• State TI Thinks Resolved
  • Locked Locked
  • Replies 14 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 6209 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

UCC28180D- PFC controller IC

Smitha R
Intellectual 790 points
Other Parts Discussed in Thread: UCC28180

We are using UCC28180D IC in our power supply design.  when our board is powered on from programmable AC source(in which we can limit input voltage and current) our power supply design works as expected but if the same board is powered on from raw AC mains(230V ac, 50Hz) we see a blast on current sense resistor and MOSFET  is going bad. One of the fuses on main line is blown. Same is observed on other 3 working boards.  Kindly request you to let us know what could be going wrong. when we captured the inrush current during power on, it measured around 50A.  Schematic is inserted.  

Awaiting for your answers!!!

  • 0
    TI__Mastermind 25570 points

    Hello Smitha

    I suggest you add a small signal diode from the ISENSE pin to GND to limit the voltage at that pin. Any small signal diode will work. Do not use a Schottky type device because the Vf will be too low. The normal operating range of this pin is from 0 to -438mV so the diode will have no effect on normal operation. Below is a schematic - please re-post if the schematic is not visible.

    Please let me know how you get on.

    Regards

    Colin

  • 0 in reply to Colin Gillmor
    Intellectual 790 points

    Hi Colin,

    Thank you for the quick response.  We will surely try out your suggestion but since we have already risked 4 working boards, we would request you to advice us for the below queries.

    1. what could make this Isense resistor to blast when the board gets powered from raw AC mains? what would be the root cause/possible reasons?

    2. How could adding diode would help in solving above? will there be any negative glitches during the power on?

    kindly request you to clarify.

    Thanks,

    Smitha R

  • 0 in reply to Smitha R
    TI__Mastermind 25570 points
    Hi Smitha

    The problem can occur because the UCC28180 can run at max duty cycle if the ISENSE pin is more than about -1V negative. Significant currents flow and this overloads the current sense resistors.

    The inrush current from a raw AC mains can be enough to cause the ISENSE pin to go more negative than -1V - it depends on things like the phase angle at re-connection, the line source impedance and the residual charge in the bulk capacitor. The inrush current from an AC source will most likely be less than that from a raw AC line. They normally turn on at 0 phase angle, where the instantaneous voltage is close to zero. Then the ISENSE pin won't go as negative and the problem doesn't happen.

    The added diode prevents the ISENSE pin from going so negative and the UCC28180 will operate correctly.

    Regards
    Colin
  • 0 in reply to Colin Gillmor
    Intellectual 790 points

    Hi Colin,

    Datasheet says Soft overcurrent limits the input current when it senses the threshold of -0.285V(at ISENSE pin) and Peak current limit operates when threshold reaches -0.4V(at ISENSE pin) terminating the active switch cycle.  But in our case why these two protection funtions are not getting enabled ? If inrush current could make the voltage at ISENSE pin as -1V which is more than the mentioned thresholds, then peak current limit/ soft start control should have been terminated /reduced PWM duty cycle. 

    Kindly request you to advise. Thank you for the quick support.

  • 0 in reply to Smitha R
    TI__Mastermind 25570 points
    Hi Smitha
    You are correct and of course, the soft overcurrent and the peak current limit functions do work under normal running conditions. However, if the UCC28180 bias voltage at VCC is already present when the line is connected, the controller may start before the inrush current event has passed. If the voltage at the ISENSE pin is greater than -1V at this time, the Soft Overcurrent and Peak current functions do not work and you get the failure you reported.

    Normally the bias supply for the UCC28180 will be generated from a supply which is also powered from the ac line - and the VCC rail will not be established until after the inrush current event is over.

    The added diode is our recommended solution for this behaviour.

    Regards
    Colin
  • 0 in reply to Colin Gillmor
    Intellectual 790 points

    Hi Colin,

    So it there any possibility of a bug in this IC? Looking at the problem statement- ISENSE pin should work fine in range of +7 to -24V which we are not voilating, though there is excess voltage due to inrush.

    One more doubt, looking at your answer- do you mean that the controller should start (VCC should appear) only after insrush current has passed? In case, if residual VCC is present (becuase of no discharge path) after switching AC mains off, the controller may still be working. So shouldn't it turn main FET off instead of still giving pulses? 

    just to add In our design VCC comes from AC source (bias winding of transformer) as you mentioned, which has a long ramp down time. 

    We appreciate your support. Thanks a lot!

  • 0 in reply to Smitha R
    TI__Mastermind 25570 points

    Hello Smitha

    You are of course correct that you are not violating the Abs Max ratings of this device.

    The controller assumes that VCC comes up after the line voltage is connected and that the controller is not active until after the inrush event. It is certainly possible that a short interruption of Vline can be enough to allow the PFC stage output capacitance to discharge, causing a large inrush current when line is re-connected yet for VCC to remain above the UVLO level of the controller. If the inrush causes a negative voltage greater than -1.1V then there is the possibility for the controller to operate at Dmax - leading to the failures you experienced.

    Regards

    Colin

  • 0 in reply to Colin Gillmor
    Intellectual 790 points

    We tried by connecting the diode as advised by you.  The issue seems unresolved as fuse blew and sense resistor went bad with a blast again.

    Please let us know if you can guide us further in resolving this issue. If there is any case where others have experienced similar problems please share the details.


    Below are snapshots taken when board was powered from programmable AC source.

    [ 2nd channel(blue) -     voltage across diode (at ISENSE pin)

    1st channel(yellow)-     voltage across Rsense resistor

    3rd channel(pink) -         Vcc for PFC IC

    4th channel(green) -      Main current (Line) at power entry connector]

    1. Without diode

    2. With general purpose diode


    Awaiting for your support. 

    Thanks,

    Smitha R

  • 0 in reply to Smitha R
    TI__Mastermind 25570 points
    Hello Mr Smitha
    I'm afraid that the images didn't come through. I've emailed you a reply just now to what I think is your correct email address - please email me at colingillmor@ti.com if it didn't come through.
    Regards
    Colin
  • 0 in reply to Colin Gillmor
    Intellectual 790 points

    Hi,

    As mentioned in earlier posts, we are using UCC28180D in our power supply design and 100KHz is the switching frequency.  When we conducted Conducted Emission test, emission levels are as shown in image(150K Hz to 30MHz-CISPR 11 CLASS A).   Emissions that are crossing limits are multiples of PFC switching frequency.  Kindly request you to advise the recommended modifications to bring the emission levels low.

    Schematic of input side of power supply is shown below.

    Awaiting for your support!

    Thanks,

    Smitha R

  • 0 in reply to Smitha R
    TI__Mastermind 25570 points
    Hello Again Smitha

    Curing an EMI issue can be a tedious and time consuming business but there are some general guidelines you can follow. The low frequency noise - below about 10MHz - is mainly Differential Mode noise, the higher frequency noise - (10MHz to 30MHz) is likely to be mainly Common Mode.
    Do you have a second stage loading the PFC – for example a LLC – because these stages can also generate noise that appears in the EMI scans.

    First, Ground any heatsinks appropriately. Heatsinks for the PFC MOSFET and Diode should be grounded to PGND in your schematic. This will give a low impedance return path for noise currents injected into the heatsinks from the MOSFET and Diode tabs.

    Second. Check that the connections to the input capacitor, C2 are short, and especially that the loop C1, PFC choke, MOSFET/Diode, Bulk Capacitor (Cout), and back through R8 are as short and direct as possible.

    Third. Dismantle the existing filter arrangement completely (everything to the left of the rectifier bridge) and rebuild it in a straight line. Take care to minimise stray inductance in series with the X and Y capacitors and stray capacitance from input to output of the common mode choke. This will give you an indication of the basic performance of the filter components and whether the PCB layout is reducing the filter performance.

    Fourth. Check that the switching waveforms are ‘clean’. Any spiking on the switched node at the MOSFET drain will increase the signal going into the EMI filter.
    This is just general advice – if you send me your PCB layout I may be able to offer more informed advice.
    Regards
    Colin
  • 0 in reply to Smitha R
    Prodigy 50 points

    Is this possibly due to running discontinuous current mode?  Anyone reading this with experience using this IC other than CCM?

  • 0 in reply to Keith Snook
    TI__Intellectual 2910 points
    Keith,

    This thread is over a year old. If you have a specific issue we can support, please start a new thread. That approach will ensure that you receive timely feedback.

    - Daniel
  • 0 in reply to Smitha R
    Intellectual 790 points

    HI all,

    As mentioned in the earlier post, we were able to bring down the emission levels to meet only CE/RE class A limits.  but  our customer has asked to make our power supply design to pass for Class B limits.

    We are currently using the UCC28180D IC as our PFC controller and we found the emissions are more from the PFC circuit.

    We have referred TI applications notes and evaluation board and found that the evaluation board supports only class A requirement.

    Do you have any better suggestions to meet the above requirement? Please let me know ASAP.