• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 65 subscribers
  • Views 2336 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.

UCC28950 cycle by cycle current limit issue

Daniel Eindhoven
Prodigy 40 points
Other Parts Discussed in Thread: UCC28950

I have a problem with our design with the UCC28950 phase shift controller. We have the device in current mode and apply a setpoint on the err input with a potentiometer.

When we approach our maximum power level the cycle by cycle overcurrent protection steps in. It shuts down de gating of D for one cycle, but not on the other gates. This causes the transformer to saturate and the current to rise drastically. In the next few pulses everything goes wrong and is unstable. In some cases the mosfet's cant handle this current and fail.

I added a scope image of the phenomena, AB = orange, CD = blue, Gate D out of  the controller = green and purple is CS before the filter.

7317.TEK0000.TIF and 6138.TEK0003.TIF

It almost looks like this is a problem with the controller itself. I think it handles cycle by cycle overcurrent situations wrong. Can anyone help with/ elaborate this problem?

  • 0
    Prodigy 40 points

    Apparently I am not the only one who encountered this problem. I found a few topics with have a striking resemblance to this issue, but with different descriptions:

     http://e2e.ti.com/support/power_management/isolated_controllers/f/188/t/292928.aspx

    http://e2e.ti.com/support/power_management/isolated_controllers/f/188/t/237192.aspx

    http://e2e.ti.com/support/power_management/isolated_controllers/f/188/t/256207.aspx

     None of these topics describe a proper explanation or solution. Adding a capacitor in series with the transformer will lower the fault current and decrease recovery time after the fault. In our case this prevents the mosfets from exploding.  I guess the workaround is to set the current sense and slope compensation at a level where this protection scheme will never interfere. this does not sound like a proper solution.

  • 0 in reply to Daniel Eindhoven
    TI__Guru 73105 points

    Cycle by cycle current limit generally prevents transformer saturation.  I have worked with the UCC28950 EVM and have not seen this issue when the unit hits peak current limit.  http://www.ti.com/tool/ucc28950evm-442

    When I look at your current sense waveform there does not appear to be any turn on delay in your FET gate drive signals.  Your AB and CD delays might be set too low.  You might want to check the recommended minimum Rdelab/cd values. 

    If there is too much filtering on the CS pins or the slope compensation is too high, the converter will run in voltage mode control.  If you don’t have a DC blocking cap in voltage mode control the transformer will saturate also.

    Regards,

     

  • 0 in reply to Mike O'
    Prodigy 70 points

    Hi

    I want to control cycle by cycle current with UCC28950. I have seen slura08 application with UCC28950.

    This design can provide constant 54V and max 30A current.

    Is it possible to change the design current between 0A and 30A during operation?

    or

    Can you suggest I can do is there another system?

  • 0 in reply to Mike O'
    Genius 3295 points

    Hello Mike O',

    You say that if you run in voltage mode, the transformer will saturate. You also say the cycle by cycle current limiting prevents transformer saturation. I have been running a 1500 W design now, in voltage mode, without a blocking cap. I had one in there but took it out because the UCC28950 has current limiting even while operating in voltage mode. So why do you need a blocking cap in voltage mode if the IC has cycle by cycle current limiting?

  • 0 in reply to Chuck Sampson
    TI__Guru 73105 points
    Hello Chuck,

    A DC blocking capacitor is generally used to prevent transformer saturation when the design is being used in a voltage mode phase shifted full bridge converter.

    The peak current limit might save the design from being destroyed if the transformer saturates. However, it does not guarantee voltage second balance across the transformer during large signal input transients. Remember that that a phase shifted full bridge delivers energy with switch (A and D) or (B and C) are on at the same time. So what happens if the duty cycle between (A and D) and (C and D) changes between energy delivery cycles? The transformer should saturate.

    I have assigned this to another TI applications engineer to help you investigate this further if needed.


    Regards,

    Mike
  • 0 in reply to Mike O'
    Genius 3295 points

    Thanks Mike O'. I used a blocking cap in my last FBZVT design. I was hoping I could spare the extra cost with cycle by cycle limiting, but what you say makes sense. The cycle by cycle doesn't balance the volt seconds during start up, I forgot about that. Oh well.