TPS25741: Design with 4-Switch Buck / Boost Converter PMP10698

So I think its probably best if I use the TPS65981. Can someone please advise if this works with the PMP10698? Is the chip configured via software to advertise the proper voltage levels?

Thanks,

Kai

Hello Kai,

We have the attached design using TPS25741. But, to answer your last question, TPS65981 could also interact with LM5175./cfs-file/__key/communityserver-discussions-components-files/138/0576.PR2254E2_2800_Variant-of-TPS25741-Referecne-Design_29005F00_Sch.PDF

Hi Eric, 

Thanks for the reply. My question is actually whether or not the TPS25741 is able to detect the cable and adjust its current advertisement accordingly? I have reviewed the schematic you posted, I do not see the TPS65981 in the schematic, but I am assuming the CTL1/2 pins out of the LM5175 go into a GPIO of the TPS65981. Perhaps it is the wrong schematic?

Thank you for your help!

Kai

Thanks Patrick,

Is there any other downsides to using the TPS25741 over the TPS65981?

Kai

Hi Kai,

TPS25741 is the simplest part to achieve PD charging with data within 60W(100W wiht captive cable). It does what it does and requires no firmware.
While TPS6598X is a more flexible yet more complicatied part(also more expensive). It enables DRP, e-marker cable communication, video, etc..
I wouldn't call TPS25741 has downsides over TPS65981. it's just they have different level of integration and therefore target different applications. May I ask what particular applicaiton you're aiming for?
Thanks.

Regards,
Patrick

Thanks, I think the TPS25741 is what I am looking for. We essentially have an ATMEL SAMA5D31 processor, and we are using the ports mainly for power delivery. The USB protocol will only be 2.0. Based on what you have said, and what I have researched. I think the TPS25741 is a better choice, given we will also have 3 Ports capable of full 100W PD. We won't be transmitting video etc.

So maybe my solution is this, could I have a jumpers between HIPWR and GND, HIPWR and DVDD, so that we could select whether 3A or 5A is advertised? This might be a dumb question... but will the TPS25741 limit the current from the LM5175 configuration based on the state of HIPWR?

Thank you so much!

Kai

thanks, I think I actually decided I will use the TPS65981 for the port that can be both device and host (OTG), and use the TPS25741 for the other two ports!

Would someone be able to look this over for errors? This is my design so far using the TPS25741.

Thanks in advance,

Kai

Hi Kai,
Generally, your schematic looks good. It looks like you've used PMP10698 for the power supply which is fine. In PR2254E2 (earlier attachment I sent), we've changed a few component values such as the LM5175 COMP pin zero (1k and 0.027uF) and also the slew caps around the FB pin (upper should be 0.022uF and lower = 0.1uF). You could also combine the CTL1/CTL2 resistors into one value as shown in PR2254E2 (the "T" network is replaced by the slew caps around FB. Pin programming of the TPS25741 looks good (93W at PSEL and 5A or 3A at HIPWR).

The current sense resistor value showed as 8 milli-ohms rather than 5 milli-ohms and I'd recommend using the R-C filter on Rsense for starters (you could remove it later if not needed). The CC1 and CC2 caps show as 300pF. We'd recommend 470pF. There is only one FET which is OK for a captive cable design but blocking of the LM5175 bulk capacitance is required if using a receptacle. For the Back-back FET design, see PR2254E2.

The schematic is a bit fuzzy so I may have missed some values. Please attach a PDF if possible for the next round.

Hi there,

Thanks for reviewing the schematic. I made the changes as per your recommendations, see attached screenshots. I have a few questions:

1. Should the ISNS+- pin on the LM5175 be connected?

2. What is the purpose of the smart diode on the PR2254E2? Would you recommend I include it?

3. Is it ok I am powering the TPS25741 off of the VPWR pin and not VCONN?

4. PR2254E2 shows diodes D6/D7 (3.2V) attached to the main bus, what are their purpose?

5. As I understand it, the pins "DEBUG,AUDIO,POL,UFP,GDPG,G5V" are for debugging, is this correct? They are not needed to operate the chip? Would you recommend test points on any of them?

Thanks!!

Kai

Thank you for the detailed reply.

I have one more question! More on the data side of things, given that D+- pairs are on both sides of the connector. Would I also need TS3USB30 High-Speed USB 2.0 (480-MBPS) chip in order to multiplex the signal, based on what the POL signal is? Is this the proper way to do it? I see that the TPS65982 has this functionality built into it.

Hello,

I have decided to use the one design with the TPS25741, and one port with the TPS65986, would you be able to post an evaluation board schematic for the TPS65986 if one exists? Similar to what you sent me for the the TPS25741.

Thanks!

Kai

Hello Kai,

Did you get a chance to look at any of these designs on the TPS65986 Tools Folder? Please have a look here: http://www.ti.com/product/TPS65986/toolssoftware

Design kits & evaluation modules (4)

There is also a reference design that may suit your needs.

http://www.ti.com/tool/TIDA-01243

Hi Eric,

I am now working on the design for a USB port using the TPS65982. I have a few questions I hope you can assist with:

• As I understand, the TPS65982 is able to be powered off of Vbus, if this is the case, should my Vbus connection actually connect before the gates? Since there would be no power after the gates, unless a cable was plugged in?
• Is the external flash absolutely necessary? I am only using 480Mbps USB D+/-, and providing external power via the LM5175 circuit as previously shown.
• What is the best method to communicate with the TPS65982 for programming? I will be needing to program it to control GPIO 0/1 to control the voltage output of the LM5175 circuit.

Thanks!

Kai

Thanks Jeff,

That clarifies a lot for me. Is it possible then for me to have the TPS65982 only powered off of the external VBUS supply? Or will I need to supply 3.3V to the VIN_3V3 pin? I have attached a screenshot of my schematic so far. To the left is the 4-Switch Buck/Boost Converter Schematic, supplying the 5/12/20V @5A.

As well, what is the best protocol to program the TPS65982, should it be hooked up to my main processor via i2c?

Here is a more detailed view of the schematic. Does this seem ok? 

• I am unsure regarding the actual communication between the TPS65982 and my main processor. I currently have it hooked up via i2C.
• Can I tie together the GPIO pins not being used through the same resistor, or do they need to be separate grounded through the 1M resistor?
• I am using the reference design HVL117D, and they show a 21ohm resistor, and 4 0.01uF caps in parallel on the Vbus line, would I also need these? I have an array of capacitors on the other side of the FETs already.
• Will the TPS65982 run off only VBUS power, or is an external 3.3V regulator needed to supply VIN_3V3?

Thanks again for your help! I realize these may be dumb questions!

Hi Kai,

I will review your schematic and get back to you asap.

Thanks

Thanks Jeff, I have changed the design to also provide a 3.3V supply to VIN3V3. 

Thank you again for reviewing the schematic, I really appreciate the customer service and support that TI provides for their products.

- Kai

Thank you very much. I will post an updated schematic, and I am hoping you can catch any errors you might see!

Kai

Hi Kai,

Will do, I will compare your schematic against our EVM to see if your differences will work. Ill update u asap.

Thanks

Thanks for the Feedback Jeff. I made those changes as per your recommendations!

Kai