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.

TIDA-050012: TIDA-050012 Bypass DCDC Protection

Part Number: TIDA-050012
Other Parts Discussed in Thread: TPS65987D, LM3489

Hi all, 

I hope everyone is doing well! On the TIDA-050012 board, if we happen to use an external power path to bypass the DCDC, are there any additional protections that need to be implemented for that DCDC converter? 

Thank you in advance for looking at this post and I hope to hear back soon!

  • Hello,

    Would you be able to elaborate more on the need to bypass the DCDC and how you plan to do that? There are a few concerns I may have, but first want to get an idea of why you are wanting to do this to see if I can give some suggestions

  • Hi Adam, 

    This is the Bypass Path that I want to follow from the User Guide. As the guide mentions, this minimizes losses in the TIDA. Please advise and thank you for your consideration.

  • Sorry Adam, the example should be correctly uploaded now

  • Hi Filipe,

    Yes, the purpose of this bypass path is to directly pass the voltage provided from the barrel jack to the Type-C connector. The Barrel jack has a 19.5 V output so the PD controller enables this bypass path whenever a 20V contract is negotiated. 

    With a architecture like this, you are relying on the protection capabilities of the source connected to that bypass path. As long as you implement the bypass path for one specific voltage contract (PDO) then there should be no issues.

    Can use the TIDA-050012 schematic as reference 

  • Hi Adam,

    Adding onto this, reviewing the BOM for TIDA-050012 I see that the two parts crossed out are not included in the BOM. Also, the two 0.1uF caps that are, are different parts (One is X5R, the other is X7R), do you know why this was done?

    Thanks,

    Michael

  • Hi Michael,

    For the cap differences, I don't think there was a specific reason for the different cap rating since there is no need to have the cap rated differently for this portion of the circuitry. 

    As far as those two components not in the BOM, they are actually marked as DNP in the design files which is why they are not included in the BOM. They components are optional and are not needed for successful operation. Apologies for the confusion 

  • Hi Adam,

    Thanks for the info, this makes sense. I have a follow up question: this PFET in the bypass circuit seems redundant with the internal TPS65987D FETs for a sink application. I'm assuming this FET is used when sourcing, if I am only operating as a sink do I need this FET?

    Thanks,

    Michael

  • Hi Michael,

    No, I believe you would still need this FET even if you are only using it for one direction. The purpose of this FET in a sink application is for the blocking diode when the power path is disabled.

  • Hi Adam,

    I guess I'm still not sure what purpose the second PFET serves. If the power path is disabled, and I disable the buck converter, then I have 20V on my input and 20V at the output of the buck/start of the TPS65987D power paths. So what exactly is the blocking diode being used to block? There might be something silly that I'm overlooking, thanks for your help.

    Thanks,

    Michael

  • Hi Michael,

    That's okay, I'll provide an example to hopefully help explain the importance of having back to back FETs.

    As I'm sure your aware, for a Type-C PD application, you can sink or source up to 100W of power (20V/5A), but a lot has to happen to get to that point. When one Type-C PD device is connected to another, they communicate with one another via the CC pins in order to determine the power capabilities, alternate mode capabilities, etc... They then negotiate a PD contract to transition from a implicit 5V contract to a negotiated 20V contract. 

    However, lets say there is a non-compliant device connected to the Type-C port that presents a hot 20V. Our PD controller will see the abnormal behavior, trigger this as a over-voltage event, and never enable to sink path. Referring to the schematic capture above if you only have a FET populated where Q10B is, even though the PD controller will not have the external path enabled, the body diode of the Q10B will still allow for voltage to pass through. In this example, the body diode in Q10A will act as the blocking diode

  • Hi Adam,

    I think we might be talking about two different things. I'm looking at the back to back FET bypass circuit that goes around the buck converter before the PD controller:

    If there is an overvoltage event on the Type C connector, the internal power path of the TPS65987D will shut down and prevent that voltage from getting to the controller input/buck output correct? So then the first FET of the bypass path has nothing to block. Sorry again if I'm just missing something.

    Thanks,

    Michael

  • Hi Michael,

    Apologies, I was referring to the wrong FETs. Thank you for recognizing the confusion and clarifying by sharing the picture. 

    You still need both FETs in this type of scenario. The barrel jack has 19.5V applied to it. The LM3489 is used regulate this 19.5V input to lower voltages needed for a Type-C PD contract. The bypass path is used to connect the barrel jack directly to the TPS65987D so that their is minimal power loss. When this bypass path is disabled, you need the FET you have circled above to act as a blocking diode so that the voltage the LM3489 outputs does not feed back to the barrel jack node