• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 499 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.

UCC28056: UCC28056 Protection understanding

user5781628
Intellectual 970 points
Part Number: UCC28056

Hi,

I have some question about OCP protection.

In the datasheet, it describe OCP1 and OCP2. 

Both of these protection are signal cycle protection, right? It means that it will trigger OCP1 or OCP2 if trigger the threshold in one cycle.

The difference between OCP1 and OCP2 is that OCP1 only provent this PWM of this cycle. but OCP2 will cause 1s long time fault. Right after 3 times OCP2 triggering, right?

My question is that:

1. In the datasheet, it says "The Ocp2 comparator will only act if there is a gross fault such as a shorted output capacitor or bypass diode". Can you help explain why OCP2 will be triggered in these cases? What is the detailed logic?

In my understanding, even in this condition, only OCP1 will be triggered, OCP2 will not be triggered.

2. In the datasheet, it says "Under some fault conditions, including output overload, inductor current may become continuous because the reset voltage is low." 

In my understanding, the reset voltage(output voltage) will take long time to make the inductor current drop to 0A. Why it say it will enters into continuous mode?

3. Continue with the last question, it says "In this case even the short Ocp1 blanking time may allow the inductor current to continue ramping up." 

In my understanding, OCP1 will turn off this PWM cycle. Why it says the inductor current will continue ramping up?

4. I think the blanking time is only used for resolve the leading edge current spike. And also OCP1 is cycle by cycle protection. It looks that OCP2 will not be triggered forever. So it looks that OCP1 is enough, OCP2 is not needed. Can you help explain why there is OCP2 is configured in UCC28056?

Thanks!

BR

  • 0
    TI__Expert 6080 points

    Hello,

    Thank you for your interest in UCC28056, please find my comments below:

    1. As you have correctly stated that OCP faults are limiting cycle by cycle current. OCP2 fault is triggered when ZCD pin voltage exceeds 0.75V, and OCP1 when ZCD pin voltage is over 0.5V. If the overcurrent occurs when in regulation; OCP1 will trigger and the UCC28056 will limit the peak inductor current thus preventing OCP2 event from occurring. OCP2 fault will occur if there is a huge current as soon as the controller turnON event, this will happen if the output cap is shorted for example.
    2. The reset voltage is the voltage across the inductor which is the difference between the input and the output voltage when the output voltage approaches the input voltage this reset voltage is very low and not enough to limit the inductor current. This means that this current will keep rising to provide the output regulation. The blanking time of 450nsec from OCP1 may not allow the inductor current to drop enough to leave the fault state. This is why we need a second threshold fault state, OCP2 fault will trigger the long fault thus allowing longer period between two ON pulses for the inductor current to ramp down to zero. 

    I hope this answers your question.

    Regards,

    Sonal

  • 0 in reply to Sonal Singh
    Intellectual 970 points
    Hi Sonal,
    1. You say "OCP2 fault will occur if there is a huge current as soon as the controller turnON event, this will happen if the output cap is shorted for example."
    First question: if the output cap is shorted, why the controller turnON event?
    Seconda question: Even if the controller turnON event, the OCP1 should react firstly, so OCP2 will not be triggered. Why you say OCP2 fault will occure?

    2. For the reset voltage is very low, there should be more off time to demagnetize the inductor current. Why you mention the reset voltage for my question? I think there is no any relationship between them.

    Is there any other engineer who could answer my question?

    BR
  • 0 in reply to user5781628
    TI__Prodigy 750 points

    Hi,

    I am V.A and I was one of the architect for this device. I will try and answer your original question and we can go from there:

    Keep in mind that OCP1 has a 500ns blanking time and the OCP2 has a 50ns blanking time.

    Can you hit trigger OCP2 before you trigger OCP1?
    Yes! It is possible. There are some conditions which are possible:

    Under what scenarios?
    1. First there are safety tests to consider such as pin open/short test conditions. If the boost diode pins are shorted or the output, many customers will not accept MOSFET failure as acceptable test condition even though the FMEA requirements do not necessarily place this requirement. Under these conditions, the desired performance is that only the fuse can fail and no presence of fire and smoke shall be present.  Hence having a OCP response of 10x is beneficial for prevention of fire.

    2. Abnormal Conditions where VIN >= VOUT such as line surge,extreme output overload conditions (in which case, the VOUT stays at 1 Vf above the rectified voltage

    In case of line transients sudden increase causes forward conduction of boost diode and CCM conditions and in some cases saturation of the PFC inductor. The level of current during turn ON may be above both the OCP1 and the OCP2 threshold.  UCC28056 does not sense the inductor current and hence cannot detect the same unless the MOSFET turns ON.

    3. Inductor Saturation:
    During cases of excess case/operating temperature, persistent overload, there can be inductor saturation. Ferrite saturation can cause the inductance to drop to 10% of its nominal value. Hence a single OCP level with a 500ns time blanking time delay can be potential insufficient to protect the MOSFET.

    4. CCM Buildup due to Vin = Vout / startup into overload

    Once the PFC output capacitor charges, the Output voltage is the same as the input voltage. If the converter attempts to start in an overload condition, then the output voltage does not increase and stays close to VIN. Now, when MOSFET is ON, the voltage across the inductor is Vin(t). When the MOSFET is OFF, the voltage is [Vout - Vin(t)] which is just 0.7V ( Vf of the boost diode). Hence the current is virtually flat for this cycle. If this phenomenon repeats itself over 5-10 switching cycle, there is a significant current build up in the inductor potentially reaching beyond the OCP1 and OCP2 threshold. To prevent MOSFET operation under these conditions, OCP2 will trigger before OCP1 as OCP2 blanking is much smaller.

    These are the conditions under which OCP2 helps where OCP1 may be too late to protect the power supply,

    Hope it helps. Let me know if you have questions.