• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 64 subscribers
  • Views 532 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.

Original question:

UCC28782EVM-030: tps65987 and PM110

UCC28782EVM-030: UCC28782 UNDER WT6636 CONTROL

robin gangopadhya18
Mastermind 6296 points
Part Number: UCC28782EVM-030
Other Parts Discussed in Thread: UCC28782, LM7480

Hello:

We have briefly discussed how UCC28782 behaves under WT6636 control if it wants to have 5V output. Even 9v will be problematic we think......maybe 12v, 15v ok. Our experience with UCC28782 is that it is very sensitive with xmfr turn ratio and 1:.2:.2 definitely will not produce 5V/9V  no matter what load or no load or what mode UCC28782 tries to get into. Specially if it is designed for 20 V 100W output.

It is far more reliable to consider  SOME step-down scheme with high efficiency which will produce all the lower voltages while 20V 5A is a pass-through.

How will that work, what topologies one can look for specially if the output current can be from 0 to 3A...meaning the integrating filter inductor will produce high ripple output in the second stage.

WE KNOW  THEY are letting WT6636 control UCC28782 but all such schemes seem to end up with failed products albeit sold "cheap".

We cannot afford such field failures.

Won't you think a better solution ought to exist? And what about 6V batteries?

A last nagging question is: does the UCC28782 controlled converter output replace the charging controller of the BQ family? Or does it depend upon the device's internal "battery charging" circuit with BQ( or equivalent IC)  to complete the charging modes with safe and long-life process most batteries require? 

This is crucial to determine how UCC28782 controlled module truly charges a standalone battery. Most "chargers" we have encountered in the past decade either did not work after a few months, or burnt up the battery or at best shortened its life at a huge loss to consumer and the environment- ending up in piles of unrecyclable batteries in the dumpster.

We can no longer afford that.

Appreciate your review of the remarks above,

r

  • 0
    TI__Genius 11350 points

    Hi Robin,

    Ulrich will respond in two or three business days.

    Regards,

    Ray

  • 0 in reply to Ray DiCecco
    TI__Mastermind 40140 points

    Hello Robin, 

    A turns ratio of 1 : 0.2 : 0.2 (restated 5:1:1) results in a reflected primary voltage of 100V for 20V output and 25V for 5V output.  The turns ratio is usually selected based on the main FET Vds-rating = Vbulk_max + V_reflected + derating margin + V_spike (if any).  Fine-tune the P-S ratio with consideration of the rating and stress on the SR-FET (or diode).  
    For USB-PD applications, the values are usually 5V, 9V, (12V sometimes), 15V, 20V, and now 28V and higher.  TI's EVM produces 5V, 9V, 15V and 20V. 
    6-V batteries are not in the USB range, other than regulated down from 9V or higher.  
    However, the fixed USB voltages are not designed for directly charging batteries but for general applications. Instead, the USB-PPS scheme is intended for direct battery charging.  A different USB controller is needed for that (the WT6636F is not designed for direct battery charging).   

    Meanwhile, the 20-V to 5-V output range is also reflected back the AUX winding at a 1:1 ratio.  This is too low for the UCC28782 bias when 9V and 5V are programmed, but the built-in boost function of the UCC28782 boosts these low voltages back up to 18V. 
    Similarly, if a low battery is directly charged via USB-PPS controller, the boost function is active all the while the Vbatt (Vout) is < 15V.

    In all cases, the charging of a battery must be controlled by a dedicated control IC to regulate the proper voltage and current to that battery at the correct amounts for its state of charge, in order to maximize the battery lifetime.  It can be a BQ controller or some other controller; they all modify the feedback current to the UCC28782 to establish the correct output voltage necessary.  If batteries are charged improperly, it is the fault of the charging controller used, not the UCC28782 that is delivering the charge demanded by that controller. 

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Mastermind 6296 points

    Ulrich:

    I got almost tortured for 6 days on this issue: too bad I had not seen this response on the day you posted!

    Thanks for absolute clarity on the issue. Unfortunately, folks seem to think such controllers do the charging just because they have " current sensing" inputs and can control Vout.

    I had worked on BQ applications in my previous life, so I knew better than to trust a plain explanation of "charging"...

    More unfortunate is that our LiPO supplier says " we only sell batteries"...HA! Who would know what they are selling? Since charging /discharging is so critical to the safe and prolonged use of the batteries.

    A lot of batteries will end up in piles of junk....chargers are being sold for  6 bucks in an A world....

    So our products will have BQ series following UCC28782 with a controller for interfacing USB C port. With 2 applications: PD and charging. They are not synonymous.

    One thing in this context.

    I also fear that applying suddenly a sink load a the receptacle with MOSFET  will cause failures! Unless  Vout is from the diode at the transformer as they show in WT6636.

    It is going to be worse fate if SR is being used. WT solution with 2 back to back switches is double trouble!

    They show common source node floating. With a charge pump at 8.6V above VCC, you guarantee switch failure when Vgs goes to 20+8.6= 28.6V

    We cannot afford to design in products with known failure modes: so have developed a 12.5mm square part that guarantees fail-safe blocking of any sink load at the port. Only when the controller is ready to enable the blocking device will it turn ON and allow PD or charging...It has TI parts in it...

    I would like to have this reviewed by you. And hopefully put it in the marlket.How do I do that?

    Appreciate any help in the matter.

    -r

  • 0 in reply to robin gangopadhya18
    TI__Mastermind 40140 points

    Hello Robin, 

    I'm not sure how to reassure you.  The WT6636 follows the established USB 3.x protocols and these protocols prevent uncontrolled loading of the USB-type-C connector.  Our EVM uses a single blocking switch.  Other designs may use back-to-back FETs as the blocking switch.  In either case, the USB protocol, implemented through the WT6636 (as well as other USB controllers) will not allow the wrong voltages to be applied at the wrong time. 

    Of course, no one can prevent somebody from attaching a USB-C plug to some non-USB certified circuit and source excessive voltage to any device with a type-C output port.  But that would constitute inappropriate and potentially damaging behavior (even possibly malicious behavior), and can not be protected against without extraordinary means.  All devices designed and certified to USB standards should work together harmoniously. 
    I have not received any reports of damage or incompatibility with our ACF EVM (which has single blocking FET). 

    I can review your special circuit to look for potential stress issues, but I will not be able to guarantee that it conforms to USB standards.  I'll send contact information through your company's website Contact screen.

    Regards,
    Ulrich

  • 0 in reply to Ulrich Goerke
    Mastermind 6296 points

    Ulrich:

    Many thanks for the thoughtful reply!

    While I am agonizing about it, I bump onto a TI chip: LM7480 which does the thing we built into a module. If our module has any raison-d'etre, it is entirely based upon this chip now. And we would convert it to use LM7480.

    "load dump protection"..who would have thought?

    But I like it.

    My fear of failure in this matter came from reviews in the A world. Cheap stuff yet where would failures come from?

    Again, thnx a lot.

    And appreciate the impending info.

    r

  • 0 in reply to robin gangopadhya18
    Mastermind 6296 points

    Hello Uli:

    got your contact thru Ganmar website.

    Will follow up on this: understood, it is purely a review without any implications of any sort.

    BTW: I have finally come to realize the following things:

    1) a PD is a power delivery adapter in the classical sense: when a USB C sink is connected to it, it carries no potential at all. This aspected  is all that mattered to clear up any confusion about WT6636 use.

    So yes, a single blocking switch will suffice.- this in order to let power flow ONLY when a contract has been negotiated ...voltage, power etc.

    1. 2) A USB C Charger by contrast MAY BE connected with the possibility of having any battery sink connected to the charger. The charger then has to establish what it is asking for. But then it cannot allow the sink to connect to any internal node. Until the battery mgmnt system figures out what  BQ chip has to deliver.  Then we might need a different blocking device, right?

    Thnx much for all this help you are extending on this matter. 

    Appreciate.

    r

  • 0 in reply to robin gangopadhya18
    Mastermind 6296 points

    Hello Uli:

    got your contact thru Ganmar website.

    Will follow up on this: understood, it is purely a review without any implications of any sort.

    BTW: I have finally come to realize the following things:

    1) a PD is a power delivery adapter in the classical sense: when a USB C sink is connected to it, it carries no potential at all. This aspected  is all that mattered to clear up any confusion about WT6636 use.

    So yes, a single blocking switch will suffice.- this in order to let power flow ONLY when a contract has been negotiated ...voltage, power etc.

    1. 2) A USB C Charger by contrast MAY BE connected with the possibility of having any battery sink connected to the charger. The charger then has to establish what it is asking for. But then it cannot allow the sink to connect to any internal node. Until the battery mgmnt system figures out what  BQ chip has to deliver.  Then we might need a different blocking device, right?

    Thnx much for all this help you are extending on this matter. 

    Appreciate.

    r

  • 0 in reply to robin gangopadhya18
    TI__Mastermind 40140 points

    Hello Robin, 

    I'm afraid that you've reached my limit of USB-PD knowledge.  I am unable to answer your 2nd question.  
    I'll see if I can find someone in the Battery Management Solutions division to whom I can direct your PD and BQ questions. 

    Regards,

    Ulrich