You only have to convert ~4.6V (peak, when we have direct solar pass-through) down to 3.3V with a device that mostly only consumes around ~2-4mA in Idle. It’s not worth the hassle adding a BB.
The design is also using a very modern LDO that doesn’t consume much power. Also the drop-off under load isn’t that big. It’s no a shitty AMS1117.
I was finding around 7-9ma in RX with the max boost then 130-160 in transmit for just a moment. Which is comparable to the rakwireless module I’ve been testing. The Ram unit definitely goes to sleep better when used as a tracker, something wrong in the HT-CT firmware not allowing deep sleep.
At about €7 for the HT-Ct and about a euro for the charge controller/LDo module it’s cheap and seems to run well from a 6w 5v solar panel and some cheap 18650. Complete build with antenna, battery, printed enclosure and 3m pole coming at around €30. Cheap enough if they get stolen or destroyed I don’t really care.
With one of the next revision I’ll replace the CR123A battery with a “classic” 18650. Turns out that there’s no big market for CR123A 3.7V LiPos and the size limits them to a maximum capacity of ~800mAh. Also there are a lot of fakes around.
The BQ25185 should have enough safety feature to even use LiPos without protection circuit. Maybe I’ll going to test this in a metal tub.
Have you done any tests to see what happens when the device runs flat? Does it re-start, when solar charged again, from when it has run flat? Does a protected battery perform differently in this regard to an un-protected one? Thanks for any info. Love the project.
Out of interest how are your programming it? Assume can flash firmware to it over the UART connector? Presumably use some sort of CP2102 or similar USB->UART convertor?
Hm, there’s nothing to add to get native usb. So it’s even simpler than a UART bridge.
You just need to wire Pad18/GPIO19 and Pad17/GPIO18 to USBDP and USBDN respectively. No extra hardware needed (unless you want to add TVS diodes for ESD protection, perhaps)