Hi, I was wondering what would be the best way to power the photon from a 3.7v (3.3-4.2v discharged & charged) LiPo source
I wanted to upgrade my electric longboard remote from an arduino with a 2.4ghz transmitter to the photon
Looking at the docs, I can power the photon by either 3.6-5.5v at the vIN pin or 3.0-3.6 at the 3v3 pin
Is there any small form factor solution that I can use to utilize the battery to its maximum potential with minimal wasted energy? I have barely any space left in the remote control
I am using a 600mA lipo battery so efficiency is key
Thanks Peekay but the shields unfortunately have a too big footprint to be useful
I’m fitting everything in a WII nunchuck controller.
My solution for the arduino was to use a 2S 7.4v LiPo that was externally taped to the nunchuck, i’m trying to integrate everything inside the contoller
@Sleziak, I definitely agree that this has been a tricky issue for quite some time.
I really like these Pololu 3.3V/0.5A buck/boost regulators that I've been using for a while:
The trouble is, those do not provide in-line charging capability, as @peekay123 said, you'll have to yank the JST connectors out just to charge the battery. I think the Sparkfun boards are perfect for prototyping, but I too do not like their size when you have to cram them into tiny spaces. Even providing us with the option to solder our own choice of lower-profile headers would have helped.
I also wonder: when will we see the final design of the Particle Power Shield?
From the posts, they look similar to the Sparkfun variants, but might have support for solar trickle charging?
The one other solution is to consider the Powerboost 500C from adafruit. (and it's bigger sibling, the 1000C. You don't need to solder the USB jack on (they don't come soldered on); but since they provide +5V output, you'll have to connect to VIN instead of 3V3. This for me is the most compact solution I have come across, as you can lay out the boards end to end and get a lower profile (6-7mm height).
That said, it's an annoying (not to mention inefficient) method to connect power – with the 3.2-4.2V LiPo boosted to 5V, then bucked back to 3.3V on the Photon's onboard regulator. But if you want convenient LiPo charging integrated on your board, this seems to be the only way I can find so far and it's worked for me.
The Photon comes with an impressively efficient 3.3V regulator, so I would have no qualms feeding it 5V via USB / Vin.
I would use an Adafruit 500C powerboost to get going, and using their circuit as a reference if you are building your own board.
Li-Po’s are very sensitive to charge rates, minimum voltages, etc, so using a decent charging device is important. Who cares about the charging efficiency unless you are using solar? It is the 5V to 3.3V conversion from the battery supply that matters, and 95% efficiency is plenty good as provided by the built in Photon regulator. Smart use of sleep cycles will save you more than the regulator loss fairly easily.
The Photon does have an efficient 3.3V regulator, (so does the Adafruit 500C’s TPS61090 boost regulator, 90%+ @ 500mA), but it’s just me feeling icky that the raw battery voltage gets regulated twice to power up the Photon. Yes there’s the 5-3.3V via the Photon regulator, but prior to that there’s also the boost circuit from a single-cell lipo from 3.2-4.2V. Hence the annoyance.
There are plenty of LiPo charging circuits, OR buck/boost regulators. What’s rarer are combinations, and the 500C/1000C is the only one I know that seems to be readily available and accessible for most of us.
I’d say a combination of judicious sleep cycle usage, and selecting the ideal charging-and-boost/buck circuit for the Photon goes a long way in determining just how long you can keep it a low-power project running on a single charge. Wasn’t referring to charging efficiency here. on that note, the Power Shield thread does have discussions on making the design compatible with solar charging, so it’ll be curious when that comes out.
Save from making our own compact 3.3V charge+power circuit, are there other alternatives out there? Would be good to know what else is out there.
I’m currently using a LiPo Rider Pro (http://www.exp-tech.de/seeed-studio-lipo-rider-pro), connected to a USB solar panel and a LiPo battery (currently 6600 mAh but my old Galaxy Nexus battery works too), it also works with 5-6V solar cells directly and doesn’t drop power when removing either power source. Today it is a little too cloudy to charge much but yesterday it nicely charged the battery while powering my Photon. Also it has been running for over 24 hours with WiFi connected. Granted the USB solar cell has 14 W but then again its just hanging behind the window and the sun is at a less then ideal angle for this setup in the summer so it doesn’t do more than 2-3 W at the moment.
Nice size, and the I2C fuel gauge is a plus! Definitely looking forward to these boards.
On a related note, a student of mine came across these very affordable TP4056 LiPo chargers on eBay that he’s using to power other mcus:
Like most cheap boards, there’s always a degree of risk involved, especially with LiPos. These boards are similar to the Pololu ones in that they do not offer step-up/down power regulation, so the LiPo can only provide useful power from 4.2V down to 3.6+V before the Photon brownouts.
Having a project needing more power with the Photon, I bought this 4000mAh battery that worked fine with the above TP4056 on projects with arduino:
But here, it appears always as charged but they do not deliver energy to power the photon. When I mesure it it is 3.890V max, I guess the PCM handling the protection.
Do you know a difference between the Particles LiPo and those others ? How could I use them with the Photons ?
Hi @peekay123 reviving an oldie but goodie. Apologies in advance if I should have started a new thread instead.
You mentioned powering the Photon with a LiPo connected to the 3.3V pin or through a boost converter.
If we go direct to 3.3V, do you need to run the LiPo through a voltage regulator like a LD1117V33?
I’m assuming that the Photon will be flakey as the battery approaches the discharge cutoff (2.75V on the Sparkfun LiPos), which is why you suggested the LiPo boost. Does the Particle Power Shield already have a 3.3 boost converter on it, or do we need to provide our own? I can’t find anything in the docs or the GitHub repo.
I remember something about the old mustache battery shield for the Core supplying 5.5V to VIN which was wasteful. Does the new Power Shield connect to 3.3V?
If I’m reading the schematic correctly, it looks like the Power Shield connects to 5V VIN on the Photon. I’m not seeing a boost chip like the TPS61200, but maybe that capacitor + inductor is acting as a boost converter?
@geoff, there isn’t a boost chip on the board. The only switcher is the INPUT regulator to convert higher voltages to +5v as input to the PMIC. The PMIC will manage the battery charging and feed to the Photon Vin depending on the external voltage source, if any. The fuel gauge chip is necessary for the Photon to monitor the battery to make sure it doesn’t over discharge. The Photon’s onboard 3.3v regulator can only go down to 3.6v at Vin so the Photon should shut down at that point to avoid problems.
In short, the battery feeds the Photon Vin directly and any devices requiring 5v will not necessarily work correctly on battery. There are other boards available to do this including the Adafruit Power Boost 1000.
Why not just use a 3.0v voltage regulator and connect it to the VBAT pin? A through-hole voltage regulator and two caps would be very small and not require a circuit board even. I believe that’s the purpose of the VBAT pin, no?
@teckel, as @ScruffR pointed out, VBAT is an INPUT that allows a battery backup supply to be supplied TO the STM32 processor to maintain the onboard RTC and a portion of SRAM designated as “retained” RAM in the Particle firmware.
Got it, so it’s not designed to run the Photon, but to just maintain state, and the microcontroller and WiFi chip are haulted. Once power is returned to VIN the microprocessor and WiFi restarted and the code resumed. It seems the VBAT definition (at least the one I read) wasn’t clear that the microcontroller was haulted if VIN was disconnected with voltage on VBAT.
You can also supply 3.3V to the 3v3 pin, but the stability/quality demands for 3.3VDC are higher.
Also cutting the Vin/3v3 supply does not halt the controller but it stop completely and start from a reset state
