TPS65987D: TPS6598x part differences

Hi David, 

I've assigned this issue to the relevant expert who will get back to you shortly. 

Hi David,

Do you think you could tell me a little bit more about your application so I can help point you to the right part? I think this could be more beneficial to you than me listing all the differences. If you could provide a brief description with info such as target application, number of ports, is this PD and data or just PD, temp range, if you support thunderbolt, do you need to support PD 3.0, etc. 

Nathan,

David asked for me, so I will chime in.

Looking to add PD to an existing BC1.2/Type-C design.  This is a portable product that in 95% power sink, but may have some cases for power source.  This switch from sink to source would be from external micro and would *not* be negotiated through automatic DRP.  Product will also be a data client (USB 2.0/3.0).   Product will only support 3A contracts, up to 15V.  No need for smart cable support, I believe.  Still need to support BC1.2 and Type-C current adapters.  Simple via construction will be used, so routing without HDI vias is a must.  Dead battery support required.

I would like to find a solution that allows for power contract negotiation only, with no need for alternative modes.  I would like to minimize code impact on the micro I have that currently programs the charger (possibly will use BQ25890/92 for charger). 

Matt 

Hi Matt,

Thank you for providing the details of your application. Based on what you have told me I think the best fit for your application would be the TPS65987D (or the 988 if this is a dual port application). These devices support dead battery and you should have no problems using an external micro to control source/sink modes. Also, it is my understanding that using this part would not be considered HDI construction but I would recommend you verify that with your board house before committing to a particular part. One thing to note is that you will need to use and external mux (which is controlled by our chip) for the USB data as this part does not have one built in.

As far as the other parts you inquired about go (981, 982, 983b), they have an internal mux but it only supports usb 2.0 so you would still have to add an external mux to support usb 3.0

Hi Nathan,

Thanks for the reply.  I understand about the external muxes and had already planned on that for USB 3.0.  

I have some concern about the BC1.2 method of detection, without having an isolation switch, like is on the TPS65981/2/3B parts or that is implemented on many TI BQ series chargers with the DSEL line.  I don't have direct control over the controller that is connected to the USB 2.0, so I have concern over the timing of BC1.2 detection and the device trying to enumerate.  Can you comment on this?  Is there a DSEL like pin that could be used to put an external 2:1 USB switch to do the isolation during detection?

I only need one port, so I would only consider the 987.

I don't need 5A, so is there any advantage to the 981/982/983B, which only supports 3A with built in switches?

Is there a code size difference between the two series of parts that would affect the external micro used to talk to this controller and then program the charger?

What are the major reasons for having two similar, but distinctly different part series (987/988 vs. 981/982/983B)?

What are the differences between the 981/982/983B, besides the one being a different package size?

I appreciate the help.

Matt

Hi Matt,

The TPS65987D is designed to work with BC1.2 without the need for the isolation switch as long as you are configured as CDP or SDP. We have an application note with the details of implementing BC1.2 on the 987D. Here is the link: 

If you check the host interface document and search for CDP or SDP you can see what registers to configure for this mode. 

As far as your question about the differences/advantages of the 981/982/983B goes, the 981 is an older part targeted more at industrial customers. It has a higher operating temperature than the 982. The 981/982 parts do not support PD3.0 as they are older parts. The 983B is a updated version of the 982 that does support PD3.0 but still requires an external power path. The 987D/988 are the newest version of out parts and they have a completely integrated power path as well as more resources available to help the customer with integration into their product. We felt that there is a high value to having an integrated power path which led to the development of the 987D/988.

The amount of code to configure these devices is similar across all of them so from that standpoint you would not see any advantages to using the 981/982/983B over the 987D.

Hope this helps. Let me know if you have any other questions. 

Nathan,

Thanks for the clarification on BC1.2.  I will look at the implementation document.

Regarding code space required on an external I2C master microprocessor, is there a link to reference code provided by TI for either 87/88 or 81/82/83 so that I could have my team analysis how much code space this will take?  I did a quick search, but nothing obvious was found.  To clarify, this would be the code we would have on an external micro, that would talk with PD controller to determine what sort of charger we have, then program the chosen charger to the appropriate rates.

Thanks,

Matt