• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 64 subscribers
  • Views 2927 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.

UC2842: How to realize Hiccup mode by using UC2842 in fly-back topology?

Sulyn ZHANG
Expert 8700 points
Part Number: UC2842

Hi Team,

My customer is using our UC2842 in old product, recently they need to improve the protection feature, they hope to realize Hiccup mode when load shorted. I'm thinking that when output was shorted, as the auxiliary transformer and the auxiliary winding polarity is same, so the auxiliary winding voltage Vaux will be also pulled low, then our UC2842 will be shut down and startup again, like "Hiccup behavior". If so, that is what customer wanted! However, the reality is cruel.... When we measure the output of auxiliary winding output voltage and Vaux, it seems the Vaux is not pulled down due to that these is leakage inductance, so it seems Vaux is not pulled down.

 

Below are two waveform with normal load status and 5V output load shorted to GND:

 Normal Load Waveform

 5V output Short waveform

When we add a 27 ohm series resistor+100pF cap to parallel with  D2, it seems the Vaux will be pulled low when 5V output short to GND, thus seems realize the Hiccup mode. However the standby current will increase about 2W which can't be accepted by customer. More detailed schematic as below for your check:

Protel Schematic_UC2842.pdf

So based on above information and description, can you or anyone give a available solution to realize Hiccup mode with our UC2842? The simple the solution the better, thanks a lot!

BTW, not sure how other customer realize the protection feature? Do you have something that can share with me, thanks a lot!

Best regards,

Sulyn Zhang

  • 0
    TI__Mastermind 32940 points
    Hello Sulyn,

    It is an interesting problem; the AUX voltage is actually a little higher during short-circuit conditions than when normal. I do not have an immediate answer, but I will think about this and discuss it with colleagues. I'll try to respond as soon as I can.

    Meanwhile, since this is an industry-wide forum, I invite other viewers, inside and outside of TI, who may already have a solution to offer their advice on this matter.

    Regards,
    Ulrich
  • 0
    TI__Mastermind 32940 points
    Hello Sulyn,

    I have a few thoughts about this, although not a definitive answer.
    It seems to me that 2W is a lot of additional standby power that the R-C snubber alone wouldn't account for. I suspect that putting the R-C across the AUX diode D2 allows it to briefly pull down the Vaux rail every switching cycle during normal mode and those perturbations might be triggering the start-up transistor T1 to briefly conduct each cycle. From high voltage, this might account for the extra power. I suggest to move the 27R-100pF snubber from D2 and apply it across the AUX winding (pins 4, 5) instead. Maybe that will snub the excess ringing, but D2 will block any drop on the Vaux rail and T1 will stay off. Your customer can test this idea 2 ways: put the R-C across the windings and put the R-C across D2 and disable T1, and observe the results of each test.

    Another idea I have is that during normal operation the AUX waveform shows ~2us on-time, and during short condition it shows ~1us on-time. This seems to be a long on-time for a protection condition. I suspect that the current sense R-C filter is delaying the peak signal too long and allowing a high primary peak current during the short condition. If this filter can be made smaller, the Ipk will be lower and maybe the leakage inductance ringing will be smaller, too. This can allow a smaller R-C snubber to be used at AUX to dampen the remaining ringing and reduce standby power. Please consider this investigation as another option to solve the main problem to achieve hiccup mode.

    Regards,
    Ulrich
  • 0 in reply to Ulrich Goerke
    TI__Expert 8700 points

    Hi Ulrich,

    Thanks a lot for your support, I will let Customer try this idea and feedback to you.

    Best regards,

    Sulyn Zhang