• Resolved

TPS51200: Inrush Current on VLDOIN & VO pins

Genius 5385 points

Replies: 13

Views: 234

Part Number: TPS51200

Dear, Sir.

My question is related to the contents of this thread;

e2e.ti.com/.../767850

My customer apply TPS51200 for DDR3. SO, they was facing the inrush

input current on VLDOIN & outout on VO pins. They apply TPS51200 for DDR3.

They tried to add 1 ohm resistor between VLDOIN and VREF pin, then

the input peak was reduced from 2.75A to 0.688A.

So, I would like to confirm followings;

1. Is it no problem on this kind of usage for TPS51200?

2. If no problem, How much voltage drop should be estimated on VLDOIN

   pin?

3. Personally, I think Trcking Start-up and shutdown is not so good for

   DDR. Because DDR has the initialization time inside.

   S3 and pseudo S5 is better to escape from unknown VTT setting

   inside of DDR. How do you think about my thinking?

I an hoping to get your advice.

Best Regards,

H. Sakai

  • Hi Sakai, 

    1. In typical DDR application, TPS51200 VLDOIN pin is connected to VDDQ, REFIN is connected to VDDQ via resistor divider. So there is no problem to add 1ohm in the circuit. 

    2. What voltage drop? Because VLDOIN pin is connected to VDDQ, the VLODIN pin voltage should be VDDQ voltage. 

    3. It depends on DDR requirement. I would suggest to check with DDR  vendor. If non-tracking startup works for DDR, then it is fine to use non-tracking startup. 

    Thanks

    Qian

  • In reply to Qian Chen:

    Dear, Qian-san. 

    Thank you so much for your advice as; 

    1. In typical DDR application, TPS51200 VLDOIN pin is connected to VDDQ, REFIN is connected to VDDQ via resistor divider. So there is no problem to add 1ohm in the circuit. 

    2. What voltage drop? Because VLDOIN pin is connected to VDDQ, the VLODIN pin voltage should be VDDQ voltage. 

    I would like to double-confirm. 

    No impact for REFIN voltage to add 1ohm resistor? Right?

    Best Regards, 

    H. Sakai

  • In reply to hideyuki sakai:

    Sakai-san,

    VLDOIN pin should be connected to VDDQ directly, because VLDOIN current is large. 

    R1 in your figure can be connected to VDDQ via 1ohm resistor, because the current going through 1ohm resistor is very small. 

    Thanks

    Qian

  • In reply to Qian Chen:

    Dear, Qian-san. 

    Thank you so much for all of your advice & sorry to disturb your jon again and again. 

    I have confirmed the real schematic of the customer. Sorry my mistake on previous schematic. 

    They connect VLDOIN to VDDQ through 1okm resistor. The voltage drop by load current

    at least will be happened. They said it is effective to reduce the input current peak. 

    Is it no problem to take this kind of circuit? 

    Sorry again!!

    Best Regards, 

    H. Sakai

  • In reply to hideyuki sakai:

    Sakai-san, 

    What's the concern of 2.75A VLDOIN inrush current? 

    Will it cause OC on VDDQ rail?

    What's maximum VTT DC current?

    Thanks

    Qian

  • In reply to Qian Chen:

    Dear, Qian-san. 

    1A current is the spec limit of 1.5V LDO as the voltage source and they don't measure

    VTT outputcurrent. There are just input current. 

    Below is the before to add 1ohm resister between REFIN and VLDOIN. 

    Also, below is after to add 1ohm resister between REFIN and VLDOIN. 

    VLDOIN voltage during inrush current flow will be 0.182V(1.5 - 1*0.688). 

    I would like to make sure this kind of usage is acceptable or not. 

    If acceptable, I would like to know the acceptable voltage range of dropped down VLDOIN. 

    Best Regards, 

    H. Sakai

  • In reply to hideyuki sakai:

    Sakai, 

    There are two concerns with 1ohm resistor between VDDQ and VLDOIN. 

    1. the power loss on 1ohm resistor

    2. With the voltage drop on the 1ohm resistor, the VLDOIN voltage may too low. 

    With VLDOIN shorted to VDDQ, can you turn on VDDQ and 3.3Vin first, then turn on EN pin to see if the inrush current is lower. 

    Thanks

    Qian

  • In reply to Qian Chen:

    Dear, Qian-san. 

    Thank you so much for all of your advice. 

    They tried to With VLDOIN shorted to VDDQ, turn on VDDQ and 3.3Vin first, then turn on EN pin to 

    utilize /RESET signal of DDR3, but it is not successful because the inrush is still higher. 

    So, I would like to reconsider the method of adding 1 ohm resistor VDDQ & VLDOIN. 

    You pointed out 2 items. 

    1. the power loss on 1ohm resistor

    2. With the voltage drop on the 1ohm resistor, the VLDOIN voltage may too low. 

    They will put more higher power dissipation 1 ohm resistor, #1 could be cleared. 

    Regrading #2, How much minimum voltage must be maintained on VLDOIN for VO=0.75V? 

    Sorry again& again, but Please give your advice. 

    Best Regards, 

    H. Sakai

  • In reply to hideyuki sakai:

    HI Sakai, 

    Before we look into the impact of voltage drop on 1ohm resistor, can you ask customer to use input filter inductor instead of 1ohm resistor, and check if the initial inrush current is reduced?

     The inductor should help to limit the inrush current. Also the voltage drop on the inductor during stead state is very small. 

    Thanks

    Qian

  • In reply to Qian Chen:

    Dear, Qian-san. 

    Thank you so much for your advice of the inductor solution instead of 1 ohm resistor. 

    I would like to make sure following. 

    1) I think the inductor solution would be expected the peak current capability of the inductor. 

        I wonder the inductance value would be higher to set lower peak current? 

    2) Please introduce more suitable product to reduce the inrush current instead of

       TPS51200 if you have. 

    Best Regards, 

    H. Sakai