TPS65987DDK: TPS65987D PD config: phone as host + sink (dongle as device + source)

Hi TI team,

I'm designing a Type-C dongle using the TPS65987D together with an external SPI Flash (W25Q16JV) to operate in standalone mode, without a host MCU.

 The goal of the system is:

• When an Android phone/tablet is connected:

  • The phone acts as a USB Host (DFP) for OTG data communication (via D+/D–)

  • The phone also receives power (Sink) from the dongle (5V @ 2A), but only when an external adapter is plugged in

  • The dongle must never draw power from the phone

• When the external adapter is unplugged:

  • The dongle should not pull power from the phone (no dead battery sink behavior)

  • USB data and VBUS should be fully disconnected

️ Our plan:

• Use TPS65987D in standalone boot mode with configuration stored in W25Q16JV SPI Flash

• Generate .bin configuration using TPS6598x-CONFIG Tool

• Configure:

  • Power Role: DRP, prefer Source

  • Data Role: Force UFP (to keep the phone as DFP host)

  • Advertise PDO: 5V @ 2A

  • Disable VBUS sourcing unless external 5V input is present

• Dead battery behavior should be avoided (don’t allow the dongle to appear as Sink when unpowered)

We would appreciate if TI could provide:

1. Any reference design, schematic, or evaluation board that demonstrates a similar configuration using TPS65987D + SPI Flash in standalone mode

2. A sample .bin configuration (PD contract with 5V source, UFP data role, and no dead battery Sink exposure)

3. Recommendations for CC line protection (e.g. using TPD4S311A) to physically isolate Rd resistors when unpowered

Our goal is to implement this without an MCU, relying fully on the TPS65987D + W25Q16JV combination.
Thank you very much for your guidance and support!

Best regards,
Hieu Tran

I am designing a system using the TPS65987D in standalone mode, with an external SPI Flash (W25Q16JV) to store the PD configuration.

If I already have the firmware .bin file generated using the TPS6598x-CONFIG Tool (v6.1.4), I would like to know whether it is possible to use my own STM32 MCU to flash this binary file to the SPI Flash (instead of using TI's USB2ANY or TIVA LaunchPad tools).

 My questions:

• Does TI officially support flashing the .bin file to SPI Flash using an external MCU such as STM32?

• Does the .bin file require any special formatting before flashing (e.g., offset, header, alignment, etc.)?

• Is there any official documentation (e.g., an Application Note) or procedure that describes how to manually program the SPI Flash (e.g., sending commands directly to W25Q16)?

• Should the .bin file be written starting from address 0x000000 in the SPI Flash?

Thank you for your guidance and support.

Best regards,
Hieu Tran

Dear TI Support Team,

I am currently working on a USB Type-C project using the TPS65987D in standalone mode with external SPI Flash. I am using the TPS6598x-CONFIG Tool (version 6.1.4) to generate the configuration binary (.bin) for the system.

To better understand the configuration process and avoid misconfigurations, I would like to request the following resources:

1. An official **User Guide** or **step-by-step documentation** for the TPS6598x-CONFIG Tool.
2. Any available **application notes** related to configuring TPS65987D for standalone mode with SPI Flash.
3. **Video tutorials** or training content (if available) demonstrating how to use the TPS6598x-CONFIG Tool effectively.

These resources would greatly help ensure correct implementation and reduce trial and error during the configuration process.

Thank you very much for your support.

Best regards,
Hieu Tran

Dear Chris,

Thank you very much for your detailed response and confirmation regarding the TPS65987D configuration.

At this time, I do not have enough time or resources to obtain and evaluate the TPS65987EVM. Therefore, I would really appreciate if you could provide a more specific and structured guideline (or step-by-step flow) to help me implement the standalone configuration using TPS65987D with SPI Flash (W25Q16JV).

My immediate goal is to support simultaneous USB data (OTG) communication and 5V charging to the phone. For now, I plan to:

- Connect the 5V adapter to power the system first
- Then connect the phone (USB Type-C) afterward
→ This way I can avoid the "dead battery mode" issue, at least temporarily, until I add a proper CC protection IC (such as TPD4S311A) in the next revision.

If there is any document, application note, or checklist describing the proper register settings, PD role behavior (Source + UFP), and flash memory structure for TPS65987D, it would be extremely helpful at this stage.

Thank you again for your support.

Best regards,
Hieu Tran

Dear Chris,

Thank you again for your continued support.

While I’ve been exploring the TPS65987D and the CONFIG Tool, I would really appreciate some high-level guidance or design insight — particularly regarding **how the TPS65987D enables the phone to act as a USB Host (DFP) while still being charged (Sink)**.

At the moment, I am still unclear on the internal PD negotiation flow that makes this asymmetric role combination (DFP + Sink) possible. Specifically:
- How does TPS65987D present itself during the initial Type-C connection?
- How can it advertise 5V Source power while remaining a UFP for data?
- Does this require mandatory support of USB Power Delivery (PD) on the phone/tablet side?
- If the phone does **not** support PD, is there any workaround to still achieve this functionality?

I understand that using DRP mode and initiating a data role swap might allow this, but I would really appreciate if you could suggest:
- A reference configuration or concept to build upon
- Or an alternative architecture that may work even without full PD support from the phone

At this point, I am still trying to fully understand how TPS65987D can coordinate this behavior in standalone mode, and I would greatly appreciate any advice or example to help clarify the system design.

Thank you once again.

Best regards,
Hieu Tran

Hi Hieu Tran,

App Customization Tool Users Guide: TPS6598x Application Customization Tool (Rev. C)

Video: https://www.ti.com/video/6122862051001

Data and power rles: https://www.ti.com/lit/an/slvaec3/slvaec3.pdf

Hieu Tran Trong said:

Any available **application notes** related to configuring TPS65987D for standalone mode with SPI Flash.

I don't know if there is a specific note, but once you have configured the GUI with the configuration you want, use the "Binary -> Save Binary" option, and generate a "Full Flash Image". Once you have the Full Flash binary saved, use a SPI Flash flashing tool of your choice to load the image to the Flash.

The GUI handles the flash binary structure. All you need to do is load it to the Flash after it has been generated.

Hieu Tran Trong said:

At the moment, I am still unclear on the internal PD negotiation flow that makes this asymmetric role combination (DFP + Sink) possible. Specifically:
- How does TPS65987D present itself during the initial Type-C connection?
- How can it advertise 5V Source power while remaining a UFP for data?
- Does this require mandatory support of USB Power Delivery (PD) on the phone/tablet side?
- If the phone does **not** support PD, is there any workaround to still achieve this functionality?

I do not know of a complete document for what you are looking for, but here are some suggestions:

Initial Type-C Connection: Port Config Register, Port Configuration Field: Set to DFP

Advertise 5V source: Configure Transmit Source capabilities to only have a 5V PDO

Hieu Tran Trong said:

- Does this require mandatory support of USB Power Delivery (PD) on the phone/tablet side?
- If the phone does **not** support PD, is there any workaround to still achieve this functionality?

Connecting to the phone to get a Data UFP Power source contract requires PD. There is no workaround that I know of.

Hieu Tran Trong said:

I understand that using DRP mode and initiating a data role swap might allow this, but I would really appreciate if you could suggest:
- A reference configuration or concept to build upon
- Or an alternative architecture that may work even without full PD support from the phone

Unforunately the only reference we have is the EVM. There are default projects in the GUI for DFP, UFP and DRP systems that can help get you started.

Thanks and Regards,

Chris

Dear TI support team,

I am currently working on a project using the TPS65987D and would greatly appreciate your help reviewing the configuration I’ve made using the Application Customization Tool (v6.1.4). My main goal is to allow a phone (USB-C Android device) to act as a USB **Host (DFP)** while also receiving **power for charging (Sink)** through the same port – essentially enabling both **OTG + charging simultaneously**.

I’ve followed the PD spec recommendation to:
- Start with the port configured as **Source + DFP**
- Advertise only a 5V PDO in Transmit Source Capabilities
- Enable **Data Role Swap** (Initiate & Process Swap to UFP) so that the phone can become DFP (Host), while TPS remains the power source.

I’ve created the `.pjt` file based on this logic, and exported a `.bin` file for flashing to external SPI flash (W25Q16JV).

Would you be kind enough to review my `.pjt` file to see if there are any misconfigurations or optimizations needed to achieve this goal?

Any feedback or suggestions – even small ones – would be extremely valuable to me.

Thank you very much in advance!

Best regards,
[Hiếu Trầntestx.pjt]

I’d like to share the complete hardware setup used in my testing. I’ve attached a hand-drawn block diagram of my experimental circuit. In this setup: - The left USB-C port connects to the smartphone. - The right USB-C port is the power input (5V charging). - The TPS65987DDH is used as the core PD controller, with SPI flash (W25Q16) storing the configuration (.bin) file. - USB D+/D− lines from the smartphone are routed through the TPS to an external peripheral (e.g., a USB mouse). To avoid dead battery conditions, I always connect the power input (charger) first, then connect the smartphone. However, even when doing this in order, I noticed that the system still does **not support simultaneous charging and USB peripheral communication**. I also experimented with pulling the **CC pin on the smartphone side to GND using a 5.1kΩ resistor**, but that did not help trigger . Could you kindly help review if there’s something missing in the configuration or schematic that might prevent proper USB host behavior alongside charging? Again, I greatly appreciate your guidance and support.Dear TI team,

Thank you very much for your time and support.

I would like to kindly ask you to review the attached `.pjt` file I’ve configured. Based on this setup, the TPS65987D does not appear to successfully enable both charging and USB peripheral communication at the same time when connected to a phone.

Individually, the system can either:
- Charge the phone (power delivery works), or
- Allow the phone to act as a USB Host and recognize a peripheral like a mouse (OTG works)

However, when attempting to achieve both simultaneously (phone acting as Host and also being charged), it does not function as expected.

I truly hope you could help me verify whether there is anything misconfigured or missing in the current `.pjt`8551.testx.pjt
. Any insights or corrections would be deeply appreciated.

I understand your time is very valuable, and I sincerely thank you for considering this request.

Best regards,
[Hiếu trần]








I’d like to share the complete hardware setup used in my testing. I’ve attached a hand-drawn block diagram of my experimental circuit. In this setup: - The left USB-C port connects to the smartphone. - The right USB-C port is the power input (5V charging). - The TPS65987DDH is used as the core PD controller, with SPI flash (W25Q16) storing the configuration (.bin) file. - USB D+/D− lines from the smartphone are routed through the TPS to an external peripheral (e.g., a USB mouse). To avoid dead battery conditions, I always connect the power input (charger) first, then connect the smartphone. However, even when doing this in order, I noticed that the system still does **not support simultaneous charging and USB peripheral communication**. I also experimented with pulling the **CC pin on the smartphone side to GND using a 5.1kΩ resistor**, but that did not help trigger . Could you kindly help review if there’s something missing in the configuration or schematic that might prevent proper USB host behavior alongside charging? Again, I greatly appreciate your guidance and support.

Hi Hieu tran,

Is there a difference in the pjt when the USB works, vs. when charging works?

What changes in the system when one works vs. the others.

For the pjt that you provided, does it consistently exhibit a specific behavior?

If you read the status register after connecting the phone, what power role and what data role do you see the port settle into?

Thanks and Regards,

Chris

Hi Chris,

Thank you so much for your prompt and helpful response.

To clarify your questions:

Is there a difference in .pjt when charging works vs. when USB data works?

No, I used the same .pjt file for all my tests. There is no difference in firmware/config between the two conditions. The observed behavior changes depending on the physical connection sequence and the phone’s role detection.

What changes in the system between the two working cases?

• When I only connect the smartphone, it enters USB Host mode correctly, and a peripheral (e.g., mouse) is recognized.

• When I only connect the 5V charger, the TPS65987D powers the phone correctly (charging works).

• But when both are connected, especially in the order charger first, then phone, I expect the TPS to act as USB Device (UFP) while remaining as power Source, which should allow both charging and USB data to work.

  • However, only one function operates at a time — either data or power, but not both.

What roles are observed in the status register?

After connection, I read the Status Register via I²C. The results show:

• Power Role: Source  (as expected)

• Data Role: Stays at DFP  (should switch to UFP to allow phone as Host)

Even though I set:

• Initiate Swap to UFP = TRUE

• Process Swap to UFP = TRUE

• Initiate Swap to DFP = FALSE

• Process Swap to DFP = FALSE

The data role does not change to UFP.

Additional notes:

• I’ve attached a block diagram in a previous message to illustrate the setup.

• I also tried manually pulling CC to GND via 5.1kΩ resistor on phone side to simulate Rd pull, but it didn’t help.

What I hope to confirm:

• Whether this .pjt is sufficient to trigger data role swap → UFP after initial connection.

• If there’s any constraint in PD spec or TPS65987D that prevents simultaneous UFP+Source behavior in this use case.

Hi Hieu Tran,

Your config seems to be good, but there are cases where the PD controller can attempt a DR_Swap to UFP and the far end rejects the DR_Swap. Ideally, you would have a PD analyzer and we could see the communication on the CC lines.

Is the phone connected to the system over a type-C to type-C cable? Are you sure that the phone supports the PD protocol?

Do you have an EVM or another Type-C PD DRP device that you can test with?

Could you try changing the Port Configuration to UFP only, so that the initial connection is forced to UFP? The initiate swap to source should attempt to source and charge. Report the Status fields for the power and data roles again.

It is starting to sound like the far end is rejecting the power/data role swaps.

Thanks and Regards,

Chris

Thanks. Although my phone supports PD technology as shown in the image, I cannot charge and transfer data simultaneously. However, I tested it on a Samsung device and it worked perfectly.
Please explain the cause and how to fix this issue.

Hi Hieu Tran,

At this point, it is difficult to tell what is failing without additional logs, specifically the PD logs. Your config should be attempting to Data role swap into the desired data role, but we cannot tell if it is even attempting this correctly without the PD log. The last thing I can think to recommend is to monitor the connection status if you have an I2C controller, use the 4CC commands SWUF or SWSr to force a power role or data role swap manually. If you had a PD analyzer, we may be able to see if the phone is rejecting the role swaps.

I did a quick test with your testx.pjt and an EVM, and the image seems to be working properly. When a phone is connected, the TPS65987 becomes the source/UFP as desired.

Thanks and Regards,

Chris

I understand that. I have tested with many phones – except for Xiaomi and Realme, most other phones like Oppo and Samsung work.
My configuration is correct, but very few phones actually support this feature.
Below is an image of the Samsung A35 that I tested.
Although its specifications do not mention PD as the charging technology, this phone still supports simultaneous charging and data transmission.

.

You suggested that I use a protocol analyzer to monitor the role swap communication, but to be honest, I currently don't have the equipment or tools to do that.

Instead, I used an oscilloscope to observe the signals during the connection and successful role swap with two phones – Samsung and Oppo – and I saw that both started with identical pulses.

For other phones where the role swap failed, I only saw a very small pulse.

I also identified a critical design issue: adding a decoupling capacitor – even just 1pF or 1nF – between CC and GND prevents any phone from detecting the role swap request.

Another issue is that when I connect both CC pins of the TPS65987 to both CC pins of the phone, it doesn't seem to work. The functionality only works properly when one CC pin of the TPS65987 is connected to one CC pin of the phone.

I would like to know how you configure it to increase the phone's charging power.
In the schematic I previously sent, I configured the power source to 5V at 3A, but during testing – even when the phone battery is at 50% – the charging current is only about 1A on all the phones I tested.

Hi Hieu Tran,

Please translate the message to english.

Thanks and Regards,

Chris