• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 64 subscribers
  • Views 583 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.

Half-bridge using LM5033 - CS pin current limit

Josh Deng
Prodigy 30 points
Other Parts Discussed in Thread: LM5033

Hi all,

I’m updating an existing half-bridge design that uses the LM5033 in order to implement the CS pin (which was previously unused and grounded). I’m using the CS pin short-circuit protection, and I’ve set the CS pin so that it trips when current through a sense resistor is ~3.85A. Our design is for 160V minimum Vin, and up to 200W out at 27V out; our normal operating conditions should not draw more than 1.4A rms input current.

The problem I’m having is that while short-circuit protection works great, the downside is that current limit also trips during turn-on, when the output capacitors of the converter charge up. (Our output capacitance is 5x 6.8uF, but we spec up to 250uF external capacitance as well.) I see the CS current limit cycle at least 20 times before Vout slowly charges up enough not to trip current limit; then Vout rises as expected.

Does anyone know how to implement current limit on this controller, while preventing it from also tripping during turn-on? Is my current limit set too low? My concern is that if I set current limit higher, my half-bridge will become unbalanced during short-circuit before the current limit kicks in.

Any help or feedback is appreciated!

  • 0
    TI__Expert 3545 points

    Josh,

    Can you give me a little more information, current sense resistor value and transformer turns ratio.

    Terry

  • 0 in reply to Terry Allinder
    Prodigy 30 points

    Hi Terry,

    The current sense resistor is a 0.33ohm; the transformer is 8:2 turns. I have a voltage divider from the current sense resistor (multiplies Vsense by 0.394), along with a blanking cap. Let me know if you need anything else.

    thanks,
    Josh 

  • 0 in reply to Josh Deng
    TI__Expert 3545 points

    Josh,

    I normally set the current limit set point at 120% of my load.  When the power supply is starting up, it is running open loop and it is not un-common for it to start-up in pulse-by-pulse current limit until the output capacitors are charged up and the output voltage get close to its set point (in your case 27V).  You can control the peak current by adjusting the LM5033 soft-start capacitor.

    Terry

  • 0 in reply to Terry Allinder
    Prodigy 30 points

    Thanks Terry. It actually seems that the biggest input current spikes are caused by the first few switching cycles, where the output cap is charging, and it seems like this is regardless of how long my soft start is. I set it with a 1uF cap (which puts turn on at a very long 30ms), and it still goes into current limit.

    I am ok with a few cycles of current limit, but 20+ cycles will delay my turn-on time too much (we have a spec we're aiming for). I'm continuing to try to lower these current spikes (or delay the current limit tripping), but I have to wonder, is this inherent in the controller design? If instead the LM5033 used a true "hiccup mode" current limit, would it dissipate too much power, or imbalance my half-bridge?

    Thanks,
    Josh 

  • 0 in reply to Josh Deng
    TI__Mastermind 25130 points

    Josh,

    The converter maxamum alllowed duty cycle will be a function of the SS voltage and the internal ramp. During startup the voltage on the output is near zero so any pulses across the output inductor will be the voltage on the output of the transformer, so the di/dt in the output inductor will be high. Since there is no voltage on Vout the current through the output inductor wil not deminish when the duty cycle turns the FETs off. This means that at the next pulse the output current will continue to increase further as the output capacitor voltage is increased. Within a few cycles, the output inductor current will reach the current limit point and that is where it will stay as the converter trips off with the peak current limit.

    With hard spec numbers that require that much output capacitance and a short turn on time, you will have to verify that your magnetics and semiconductors can handle a higher current and then change the limit on your current sensor circuit to allow for the higher levels during startup. The higher current limit will have to be designed so that the current needed to charge the output capacitor in slightly less than the time required by the specification.

    Once teh converter is started you can revert the current limits to the levels needed for you star=ndard load.

    You will probably want to set the on chip SS to a low value (near zero) and use a secondary side softstart to allow the controlled rate of rise on the output but the first thing is to make certain that the power components can handle the short term transient high current needed for the startup. Thermally you can still design for your nominal maximum load.

    Regards, 

    John