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.

TPS259827LEVM: Parallel operation

Part Number: TPS259827LEVM
Other Parts Discussed in Thread: LM5069, TPS2491, TPS25947

Hi,

Due to high capacitive load I want to operate two TPS259827L in parallel as in the SLVA836 Application report.
The application report covers the TPS259241 and shows how it should be connected.
Can it be applied to the TPS259827L device as well and has it been tested?

Best regards,

Jonas Forssell

  • Hi Jonas,

    Theoretically they can be paralleled but sometimes due to overall path impedance mismatch it can happen one device taking up all inrush current. So we recommend using one device for startup purpose. Are you facing startup issue with one device?

    Regards

    Kunal Goel 

  • Hi Kunal,

    Thank you for the answer. Yes the circuit disconnects due to thermal switch off exactly as described in the datasheet, due to a large capacitive load and start-up time (we can't start slower 37 ms). Using the design tool indicates that 2 parallel devices would solve the problem. The eFuses are a good option because the design is integrated and with temperature monitoring. The other way to solve it is to use a controller and an external MOSFET. We have used the LM5069-2 but it has a drawback with a strong pull down of 230 mA on the gate under PORit (7.6 V). Since there is a sneak path from OUT through the gate protection diode (internally in the LM5069) to the pull-down current sink (230 mA). So it will have a high power loss internally if the device has charged capacitors on the load side or is connected to another supply on the load side. This could be a problem if it persists for a long time as it is 1.8 W and will heat up the controller and maybe cause reliability issues. The TPS2491 just sinks 2 mA in this case causing a minor loss (17 mW). Please advice, should we go for a discrete solution or stay with two eFuses? If we go with the discrete solution should we use the TPS2491 or LM5069?

    Best regards,

    Jonas Forssell

  • Hi Jonas,

    What is the voltage level eFuse VIN would see?

    Regards

    Kunal Goel

  • Hi Kunal,

    The voltage level is 18 V. The reason I asked about TPS259927 is because it's used in a different design on the board with higher voltage level so it was easy to reuse. But as i understand it by reading through datasheets and application notes there is a limited number of eFuses that are suitable for parallel operation, and that it should be specified if they can operate that way. Otherwise the function may not be guaranteed. Especially in this case when they must share the current during ramp-up.

    Best Regards,

    Jonas Forssell

  • Hi Jonas,

    We will check and get back if some other solution can work here.

    Regards

    Kunal Goel

  • Hi Jonas,

    Can you share details of max COUT cap and max steady state load current? Also you have to start within 37ms?

    Regards

    Kunal Goel

  • Cout max is 1300 uF and the load current is on average 0.5 A with a short peak at 2 A (20 ms duration) after ramp-up (due to start of subsequent loads). And yes the ramp-up is 37 ms.

  • Hi Jonas,

    Let me check and get back.

    Regards

    Kunal Goel

  • Hi Jonas,

    Single TPS25947 device can work here. It has ITIMER feature to allow transient of 2A. 

    Copy of TPS25947xx_Design_Calculator.xlsm

    Regards

    Kunal Goel

  • Hi Kunal,

    Thank you this looks promising. The TPS25947x has a smaller footprint and worse Rtja than TPS25982x but it has better transient thermal response in the 30 ms range. So this will probably work out fine. I will look at this solution (preferred) and a solution with TPS2491 and a discrete MOSFET as well.

    Best Regards,
    Jonas Forssell