• State TI Thinks Resolved
  • Locked Locked
  • Replies 22 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 1666 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.

UCC28780: Understanding OPP curve in Mathcad worksheet

Gerrit Barrere
Expert 1245 points
Part Number: UCC28780

I am working with the Mathcad worksheet for this part from sluc644.  In the "Neutron..." worksheet I don't understand the calculation for VCST_OPP(Vin).

The graph for VCST_OPP(Vin) looks like it is meant to duplicate Fig 34 in the IC data sheet, but with Vin as the x axis rather than IVSL.  Is that correct?

Trying to correlate the two, I don't understand the formula for Vin_IVSL(Vin).  Does this function calculate the voltage on the VS pin as a function of Vin?  If so, shouldn't the resistor values Rvs1 and Rvs2 and the Naux transformer winding ratio be part of that?  What does the fixed 106 V in the function represent?

And I don't understand the fixed voltages in the piecewise function for VCST_OPP(Vin).  It seems like they are intended to create the breakpoints in the OPP curve like Fig 34, but since Fig 34 is a function of IVSL shouldn't the breakpoints depend on at least Rvs1 and Rvs2?  Maybe that would be taken care of if Vin_IVSL(Vin) depended on Rvs1 and Rvs2?

Thanks for any help.

  • 0
    TI__Mastermind 47245 points
    VCST_OPP is one group of specs for this device. In the spces table on ds page 8, it is VCST(OPP). The test conditions for each VCST(OPP) are listed in that table. Figure 34 shows the IVSL range for each VCST(OPP). IVSL is the current through Rvs1 when VS clamps negatively during primary power MOSFET (low side) turns on. IVSL can also be referred in the block diagram on ds page 13. I like to pause here and to see if my description makes sense to you so far.
  • 0 in reply to Hong Huang
    Expert 1245 points
    Thanks, Hong. Yes, let's take this step by step.

    I follow what you're saying. IVSL is a line-sensing current source which flows out of the VS pin. It looks like it will flow into the Rvs1/Rvs2 divider and cause a voltage on pin VS which would depend on Vaux, Rvs1, and Rvs2.
  • 0 in reply to Gerrit Barrere
    TI__Mastermind 47245 points
    When VS clamps negatively, the VS negative clamping voltage is specified in the ds page 7 as VVSNC typical 250mV (negative so it is a number below GND). So Rvs2 effect to IVSL is ignorable as 250mV compared to VAUX is too small to make meaningful difference. IVSL is calculated as VAUX/Rvs1 = (VBULK/NPA)/Rvs1. Let me know if the above makes sense so far.
  • 0 in reply to Hong Huang
    Expert 1245 points
    Yes, thank you, I was just reading section 7.4.10.1 and realized that my description wasn't right. I see now that VS clamps at about -250 mV and IVSL is caused by Vaux going negative and pulling current through Rvs1. Now I'm with ya.
  • 0 in reply to Gerrit Barrere
    TI__Mastermind 47245 points
    Fig 34 provides typical (roughly) VCST(OPP) value when IVSL is obtained, i.e., for given Rvs1, NPA, and VBULK, a corresponding IVSL can be obtained. Then we will know its corresponding VCST(OPP).
  • 0 in reply to Hong Huang
    Expert 1245 points
    Yes, that's clear. VCST(OPP) is then the threshold voltage which triggers OPP when the CS pin exceeds it. The OPP threshold varies with IVSL as shown in Fig 34.
  • 0 in reply to Hong Huang
    TI__Mastermind 47245 points
    VBULK is in relation to Vin(ac) as VBULK = Vin(ac) * sqrt(2) in nature (ignore ripple, voltage drops across rectifier etc).
  • 0 in reply to Hong Huang
    Expert 1245 points
    That's clear.
  • 0 in reply to Hong Huang
    TI__Mastermind 47245 points
    The MathCAD file was created on the assumption of universal input voltage, i.e., 85Vac to 264Vac, as this device was particularly defined for such applications.
  • 0 in reply to Hong Huang
    TI__Mastermind 47245 points
    In such applications, 80Vac is an often used number for converter output voltage to start into regulation.
  • 0 in reply to Hong Huang
    Expert 1245 points
    Maybe we could take this *two* steps at a time...? :-)

    85 - 264 VAC is a common range for AC input, but I don't see anything limiting it to that operation. Does it say that anywhere in the ds? With the proper choice of external components it should be fine with other ranges.
  • 0 in reply to Gerrit Barrere
    TI__Mastermind 47245 points
    I am replying to the below first:

    Yes, you are right, i.e., 85Vac to 264Vac is not a condition for this device to operate. with proper component selection, the device can operate in other input voltage range.

    "In reply to Hong Huang:
    Maybe we could take this *two* steps at a time...? :-)
    85 - 264 VAC is a common range for AC input, but I don't see anything limiting it to that operation. Does it say that anywhere in the ds? With the proper choice of external components it should be fine with other ranges."
  • 0 in reply to Hong Huang
    TI__Mastermind 47245 points
    But, the current MathCAD design tool assumes 85Vac to 264Vac as the design condition, although the device has no such requirement.
  • 0 in reply to Hong Huang
    TI__Mastermind 47245 points
    For different range of input voltage, the assumptions in this tool can become questionable so to cause issues.
  • 0 in reply to Hong Huang
    TI__Mastermind 47245 points
    As VBULK, Vin(ac), Rvs1, and NPA generate IVSL, and IVSL is a function of Vin(ac), and VCST(OPP) is a funciton of IVSL, then VCST(OPP) is a function of Vin(ac).
  • 0 in reply to Hong Huang
    Expert 1245 points
    You can stop now. It's clear that the worksheet does not calculate these variables correctly, even though the user specifies Vin_min and Vin_max. This is the root of the problem with RCS and ROPP calculations being wrong also.

    The answer to my original post is that I will need to rewrite the formulas to make them correct for the input range as specified.
  • 0 in reply to Gerrit Barrere
    TI__Mastermind 47245 points
    I agree.

    But it will take quite a bit time to re-write, or create a new tool, to cover different range of input voltage, and moment we do not have a person who can have time to do it.

    An alternative, then, is to use the existing tool and make manual adjustment to achieve these curves correctly then the resulted numbers can be used for a design.
  • 0 in reply to Hong Huang
    Expert 1245 points
    The whole point of a MathCad worksheet is to calculate values depending on inputs. These calculations are long, complex, and interdependent. If you just short-circuit the whole process by overriding the calculation results, who knows what you'll get?

    For a trivial example, say Y and Z depend on X, and X is in error. If you jam a number in for Y, Z will still be in error.

    Gosh, I'm sorry you don't have time to produce a correct worksheet. I have all the time in the world, and I'll get on that right away.
  • 0 in reply to Gerrit Barrere
    TI__Mastermind 47245 points
    I am not going to argue with you. I am just telling the fact we have.

    By the way, if X is in error, Z can be ok as we have curves as another check.
  • 0 in reply to Hong Huang
    TI__Mastermind 47245 points
    The original questions have been answered. For other associated questions such as to correct the existing, or create, a new tool to cover different input voltage, please create or use a different string. I will take "TI Thinks Resolved" if I do not receive any further questions in relation to the original questions in this string.
  • 0 in reply to Hong Huang
    Expert 1245 points
    For the record, precisely none of my questions have been answered and you have been gaslighting me for days on this and other threads. The worksheet is in fact not correct for input voltages different than the usual range, which you finally admitted yesterday. The worksheet itself makes no mention of this either, and of course the fixed numbers and wrong calculations in it make no sense without knowing that.

    I have a lot more work ahead of me now, trying to determine where the worksheet is making assumptions and how to fix them from the data sheet. I might as well be doing the calculations from scratch, and the worksheet is only marginally useful now.
  • 0 in reply to Gerrit Barrere
    TI__Mastermind 47245 points
    You have to tell me which of the below not answered. If no specific, I will close this string.

    "The graph for VCST_OPP(Vin) looks like it is meant to duplicate Fig 34 in the IC data sheet, but with Vin as the x axis rather than IVSL. Is that correct?

    Trying to correlate the two, I don't understand the formula for Vin_IVSL(Vin). Does this function calculate the voltage on the VS pin as a function of Vin? If so, shouldn't the resistor values Rvs1 and Rvs2 and the Naux transformer winding ratio be part of that? What does the fixed 106 V in the function represent?

    And I don't understand the fixed voltages in the piecewise function for VCST_OPP(Vin). It seems like they are intended to create the breakpoints in the OPP curve like Fig 34, but since Fig 34 is a function of IVSL shouldn't the breakpoints depend on at least Rvs1 and Rvs2? Maybe that would be taken care of if Vin_IVSL(Vin) depended on Rvs1 and Rvs2?"