• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 2 answers
  • Subscribers 65 subscribers
  • Views 617 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.

UCC28782EVM-030: tps65987 and PM110

robin gangopadhya18
Mastermind 6296 points
Part Number: UCC28782EVM-030
Other Parts Discussed in Thread: UCC28782, TPS65987, , UCC24612

Hello:

In the UCC28782EVM document, they mention PM110.

It cannot be found anywhere.

We are not yet ready to fork out $1000 bucks for testing with anything other than Electronic load, maybe HPIB.

A real question is : does the UCC28782 circuit maintain its efficiency when outputs go down to 5V ?

How does the xmfrmr behave over this large change in output? What specific control aspects of UCC28782 kick in the make this possible?

We understand the role of the WT6636 chip: but surely cannot control UCC2878. How do you compare TPS65987 WITH WT6636?

Any clarification on this will be highly appreciated

  • 0
    TI__Mastermind 24375 points

    Hi Robin,

    This has been assigned to a product expert, please allow one or two business days for a response.

    Regards,

    Harish

  • +1 in reply to Harish Ramakrishnan5
    TI__Mastermind 40140 points

    Hello Robin, 

    It seems that PassMark Software has discontinued the PM110 tester and replaced it with the PM125 USB PD Tester.  At the time that the UCC28782EVM-030 User Guide was written, the PM110 USB PD Tester was available and used by our team to interface with the USB output of the EVM. 

    It is necessary to use some sort of USB-PD test device to interface with on-board USB controller to adjust the output voltages, but it is not required that a PassMark device be used.  At present, a web search for "USB PD tester" yields multitude results including apparently dozens of after-market USB testers promoted on a very prominent consumer mail-order concern (beginning with "A") all of which were listed at less than $50.  Most testers manipulate the USB signals and pass the output power through to a separate load, which may be an electronic load. TI does not recommend any particular brand, but merely reports the tester that was used. 

    The UCC28782EVM-030 User Guide lists the efficiency measurements and plots the curves for all of the output voltages across load and line.
    The efficiency at 5-V output is lower than those of the higher voltages, however it still exceeds the requirements of the European Code of Conduct.

    The UCC28782 controller moves through a series of different operating modes and the overall output power changes from high to low and back again. 
    The WT6636 USB-PD controller does not directly control the operation of the UCC28782, except in the sense that the USB controller generates the error-signal that the drives the feedback current (through the optocoupler) which determines the operating modes and current-sense thresholds for each mode. 
    The USB controller behaves similarly to the usual TL431-type shunt regulator.  

    The TPS65987 is a digital USB controller involved with USB-related communications and power steering, but does not have a means of regulating the output voltage.  Hence it is unsuitable for application on the EVM, which is simply a power source.  

    I hope this clarifies the operation and role of the WT6636 controller. 

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Mastermind 6296 points

    Thanks a lot Ulrich!

    Good to hear back from you...

    -robin

  • 0 in reply to Ulrich Goerke
    Mastermind 6296 points

    Hello Ulrich:

    We are finally at the confluence of UCC28782 power output 20V @ 65W (hoping to get to 100w after more testing)...and USB-C PD. Testing can begin for the USB-C PD interface. We are having discussions with WT6636F folks... a lot is unsaid.

    Most importantly the Gate pin behavior.

    None of their d/s, or explanations seem to stick with us here.

    So seeking some counsel from you!

    - Our concern is that the sink side could be anything an average user might hook up albeit with a USB-C PD cable/plug. One has to imagine the worst combo. The plug has a fully charged 20V battery when they decide to hook it up... No matter how slow the MOSFET WT6636 fellas think long gate time constant is, MOSFET WILL switch in 10 of ns after Vgs > Vth. Unless the MOSFET itself is a super slow type(there are many such) but CSD175xx is superfast.

    Our suspicion is it will blow! Taking down a chain of other devices. (reports of cheapo PDs in a world seem to indicate quick failures engendered by a similar mechanism)... we simply cannot afford that.

    Or Source is at 5V and sink connects to 20V/100W... slow switching will overheat the switch. May even blow it.

    Questions are:

    - What should determine the switch turn-ON OFF speed?... our take is that FAST: if we have a current source from UCC28782 converter, then this speed is moot. Yet, dissipation in the switch will be min. 

    - What should determine how the switch should block the sink side until WT6636 has determined what is connected & regulate UCC28782 accordingly? Our thinking is that the switch should be selected so that it blocks it until Vgs is appropriate for it to turn-ON. Does WT6636 do it? I doubt.

    - High-side connection of a MOSFET always requires an appropriate "floating " drive. A resistive divider is not ONE of the appropriate mechanisms to do so.

    Unless like UCC24612 (or similar) which derives gate drive with respect to high side Source because of the D & S connections to it from the high side switch. It is not clear HOW this is accomplished in WT6636- maybe not at all possible given that their approach seems to be based upon a slow resistive divider.

    This is key lacunae in the drive scheme we think.

    And we need to figure out a way to make it fail-safe & satisfying on-off criteria as we mentioned above.

    Comments to resolve these issues will be highly appreciated.

    We are about to make a board on the testing side with a WT6636 chip.

    -robin

  • 0 in reply to robin gangopadhya18
    TI__Genius 11350 points

    Hi Robin,

    It sounds like you are getting close with your design but I wonder if you can wait about three business days for a response. Ulrich is our UCC28782 product expert and it's probably better for him to respond to your questions. However, he is enjoying a few days away from the office, so if you need an answer sooner than that, I or one of our other team members will try to help. Let me know.

    Ray

  • 0 in reply to Ray DiCecco
    Mastermind 6296 points

    Ray:

    give him time to enjoy the summer vacation...no hurry.

    We have many things to sort out still.

    thnx

    r

  • 0 in reply to robin gangopadhya18
    TI__Genius 11350 points

    Thank you Robin. Have a great weekend.

    Ray

  • 0 in reply to Ray DiCecco
    TI__Mastermind 40140 points

    Hello Robin, 

    The Weltrend WT6636F USB-PD controller used on the UCC28782EVM-030 board has a built-in charge-pump that generates sufficient voltage to drive the PD-switch gate.  So Vgs level should not be a concern. 

    The PD controller is programmed to always start up with the PD-switch off and programs the feedback current to keep the internal output voltage at 5V.  It will only raise internal-Vout to 9V, 15V, or 20V when commanded to by an external USB-PD interface controller via the CC1 and CC2 pins.  And it will not raise the voltage until it has negotiated with the down stream load (sink) that it is okay to do so.  Then it will turn on the PD-switch first and raise the voltage in small steps thereafter so that there is no huge out-rush of current to the sink from a reservoir cap pre-charged to 20V.  There are specific timing intervals defined by the USB 3.x specification for these interactions and voltage changes. 

    On the other hand, if there is an external 20-V source on the cable and it is plugged into the USB Type-C connector, there is nothing the WT6636F can do to prevent an inrush of current through the PD-switch's body-diode.  Fortunately, the internal components are already rated to handle this voltage so it mainly represent a brief thermal disturbance to the switch and an over-voltage condition to the USB controller feedback (it is expecting 5V as a default).  It will keep the PD-switch off due to this apparent OV condition until the source is removed.  
    The only way to prevent this situation is to use back-to-back FETs for the PD-switch, which will reduce efficiency a bit.  The same GATE signal can drive both FETs. I don't think turn ON/OFF speeds are a concern since the internal regulated voltage (default to 5V) can only change when the USB controller is satisfied by compatible conditions on both sides of the PD-switch.   

    I hope this alleviates your concerns.  

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Mastermind 6296 points

    Hello Ulrich:

    Got it!. We have to use back-to-back switches. It is not a big issue.

    Closing the post,

    Thnx a lot.