• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 62 subscribers
  • Views 138 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.

UCC28180: Square wave DC Input

WB
Intellectual 330 points
Part Number: UCC28180

Hello,

Could you please advise how the chip will behave if the input is a Square wave DC, switching between 0V and 400V at 400Hz? I am wondering if this controller is going to limit the inrush current to the bulk cap at every switching edge. If not, do you have any solutions you would suggest?

Thanks

WB   

  • 0
    TI__Mastermind 30900 points

    Hello WB,

    Please clarify the input waveform and your input structure for the UCC28180 circuit. 

    Is this the normal AC-input PFC converter with diode bridge rectifier up front, to which a DC square wave is applied into the bridge? 
    In that case, the square wave will also be rectified.

    That brings up the question of the square wave parameters.  Is this a square wave with amplitudes of +400V and -400V about 0V, at 100% duty cycle? 
    If so, then the rectified voltage will be a solid +400V all the time and will keep the bulk cap charged all the time without significant peak currents at the transition edges. 

    If the normal PFC output voltage is set for some voltage higher than 400V, then the UCC28180 will continuously boost the input to the higher output voltage. 
    There will be no inrush currents except any time that output voltage is lower than the input voltage.  That is true whether DC or AC input, and the controller cannot control this inrush at all.  In a non-isolated boost-PFC topology, no controller can limit the inrush. 

    If the output voltage regulation is set higher than the DC input, then again, there will be no inrush currents at any input switching edges (even if duty cycle is less than 100%) once the controller has boosted the output to above the maximum input. 

    It seems like limiting inrush current is your main concern.  Boosting Vout > Vin(peak) is the way to avoid repetitive inrush currents. 

    As for the controller performance with a 400Hz DC square wave input, if the duty cycle is significantly less than 100% (that is, significant 0V time between +/-400V levels), I'm not sure the controller will be able to follow the rising and falling voltage edges very closely.   I suspect that there may be some time delay and overshoot in the current on the rising edges of the rectified waveform as the current loop settles.  On falling edges, assuming the input goes to zero almost instantly, the current will also drop to zero just as instantly. 

    Regards,
    Ulrich  

  • 0 in reply to Ulrich Goerke
    Intellectual 330 points

    Hi Ulrich,

    Thank you for the feedback. 

    The Input is a square wave with amplitudes of +200V and -200V. The idea is to create a lossless active inrush current limiter by utilizing a standard PFC or Boost converter. The converter will need to support a synchronization signal to detect the rising edge and limit the peak current to a predefined value.  

  • 0 in reply to WB
    TI__Mastermind 30900 points

    Hi WB, 

    Thank you, your diagrams clarifies a lot of things. 

    First of all, you really don't need a PFC controller for this boost function.  Any general-purpose PWM controller (such as of the UCC28C4x or -C5x families, for example) configured for boost conversion can drive this converter.  

    Secondly, as before, to avoid peak current inrush on each rising 400V edge, Vout must be higher than Vin.
    So if possible, I suggest to regulate Vout to 435V or higher (for margin and load steps) to allow for the 30-V droop during the input deadtime. 

    If it is not possible for Vout > Vin + 30V, then only the inductor impedance is able to provide any inrush peak limiting.
    Adding a series resistor can limit more, but becomes lossy. 

    Adding a high-side (or low-side) series switch can limit the edge currents, but control of the switch becomes complicated.
    However, this would be the least lossy method for current control other than Vout > 435V (in my opinion). 
    A series switch can PWM the current using the boost inductor at the input rising edge, then change to continuous ON for the duration of the high input. 
    If the series switch can be synchronized with the boost FET, it may simplify the control of the PWM drive.
    The OR-in the continuous ON drive after a suitable time delay or peak current level. 

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Intellectual 330 points

    Hi Ulrich,

    • I forgot to mention that the output is 1KW, isolated, and should not exceed 400V. To account for a possible voltage drop, one option could be boosting above 400V and then using a two-switch forward converter to step it down to 400V. This approach is efficient but requires two stages.

    • The series inductor was on my list of potential solutions, but it poses some challenges: it would need to be custom-made, which could be expensive, and it would likely be large in size. I've calculated a required inductance of 2mH. Additionally, the inductor could form a tank circuit with the circuit capacitance, potentially leading to resonance issues.

    • I've been hesitant to modulate the rising edge, as this method tends to be noisy. The switch would need to operate at high speed, necessitating the consideration of a SiC FET to minimize losses. An LC filter would also be necessary. Furthermore, if the switch is placed on the high side, an isolated 12V rail would be required.

    • I am intrigued by your suggestion of synchronizing a series switch with the boost FET. Could you please elaborate on this idea and share a reference design if you have one?

    Thank you,

    WB

  • 0 in reply to WB
    TI__Mastermind 30900 points

    Hi WB, 

    On point 1, your boost diagram shows a non-isolated boost (same as a standard PFC), so to provide an isolated output you will need a second stage anyway.
    That will allow you to boost well above 400V and tightly regulate to a 400V isolated output, as you concluded, regardless of input and load variations.  

    On point 2, the series inductor I was referring to IS the boost inductor, already there.  No additional inductor. Every boost stage does have circuit capacitance on the switched node, but resonance can happen only if the inductor current becomes DCM.  No resonance issues while in CCM.

    On point 3, boost switching is already noisy, although another switch will probably just increase the noise during the brief intervals at the rising input edges when the series-switch is switching.  I concur, a small LC filter may be necessary to mitigate this. 

    On point 4, I don't have any or know of a reference design, I was just thinking off the top of my head. Assuming that the boost inductor current is low or zero at the start of the input 400V edge, then its current must be ramped up to the steady-state level (over most of the 400V high time) to avoid a significant current  overshoot.  This suggests to me that the series switch and the boost FET turn on at the same time, but the series switch turns off before the boost Fet does to keep control over how how peak inductor current grows.  Once the peak reaches the steady-state level, the series Fet stays on for the rest of the input half-cycle.  It may be possible that this can be an open-loop control, with series duty-cycle increasing at a fixed rate regardless of load.  I don't know for sure, but this possibility can be investigated. 

    A free-wheeling diode will be needed from inductor input to GND if a series switch is added. 

    Regards,
    Ulrich