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.

UCC3895: About RDEL settings

Part Number: UCC3895
Other Parts Discussed in Thread: UCC2895

Hello,

I would like to ask you about the RDEL settings for the UCC3895.
When I designed it before, the data sheet I referred to didn't mention the recommended range of RDEL.
Therefore, the current board has a 510Ω resistor mounted on RDEL.
If a resistor outside the recommended range is mounted, is there a problem in operation?
Please tell me the possible problems.

Best regards,

  • Hello,

    The resistor value you have selected is outside the recommended/characterized range.

    The turnon delay most likely will not match the equations presented in the data sheet.

    You should evaluate your design just to see if you can a resistor in the recommended range.

    Regards,

  • Thank you for your reply.

    It is said that it may not match the formula shown in the data sheet. Is there a possibility that the following phenomena may occur due to the effect of implementing a resistor that is not recommended?

    For example
    1. Suddenly, the behavior such as dead time and phase shift overlap angle changes.
    2. Even if the same resistance value that is not recommended is implemented, the dead time value changes depending on the IC.

    Currently, we are mass-producing products using this device, but if there are serious problems with using non-recommended resistance values, I need to modify the board.
    So I need your advice.

    Best regards,

  • Hi,

    Our offices are closed today for the US Fourth of July holiday. Michael will respond tomorrow. 

    Best regards,

    Don

  • Hello,

    If you are using a resistor outside of the recommend range the equations that are presented in the data sheet for calculating adaptive delay are not accurate.  We recommend that you use a resistor that is recommended in the data sheet.

    1.  The phase shift and delays will change with loading if you are using adaptive delay.

    2.   When it comes to using a resistor outside the range it is unpredictable what the delay times may be.  You may want to check the devices that you have to see if the delay times match the calculations when resistors in the recommended range are being used.  The UCC2895 dead time is not guaranteed when the resistors are used out of the recommended range.

    I did look at  equation 3 and with a VDEL of 0V and a 510 ohm resistor your delay time would be 37.8 ns.  With a VDEL of 2V it would be even less.  This seems pretty fast half the resonant period.  Are you shore you need this time?  I would double check your delay settings vs the switch node to confirm this actually what you need for your design.

    Regards,

     

  • My idea about equation(3) is different.
    VDEL should be clamped to a minimum of 0.5V, is my perception correct?
    Besides, if you decrease the value of VDEL, TDELAY should increase.
    Is the previous term of the TDELAY calculation formula 12.8ns when VDEL = 0V?

    Also, when RDEL is 510 ohms and VDEL = 2V, TDELAY = 31.4ns. Is this result correct?

    Best regards,

  • Hello.

    I reviewed the block diagram and VDEL will be a minimum of 0.5 V as you stated.

    This being the case my calculation was off 0.5V.  With this value the delay would be 50.5 ns when CS = 0 V.

    I do get 31.4 ns in the calculation when using 510 ohms and 2V in the calculation.

    So the 510 ohms is outside the recommended values for the RDEL pin.  The turnon delay most likely will not match equation 3.  TI will not guarantee the UCC2895 when the device is outside the recommended values.  However, the device can and still function, it just may not operate as expected. 

    I had suggested you check to see how much turron delay you actually need for the design.  31.4 ns seems kind of fast.

    Regards,

  • I would like to ask you one more question about the impact on VREF.
    (question 1)
    Is it expected that the voltage of VREF is likely to fluctuate because the resistance value of RDEL is lower than the recommended value?
    (Question 2)
    Also, checking the data sheet, the recommended current for VREF is 5mA, and the short circuit current is 20mA (typ).
    If RDEL = 510Ω is used, a current of 5V / 510Ω = 10mA will flow.
    What happens when the VREF current reaches the short circuit current?
    Is it constant current control? Or is it in hiccup mode?

    Best regards,

  • Kaji,

    Mike is currently out of the office, so I have asked someone else from the team to look into this request. Please allow some extra time for them to review the thread and ensure the answer they provide is sufficient to address your questions. 

  • Kaji,

    The operation of UCC3895 with the components which are out of recommendation is confident, and the TI doesn't guarantee UCC2895 can work normally at such condition. VREF is very critical because VREF is the power rail for many internal function block. The switching frequency is defined by VREF, the deadtime is defined by VREF. Whatever VREF is constant current control or hiccup control, the system stability would be affect if VREF is not stable or less than 5V too much. 

    So, it is highly recommend you to design the parameters by following datasheet.

    Yunsheng