• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 63 subscribers
  • Views 799 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.

TPS7A4901 current limit at startup into highly capacitive load

Jan Volkering
Expert 8190 points

Hi,

I am using a TPS7A4901 and a TPS7A3001 in a +/-15V powersupply for an audio application.

The nominal load is around 120mA but the total capacity on the positive and negative 15V due to local opamp decoupling adds up to 600uF.

At startup the front side switcher generating +/-16V does not start up due to the higher inrush currents of the capacitive load.

I assumed the following designing this powersupply:

The LDO's will start up in current limiting thereby charging the 600uF capacitance until they reach their regulation voltage.

Is that assumption correct or do they have a different I-V characteristic when in current limit mode.

In this way the LDO's act as constant current sources charging the bulk caps to their nominal +-15V voltage levels but apparently is this limited current value still too large for the front side switcher to startup

One way to further limit the inrush current is to increase the softstart caps values on the LDO's. 

Is there a problem to increase the cap at the NR/CC pin to lets say 47nF or even 100nF?

  • 0
    TI__Guru 55940 points
    Hi Jan,

    Please note that the NR capacitor will change the startup waveform from linear to an RC charging curve.

    Another way to help provide the inrush current is to increase the input capacitors for the LDOs and delay the enable of the LDOs. The input capacitors will then help supply the inrush current.

    Very Respectfully,
    Ryan
  • 0 in reply to Ryan Eslinger
    TI__Expert 8190 points
    Ryan,
    If I would remove the NR cap the output will rise linearly right?
    Does that means that under a large capacitive load (600uF) the current limit kicks in and the output voltage will rise linearly with a 220-500mA constant current (TPS7A4901) until the buffer caps are charged and the output is in regulation?
    Putting a higher input cap at the LDO will not help much since then the front side DCDC will go into current limit on that big LDO input cap. I rather have the LDO do that.
  • 0 in reply to Jan Volkering
    TI__Guru 55940 points

    Hi Jan,

    Yes, without the NR cap, the output will rise linearly as shown in Figure 19 of the datasheet.  

    There is a short period before current limit will engage where the LDO will allow inrush current to charge the output capacitors.  After this delay, the current limit will engage.  At this point the current will be a constant 309mA (typ) until the capacitors are charged.

    Please note that current limit is designed to help protect the regulator during fault conditions and is not intended for extended use.  You may want to consider a higher current DCDC for a more reliable design.

    Very Respectfully,

    Ryan

  • 0 in reply to Ryan Eslinger
    TI__Expert 8190 points
    Ryan,
    My experience now is that for highly capacitive loads powered by post LDO's it is better to have some form of constant current limiting in place at startup when there is a DCDC front side switcher in the system. Or as you suggest to over dimension that front side DCDC. I want to have more control over inrush currents flowing so I prefer constant current limiter at startup. For single supply designs TI has plenty of solutions available but for a dual powersupply virtually none, at least none that support dual supply tracking during startup to make sure both supplies ramp up equally avoiding opamps to generate large output pulses at startup. The idea is to place that dual supply current limier between the front side DCDC and the post LDO's. This means that the input voltage to the LDO's ramp linearly with an adjustable ramp to acomodate the desired inrush current equal for both positive and negative supply.
    I came up with a circuit idea how to do that I could share off line with you.