Hi there,
I have a power supply hardware design question for Boron / Argon / Photon2:
The power circuitry for the Boron looks as shown below, where VUSB can go as high as 17V:
The circuitry for Argon / Photon2 is similar but the diode is placed slightly differently and VUSB cannot exceed 6.2V (Argon) or 5V? (Photon2):
I have a design where a) I would like the option to power from an external source VEXT as high as 16V and b) I need to generate 5V for other purposes. My first attempt looks like this:
Having built the circuit, it works pretty much as expected when VEXT is greater than Li+ and it can successfully / safely power Boron, Argon, Photon2 from VEXT. However, if there is no external supply (VEXT=0), then the 5V is generated from the LiPo supply and I had not fully considered the implications of having this 5V connected to VUSB (dotted line in the figure).
In particular, the PMIC now sees two power sources, the LiPo and the generated 5V. The orange charging light turns on, suggesting that it is using the 5V source to recharge the battery from which the 5V source is being generated! So, one question is whether this will actively cause damage, or (best case) if it will just discharge the battery due to inefficiencies in the ‘self-charging’ loop?
The other question is how to fix this. I imagine in software I could disable battery charging if VEXT less than LI+. This doesn’t require hardware modification but it is not easily automated and puts the onus on users to take this action when needed.
Another possibility is to modify the hardware design to look as shown below which breaks the ‘self-charging’ loop but now limits the (no longer regulated) VEXT to the specified supply voltage limits of the Boron / Argon / Photon2 respectively (ie. it doesn’t fully satisfy goal a))
I know there are a lot of very smart people out there + I feel a little embarrassed to be airing my design failures in this way, but any thoughts / comments / suggestions would be much appreciated.
This is a scenario supported by the Tracker One, so you may want to look at that for a reference. The solution we used was to put a power supply in front of VIN for external power. It's a little different than the Argon/Photon 2, however, because the Tracker only supports the Tracker SoM which supports 17V on VIN like the Boron.
In your case, your external power supply could be a voltage regulator than converts down to 5V, which powers VIN/VUSB on the devices you have selected.
This does not, however, power the 5V for your uses. For that, you add a buck-boost converter. This is like the CAN_5V buck-boost converter on the Tracker. This is powered using OR-ing diodes from Li+ and VUSB for the Boron/Argon/Photon 2. This allows powering from battery, USB, or your external supply, while preventing the loop you described in your circuit.
Thanks for the quick and helpful reply! I took a look at the Tracker1 and it does seem a bit different from the Argon / Photon2. But for my case, it looks like you’re suggesting replacing the single buck/boost TPS63070 in my original circuit with two buck/boost regulators, one to power the system, the other to generate the 5V I need for other purposes. That would indeed avoid the loop, but at the expense of more components. It would look something like:
a) Is this what you meant?
b) Do you know if the undesirable ‘self-charging’ in my current design will actively cause damage, or will it merely cause battery depletion (hopefully not too fast)?
That is what I meant, using two converters. I can't say for sure whether a single converter will work or not, but there are many edge cases that make it questionable. Also it can behave very differently between the Boron (bq24195 PMIC) and Argon/Photon 2 (no PMIC, just a charger chip) so make sure you throughly check both cases.
Any thoughts / comments about this for another possible solution? It might be better to power the comparator from the generated 5V but I was a bit worried about input voltages that exceed the comparator's Vss (though the AP331 is quite tolerant).
