• State TI Thinks Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 2 answers
  • Subscribers 64 subscribers
  • Views 454 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.

UCC28950: Sometimes at startup, after 60..65 ms, the primary full bridge loses the "symmetrical block" current.

Koen Byls
Prodigy 10 points
Part Number: UCC28950
Other Parts Discussed in Thread: UCC28951

Tool/software:

Problem:
=> Does happen now and then (about 1 startup of 15)
=> Happens ALWAYS at about 60..65 ms after the on command.  The outputvoltage has already reached the nominal value of 75V.  The primary current looks stable, until 60..65 ms.
=> Not working in synchronous rectification (yet). 
=> The gate drive signals A, B, C, D look OK.  

What could be the reason of this misbehaviour ? 

See picture: Osc 20240618 13:28 Ch1 = Iprim shim; Ch2 = Gate_A; Ch3 = Gate_B; Ch4 = Gate_C; Ch5 = Gate_D; Ch6 = CRC_ON; Ch7 = Gate_E; Ch8 = 75V.

  • 0
    TI__Guru* 86186 points

    Hello,

    I am reviewing your inquiry and will get back to you shortly.

    Regards,

  • 0 in reply to Mike O'
    TI__Guru* 86186 points

    Hello,

    When cycle by cycle current limit is hit it will cause output C and D to look like like they have larger than 50% duty cycle.  However, what is happening is these outputs are phase shifting after cycle by cycle peak current limit is hitting.  I see this happening on your C and D pulses.

    To verify this you might want to study the CS, EAout, Voltage across the primary of the transformer and SS?  If the CS is hitting 2 V and the duty cycle across the transformer is decreasing you are hitting cycle by cycle current limit.

    If you remove the cycle by cycle current limit fault your  waveforms will look similar to what is expected.

    Regards,

  • 0 in reply to Mike O'
    Prodigy 10 points

    Hello Mike, 
    I did measure the CS signal.  
    At the moment where the strange behavior occurred, I only measure 0.424V peak.  Normal, since the converter runs at only 150W (of the 1600W max). The primary current at that point (Ch1 - yellow = Iprim) was "in regime", only 5.2A ptp.  


    Since this problem rarely occurs. I did suspect some noise crosstalk.  I did add a 10pF on the input of the input of the isolated MOSFET Gate Drivers.
    Now trying some startups under the same conditions, and see if the problem shows up.


    New measurement: Ch3 = voltage on the current sense line.  No problem occured... yet?

    Osc 20240619 14:42 Ch1 = Iprim shim; Ch2 = Gate_A; Ch3 = V PRIM-CUR-SENSE; Ch4 = Gate_C; Ch5 = Gate_D; Ch6 = CRC_ON; Ch7 = Gate_E; Ch8 = 75V.

  • 0 in reply to Koen Byls
    TI__Guru* 86186 points

    Hello,

    Is your scope probe capacitor on the CS pin helping with noise reduction?  Try adding a 22 pf capacitor on the CS pin to see if it helps.

    Most of these design will use a 1 k ohm 220 pF filter from the current sense resistor to the current sense pin.  Put the R and C as close to the CS pin and and UCC28950 ground as you can get it.

    Regards, 

  • 0 in reply to Mike O'
    Prodigy 10 points

    Hello, 
    I did measure the CS-line with a 1K resistor between the CS line and my probe.  This way of measurement did not influence the "problem".  It shows up with and without the probe. 

    Questions : 

    1. I see that the problem occurs at 51 .. 55 ms after the startup (=first pulses).

    Is there some kind of check in [the statemachine of] the UCC28950 which is activated after 50-60 ms ?  Which could give a lead in what way I have to search for a solution?

    In the measurements I did, I do not see any issue in the primary current.  The Gatedrive of the synchronous rectifier is not active, at the moment the problem occurs.  


    2. Do you have knowledge of other applications, which also have issues that occur 50..60 ms after startup?  

  • 0 in reply to Koen Byls
    Prodigy 10 points

    Additional: the information of the oscillograph above: 

    Osc 20240624 10:53 Ch1 = Iprim shim; Ch2 = CS-pin, measured over 1K; Ch3 = Gate_B; Ch4 = Gate_C; Ch5 = Gate_D; Ch6 = Gate_E; Ch7 = Gate_F; Ch8 = 75V.

  • 0 in reply to Koen Byls
    TI__Genius 12795 points

    Byls,

    Please give us some time to respond.

    Thanks

  • 0 in reply to Ning Tan
    Prodigy 10 points

    Hello, 
    Some additional info.
    1. It seems the problem occurs with all loads, always at 50..60 ms.  I did test it at 75V/1.88A; 75V/3.75A and 75V/11.25A.
    2. I am using the UCC28950.  Dutycycle is quiet large.  Can it make a difference if I should use the UCC28951?
    3. Can I contact a TI application engineer directly, so communication can be faster and more in detail. I could measure things which are suggested by the TI-application engineers.   

    Measurement 

    Osc 20240625 09:28:57 Ch1 = Iprim shim; Ch2 = Gate_A; Ch3 = Gate_B; Ch4 = Gate_C; Ch5 = Gate_D; Ch6 = V PtoA-AtoP; Ch7 = CS with 1K; Ch8 = Iprim transfo.


    Detail running before (50.2 ms) and detail on which the switching changes.(51.2 ms)



  • 0 in reply to Koen Byls
    TI__Genius 12795 points

    Byls,

    You are interacting with the application engineer directly. There is no middle man in between.

    Please leave your email here and we can contact you via email directly.

    BTW, you can also try UCC28951 which supports higher duty cycle.

    Ning

  • 0 in reply to Ning Tan
    Prodigy 10 points

    Hello, 

    Thanks. 
    Koen.Byls@barco.com

  • 0 in reply to Koen Byls
    TI__Genius 12795 points

    Byls,

    We will contact you shortly. Just to confirm, have you contacted our sales team?

    Thanks,

    Ning

  • 0 in reply to Ning Tan
    Prodigy 10 points

    No, I did not contact sales.
    Tried to find technical support.

  • 0 in reply to Koen Byls
    TI__Genius 12795 points

    Byls,

    OK, noted.

    Ning

  • 0 in reply to Ning Tan
    Prodigy 10 points

    Hello, 
    After a lot of searchwork, I came to the following.
    It seems that the startup of the UCC28950 causes some noise, which interferes with a primary 12V supply. This supply feeds the primary mosfets driver, and the voltage falls below 8V after 50..60 ms which disturbs the drive of the primary bridge.  
    Conclusion: no issue with the UCC28950.
    Thanks for the help and for guiding me to find the rootcause of this.

  • 0 in reply to Koen Byls
    TI__Genius 12795 points

    Glad to see the issue is resolved.