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.

WEBENCH® Tools/UCC28630: UCC28630DR

Part Number: UCC28630

Tool/software: WEBENCH® Design Tools

We have developed the AC-DC Adapter circuit using UCC28630DR for 160V to 260V input AC supply working upto 100 watts. We faced the following problem during testing of the circuit.

The AC input voltage was set to 170V AC and unit was turned ON, we were able to get DC voltage.Capacitor C2 output around 280V DC, this is fed to primary points of transformer. The supply point VDD of IC was13V initially when we switch on the circuit and voltage varies in between 13V to 8V later. But no voltage was observed across secondary and auxiliary points of transformer. At this stage R8 resistor valve was 6.49 Kilo ohms (Rbld).  We decreased R8 resistor valve to 3.75 Kilo ohms (Rbld) but still no voltage was observed across secondary and auxiliary points of transformer.  VDD of IC was13V initially when we switch on the circuit  voltage varies in between 13V to 8V. We request you to advice us to solve the issue.     

  • Hello Jayaprakash,

    Thank you for your interest in the UCC28630 flyback controller.

    There is insufficient information here to be able to help you.  Can you please provide a schematic diagram of your converter?  I could find no reference to "Rbld" in the datasheet and "R8" or "C2" provide no reference as to their function.

    Your description of VDD cycling between 8V and 13V is indicative that the controller is not satisfied that all conditions are suitable to begin switching.  The controller monitors several inputs at the peak of the 13V to determine that there are no fault conditions and that switching may commence.  If one condition is not suitable, it will allow VDD to decay to 8V and then restart the cycling.  If you see 3 gate drive pulses at the 13-V peak each time, then it is NOT a latch-off fault.  If you see no gate drive pulses at all, then it is a latching fault.  That will help narrow down the problem possibilities.  

    Please check the fault protections listed in Tables 3 and 4 and some timings of Table 6 (in the UCc28630 datasheet) to determine which start-up condition is not being satisfied.

    Regards,

    Ulrich

  • The schematic of the ac/dc adapter is attached.

  • Hello Jayaprakash,

    Thank you for the schematic diagram.  I see that R8 is minimum load resistor.  Changing that value will not affect start-up of this converter. 

    I see nothing obviously wrong with the diagram, except that I assume the value of T1 is 269uH, not 269mH, right?
    Please provide the turns ratios of T1.

    As I mentioned in my first reply, the controller is not starting because it is detecting some fault condition.  You'll need to determine whether it is a latching fault or a non-latching fault, based on the presence or absence of pulses at each peak of VDD.  Then consult the fault tables to see which of the fault conditions (latching vs. non-latching) might be existing to prevent start-up.   Then systematically investigate each potential fault condition to see if your circuit matches that condition.  Then correct those situations that do, so that the controller sees no faults prior to start-up.

    Regards,
    Ulrich  

  • We are trouble shooting the pcb as per your guidance.

    The transformer details  is attached. As wurth electronics was not responding to our request we purchased the PQ2625 from epcos. Only difference is it has no air gap. We created the airgap to get the primary  inductance value of 269uH.. transformer_report_2014744_21_354160214.pdf

  • Hello Jayaprakash,

    Thank you for providing the transformer specification.  I don't see anything obviously wrong. 

    One thing that I noticed is that the winding guidance says to wind the primary clockwise and secondary and auxiliary to be counter clockwise.  There may be a possibility that the winding connections have confused the polarities in the system.

    Please check your transformer polarities with respect to the pinning of the schematic diagram, either in circuit or outside the circuit with a signal generator.  (For example, using 10~20kHz at 1.0V on AUX should give you 1.6V in-phase on the SEC and 7.8V in-phase on the PRI windings.)  If one of the windings is in the wrong polarity, it can be causing the start-up failure.

    Secondly, another possibility may be that if you created the gap by shimming the outer legs of the core, then fringing flux from the outer gaps may be inducing noise in to the VS or other signals and interfering with start-up.  If you do have gapped outer legs, then try adding a shorted flux-keeper foil winding around the outside of the core.  This could reduce your inductance a bit, but might allow start-up (if this is the problem).  If so, obtain another transformer with only the center-leg gapped (by carefully filing or grinding down the one of center posts).

    Please let me know what you find.

    Regards,
    Ulrich

  • We tested the transformer using signal generator and oscilloscope. The following are the observation.

    We had set frequency to 20KHz and amplitude to 1.0V on the signal generator and supplied to auxiliary point of transformer In oscilloscope very low level signal was observed across the auxiliary points. We calculated the auxiliary inductance using the formula ѠL=2ПFL, as per our calculation we got ѠL=0.628 ohms. Due to this value, when we feed the signal to auxiliary points very low signal was displayed in the oscilloscope.

    Instead of feeding the signal to auxiliary point we fed it to primary points of transformer. We had set frequency to 20KHz and amplitude to 1.0V on the signal generator and supplied to primary points of transformer, at auxiliary point we received in phase signal of amplitude 128 mv and at secondary we received in phase signal of amplitude 200 mv. Calculated transformer values is mentioned below

    Primary                       Inductance Value = 264uH and impedance is 331.5 ohms at 20Khz        

    Secondary                 Inductance Value = 11.7uH and impedance is 1.46 ohms at 20Khz

    Auxiliary                    Inductance Value = 5uH and impedance is 0.62 ohms at 20Khz

  • Hello Jayaprakash,

    You are right, I had forgotten that the Aux winding would be much lower impedance and load down the test signal.  Measuring the primary is the best alternative.  I was hoping to keep the reflected voltages out of the mV range for easier measurement of phase.

    Anyway, the main point of testing the windings was to establish whether the physical transformer polarities are correct when it is installed in the circuit, according to the schematic drawing.  Mis-wired polarity is one of the possible causes of not starting. 

    If there is no polarity error, then that can be removed from consideration. 
    Then it is back to narrowing down the type of fault (latching or non-latching), and follow-on narrowing of each fault in the category.

    This will be a process of elimination.  For example, if the output does not rise, then it is extremely unlikely (other than due to some strange kind of noise) that it is an output over-voltage shutdown.  Scratch that off the list, and look at the next possibility.  Unfortunately, this is a bit of a tedious chore to perform. 

    One other bit of advice: make sure all of the polarized parts in the circuit are installed correctly and that the component values are correct.  The schematic diagram may have no problem with it, but if the actual board is built with a mistake, it could take a bit of debug time to find that mistake.  Isolate the circuit into smaller sections and verify that each node voltage and current behaves as you would expect it to behave when given a specific input signal.

    Finally, try following this debug guide: http://www.ti.com/lit/an/slua783/slua783.pdf
    While not specific to the UCC28630, many of the issues and debug techniques are the same as for the UCC287xx family.

    Regards,
    Ulrich

  • Hello Jayaprakash,

    I have not received a response in about a week on this issue.  I will assume that things are going well and will close this thread.

    If you are still having trouble starting up, please reopen this thread or start a new post with the latest information.

    Regards,
    Ulrich

  • Test Setup used for testing the AC-DC Adapter is attached.

    When we switch ON the unit following observations were made:

    1. VDD pin in IC is not getting locked it is repeatedly changing between 12V and 8V (for 0.6 sec VDD is at 8V and for 0.2 sec VDD is 12V and this is repeating)

    2. DRV pin in IC, Secondary output and auxiliary output of transformer are at zero level.

    3. Voltage at HV pin in IC is also varying.we are getting rectified output of 250V for 0.4 sec and zero output for 0.2 sec.

    TEST SET UP.docx

  • Hello Jayaprakash,

     

    Thank you for the additional information. Unfortunately, your observation descriptions are not enough to be able to diagnose the start-up problem.

     

    For observation 1, you describe a rectangular waveform where VDD dwells at 8V for 600ms then jumps up to 12V for 200ms and jumps back to 8V and repeats. However the expected non-starting VDD waveform is a rising ramp from 8V up to 14.75V and then a falling ramp back to 8V and repeating.

    Please verify what the actual VDD waveform is.

     

    For observation 2, if the DRV output is always zero, then I suspect that the controller has latched off on one of the latching faults at the first application of ac power. After latching off, all switching activity will stop until VDD is allowed to fall below the latch-reset threshold of 5V.  Essentially, this requires removing AC power for a few seconds.

     

    For observation 3, the voltage to HV roughly tracks the activity on VDD, where the “zero output for 0.2 sec” corresponds to the 200-ms charge-up ramp of VDD from 8V to 14.75V. However, during the charge-up interval, HV voltage will not be zero, but should be a volt or two higher than VDD, tracking the ramp.  During VDD discharge, the internal JFET will be off and HV will be at the line peak voltage.

     

    Can you please list which of the debug steps in the Troubleshooting Guide (SLUA783.pdf) you have completed? You should have been able to see DRV output pulses the first time that VDD reaches 14.75V after applying AC power.  If there are never any more DRV pulses, then the IC is latching off due to one of the latch-off faults being detected.  Based on your description, this seems to be the case, but should be verified by cycling AC power and looking for the first few DRV pulses. 

     

    Presuming that latch-off is verified, Table 3 in the datasheet (page 42) lists a number of faults that will latch-off the UCC28630.  “VDD OV”, “OverTemp” and “SD pin low” are unlikely and can probably be easily eliminated by observation.   But don’t assume it; verify them.

     

    That leaves 5 remaining faults to examine and eliminate one by one. Table 3 lists circuit operating faults.  These are faults detected while the IC is switching (however briefly).

     

    Table 4 (page 43) lists additional latch-off faults, but these are pin faults. In other words, there is a problem with the pin connection: either open or shorted to something else.   These faults are detected before any switching starts.  Each of these faults also needs to be investigated and eliminated as the reason for not starting.  If you never see DRV pulses even at the very first time VDD reaches 14.75V, then the problem is probably one of the pin faults.

    I recommend that the faults of Table 4 be discounted first, to make sure all of the pins are properly connected before debugging the possible Table 3 faults.

     

    If you come across a waveform that you suspect may be the cause of the latch-off, but is not absolutely obvious to be the cause, kindly provide waveforms of sufficient detail and resolution for me to corroborate your findings.

     

    Regards,

    Ulrich

  • Wave forms observed at different pins of UCC28630 IC is attached.Please review.Wave forms of UCC28630.rar 

  • Hello Jayaprakash,

     

    Thank you for the waveforms; they are a good start to debugging this.

     

    From “VDD 2.jpg” we can see that VDD builds up from 5V to ~15V (actually 14.75V), then drops rapidly to 8V and then sags more slowly down to 5V and repeats. Sometimes it is a little different, and I’ll get to that shortly.   The main point here is that the slow sag is about 1 second long which corresponds to the “tRESET(long)” parameter in Table 6 (page 52 of the datasheet).   This indicates that the fault is an auto-restart fault, not a latching fault. 

     

    A latch-off fault would show VDD stuck oscillating between 10V and 8V (“VDD(latch-hi)” and “VDD(latch-lo)” in Table 6) indefinitely until AC power is removed.

     

    Sometimes VDD builds up from 5V to 15V and then sags slowly to about 12V, then drops down to 5V rapidly and restarts. This is a different case, and I believe it is related to the “tRESET(short)” parameter in Table 6, where an AC undervoltage condition or X-cap discharge event triggers a 0.5sec off time.   This provides a clue where to look for abnormal behavior.

     

    In the UCC28630 column of Table 3, only “VDD UV”, “AC brownout”, and “Overload timer” are non-latching faults. In Table 4, only “DRV-open” and “HV-short-to-VDD” are hiccupping faults (non-latching).

    Well, we can eliminate “VDD UV” and “Overload Timer” faults. This leaves “AC brownout”, “DRV-open”, and “HV-VDD short” as possibilities.   

     

    All three of these are some-what interrelated. You can check “DRV-open” by looking for gate pulses at the gate of Q1 (not on the IC pin itself).  The DRV signal is involved with VSENSE through R5, which affects the ability to sense the bulk voltage during the FET on-time (indirectly measuring the AC line).

    There may be more than one problem in the system, and start-up may fail for different faults at different times. So it is important to check them all out.

     

    You will have to use a much faster sweep speed, such as 10us/div or so to see the switching pulses. Also, it will be helpful to look at more than one signal at a time if you have more than one probe. Key waveforms are DRV, Vaux, and CS.  Place the zero-reference indicators on a reticule dotted line, please, to make it easier to read the voltage levels.  Also, please indicate the test conditions (input Vac, output load, etc.) for each waveform.

     

    Also, please tell me what the part numbers are for D6 and D7. The wrong diodes here can impair proper operation.  

     

    Regards,

    Ulrich

  • TI recommends Diode D6 value to be 1N5711WS-7-F. we have also used the same part.

    TI recommends Diode D7 value to be SK220ATR. we have used DFLS1200-7 (SK220ATR not available. TI web bench recommenbed alternate part as DFLS1200-7 )

  • Hello Jayaprakash,

    Thank you for the diode information.  These parts are suitable for the job, so we still need to analyze the switching waveforms for more clues. 

    Regards,
    Ulrich

  • Hello Jayaprakash,

    Several days have passed since my last reply to you.  Hopefully, you have identified and solve the problem. I need to close this thread.  If there is still a start-up issue, you may re-open this thread.  Otherwise, if necessary you can start a new posting.

    Regards,

    Ulrich