USB-PD DRP from a single LiFePO4 cell

[ewinek]

I am about starting a project with LiFePO4 pouch cells. The idea is to create a powerbank from a single cell. The powerbank must have an input/output USB C connector with 65W minimum power (100W ideally). The idea is to charge laptops or charge the cell from the same connector.

I am searching all around and I don't find what I need to do such configuration. There are manufacturers like TI that have all-in-one solution (DC-DC+PD controller) but that chip works only as source and it is limited to 60W.

What I understand so far (maybe I am wrong) is that the PD controller handles the negotiation with the connected device giving a signal to the DC-DC to output the right voltage. However I don't get it yet how this is done, because then the DC-DC must be:

Adjustable

• output
• buck-boost
• Compatible with the PD controller
• bidirectional (for the application I want)

The PD controller must also know which voltage is needed for the powerbank battery cell, if it is configured in a 3S Li-ion then the voltage for the DC-DC given to the cells must be 20V, if it is only one cell then it can be 5V.

Any advise?

Thank you very much in advance!

[David Hess]

Some laptops support this with their USB-C PD ports; I know my Thinkbook does.  But I think it is more involved than a single chip controller, with an external microcontroller and custom firmware.  On mine it also appears to regulate the output current based on how much extra charge current is available.

[ewinek]

Any guides or ideas how to proceed further on this project? I am a bit lost at the moment.

Thanks!

[inse]

100W from a single cell is ~27A, are you sure you are able to handle this?
That is 0.7W dissipation in every mOhm already.
Did you ever design a SMPS?

[inse]

Dead-end topic?

[ewinek]

Hi,

Actually is around 31A, a LiFePO4 works with 3.2V nominal value. I have cells of 31A with capabilities of 1C discharge, I have cells with 20A/16A with 3C discharge capabilities. The challenge for me is more in the logic of the circuit more than in the power, I can split the current over many MOSFETs and produce a decent DC-DC. I have design few in my life, but not this exact type.

The reason to do it with one cell is to avoid the BMS, avoid cell balancing, space/size and connections. Moreover, there are no many available options out there with LiFePO4 technologies in terms of BMS and battery chargers.

Some hints? All are welcome

[brumbarchris]

Apart from the power circuit, you will also have difficulties with the USB PD implementation and compliance itself. It is obviously not impossible, but only the spec itself is 800+ pages and there is generally quite a lot of firmware involved. I am going through it for a project right now and it is quite daunting!

Good luck with you work!

Regards,
Cristian