Set minimum current limit lower than 1A

[alexkost819]

Hey! Great product. Glad I bought one and have it :)

I really wanted to use the PocketPD for board bring-ups but I cannot because the 1A current limit minimum may fry components on my board. Ideally the PocketPD could support lower current limits, but if that's not possible, the current limit range should be mentioned on storefronts/Hackaday site/wherever relevant so buyers are aware of this limitation. Right now only the max current output is mentioned.

Apologies if this is obvious to everyone in the community already. I vaguely recall reading the PPS spec mentioning current below 1A is not supported, but I also feel I read things that mentioned 50mA steps are supported... suffice to say I'm no expert here.

Thanks!

[scotteshaw]

I recently stumbled onto the same thing. Based on my reading around the minimum PPS current is indeed 1A. I can think of a couple of paths.

1. A software driven CC mode could be implemented for currents lower than what is supported by PPS. Could monitor current and then lower the voltage to limit current. This would, of course be limited down the 3.3V and the response time could be limited to whatever speed data could be practically acquired from the INA226. Probably wouldn't really help too much in for the "I'm bringing up a new board and 1A would kill the board" since in many cases you'd probably hit the 3.3V minimum.
2. If a current limit below the PPS 1A minimum was selected the Pocket PD could switch to a mode where it just cuts off if the current is over a limit. The response time would be limited by whatever maximum effective rate the INA226 could be sampled. This would probably be a lot easier to implement than 1 and would be really useful, particularly in my use case.
3. If the hardware was modified I think it should be possible to make use of the alert pin rom the INA226 to either interrupt the uC or perhaps tie back to the LM73100s to increase response time. I didn't dig in detail on the INA226 to really understand the how it works though.

Thoughts?

Edit: LOL, It turns out I basically just restated what was posted in a blog entry a long time ago! https://hackaday.io/page/21022-variable-power-supply-with-usb-c-pps-picopd

[centydev]

Hi @alexkost819,
100% Agree on that we should clarify the spec of PocketPD with PPS. We will update the posting and related document shortly.
We will also need to write a post regarding PPS profile so reduce confusion on charger brick.

@scotteshaw
We did a review last week and saw that we indeed could have use the INA226 INT output, pass it through some kind of flipflop with RESET so we can directly turn off the LM73100s to increase response time. Put an EN able somewhere to allow this. We have planned to prototype this circuit in August and hope to be able test it better in the following month.
Lessons learn the hard way that we should just route all pin to the uC if there is extra. Firmware is much easier to update.

[centydev]

HW1.3 has some additional circuit that allow us to set a threshold to turn off output switch with current limit. However, the response is faster than MCU based, but much much slower than standard short-circuit protection. Example LM73100 is 500ns response time.

Done by passing INA226 ALERT signal to LM73100 Enable signal

[dreirund]

HW1.3 has some additional circuit that allow us to set a threshold to turn off output switch with current limit.

Are the pins that needs to be connected present on HW1.0 as solder pads, so that I can add the circuitry to my HW1.0 as well?

(I already did upgrade my HW1.0 to "quasi-HW1.1" by making the current sensing resistor 5mOhms.)

[centydev]

Hi @dreirund,

We are wrapping the hardware to release under CC4.0-BY-SA license. Please check back on the PocketPD-HW repo next week. There will be KiCad files.

[centydev]

Link to #42