207 V
    Rt 24.9 kOhm
    Rb 9.1 kOhm
    V_in_4V 1.070588 V
    V_in_6.5V 1.739706

    V

    ADC input can measure up to 1.207V but it is tolerant to 1.8V,

    is it ok to use resistor divider only without Zener so that the 6.5V is below 1.8V  as in calculation above?

    Do you have any other recommendation?

    Best Regards,

    d.

  • 0 in reply to d_zero
    TI__Mastermind 30255 points

    Hello,

    TPS2H160-Q1 has ±25% and TPS2H000-Q1 has ±20%. TPS2H000-Q1 can specify the current limit down to 50mA so it might be a better fit if you can get away with the increased RON- generally the higher-RON devices have a better price. No issue fitting the power profile you provided- it's just a matter if you can handle the 1ohm RON for the 2H000.

    The 2H000 has better accuracy for current sense at lower load currents:

    2H160:

    2H000:

    No issues using the voltage divider- this is a common design practice for the part. The calculations look fine. Looking at the schematic- the only comments I would have is to make sure there is some reverse current protection as described in 8.3.7.4 Reverse-Current Protection in the datasheet. Also, for EMI performance, the optimal input/output capacitor set is:

    Placing the 10 nF and 100 nF capacitors in parallel from Vs to IC_Ground has the greatest effect for EMC performance. Placing the 100 nF capacitor from Vout to System_Ground also helps, but isn't as significant. The capacitors need to be placed as close to the Vs and Vout pin as possible to prevent any additional inductances. Since you have no ground network, they are connected to the same ground.

    I saw the correct caps on VS, however did not see the VOUT side.

    Best Regards,
    Tim

  • 0 in reply to Timothy Logan
    Genius 4735 points

    Hi Timothy,

    we will use TPS2H160-Q1, because 1Ohm ron is to much handle in our situation.

    I think we don't need  8.3.7.4 Reverse-Current Protection. Wi will use TPS2H160-Q1 for FPDLINK output ( DESER side). 

    Like on the TIDA-01323 https://www.ti.com/tool/TIDA-01323 TPS4H160BQPWPRQ1 doesn't have nor metodh1 or method 2 for reverse current protection.

    Also commented here 

    https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1242825/tps2h160-q1-design-questions

    Can you please also add your comment if in this case is the diode on Vbat or gnd network needed?

    Regarding the caps on OUT1 OUT2 we use following approach like on TIDA, is this ok? or should be there more than 10uF 1210 ( 9uF eff when darted to 12V) ?

    I will ad additional 10nF to Vs

    Best Regards,

    d.

  • 0 in reply to d_zero
    TI__Mastermind 30255 points

    D,

    If there is any case for reverse current then either a diode on the input or ground resistor/diode network is required.  Usually for automotive applications there is a requirement for reverse battery that requires this protection. Many times there is a diode on the upstream voltage rail that would protect against this. Do you know if there are any significant inductances on the load? If so- then on the inductive turnoff there would be a negative voltage that might cause some reverse current.

    The TI-Design that you reference might not be optimized for EMI performance. You can change the bypass capacitors to be smaller- from our testing that we did on our device though we found the optimal component set to be what was referenced above. 

    Best Regards,
    Tim 

  • 0 in reply to Timothy Logan
    Genius 4735 points

    Hi Timothy,

    thank you for the replay.

    We use FPDLink for camera, only inductive part is the Coax cable and 2.2uH + 10uH inductor on the PoC filter of the camera, do you think this is a concern?

    Would you suggest to use Schottky diode RB161MM-20TR or an ideal diode like for this LM7310 or LM74704-Q1 or LM74701-Q1  purpose on the marked place?

    We use following capacitors for PoC

    Best Regards,

    d.

  • 0 in reply to d_zero
    TI__Mastermind 30255 points

    D,

    It looks like D2 from your diagram is acting as a freewheeling diode to handle the trace inductance, is this correct?

    A Schottky diode would work fine- I would recommend something like a B150-13-F which would be rated for 1A and could handle the load current. The RB161MM-20TR device would also work as long as there is no chance the 20V reverse voltage would break down the diode in the reverse direction.

    Just to make sure- but the cathode of the diode should be pointing to the VS pin of the high-side switch. Is this what your schematic is reflecting?

    Best Regards,
    Tim 

  • 0 in reply to Timothy Logan
    Genius 4735 points

    Hi Timothy,

    The D2 is  TVS not a standard diode , primary function for ESD and EFT protection, but it can also help in this situation form DS ?

    Yes the cathode of the Schottky is pointing towards Pin13,Pin14 of the TPSH160

    So you don't suggest to use ideal diode instead of Schottky?

    Best Regards,

    d.

  • +1 in reply to d_zero
    TI__Mastermind 30255 points

    d.

    An ideal diode could provide several advantages to the system and provide an extremely low voltage drop (around 20mV for LM74700-Q1), however the Schottky diode is a lot simpler/cheaper so if you can get away with the larger voltage drop the Schottky diode would be a fine. 

    Best Regards,
    Tim