Hello everyone,
I have read several threads about the 18650 battery shield and noticed some confusion over how they work.
These boards are available in several versions - some have a normal/hold switch and some don't, some have 3x 3.3V and 5V outputs while more recent have 5x of each. They are available with a single or dual cell holders - I have standardised on dual-cell for the greater capacity this provides. My boards are not marked with a circuit revision number - which is a frustration, but two have the normal/hold switch and one doesn't.
They are touted as suitable for running Arduino or Raspberry Pi systems and the dual voltage output makes them very attractive.
Some operating pointers:
when cells are first installed the units will not turn on until a charge cycle is performed
a short button press will turn the board on with the battery status indicated on the LED's - this also activates the output voltages and the outputs will remain active as long as a minimum current is drawn
the normal/hold switch can override the drawn current requirement when in the 'hold' position so the minimum current requirement is not required (so the board remains on)
if the board is on then connecting a charger does not appear to interrupt the output voltages (my Arduino projects don't reset)
If the board is being charged while supplying an output and the charger supply is removed (either by disconnection or mains failure) then the board turns off the outputs for several seconds and then turns it back on
A long button press will cause the board to shut down
Point 2) It is likely that the internal sense resistor could be altered to adjust the current level required to remain on
Point 5) This is a real bug bear and precludes using these boards as a power bank as whenever the input voltage is removed it momentarily removes and then restores the output voltages. Comments have been made about this in other threads but I didn't locate a solution. I have the v3 circuit but does anyone have the current (v8?) circuit. A modification that allowed power feed through but retained the ability to protect the cells from over discharge would, I believe, be welcomed by many users!
I have a power bank and I will trial how it behaves with respect to removing the charge while supplying a load. Has anyone tried a commercial power bank in this regard?
@Graham,
Do you have any links ?
Also , in your title you refer to it as a shield, but the one Terry has is a breakout board (not a standard
plug in shield) Are yours shields that plug into the Arduino UNO R3 ?
The term shield comes from the marking on the board itself which reads 'Battery Shield' and it is a stand-a-lone item not designed to piggy back another board. This naming does lead to confusion...
Here is a link to a single cell version:
I also trialed by commercial power bank with an Arduino project to see what occurred:
Charger source removed - output power is turned off
Power output remains off
Charger source reconnected - output power turns back on, Arduino now runs
For our purposes of powering an Arduino or Raspberry Pi with uninterrupted power this is less than ideal behaviour but is probably not too bad if the power bank is being used to charge another device that has internal power storage.
Back to the original question that prompted the thread: does anyone have a modification to these 18650 battery 'shield' so that they function more like a UPS?
I have one of the single-cell V3 modules like the one in the electroschematics link, and even made a video on how to modify it to make it fully functional as a UPS. It has familiar chips like the TP4056 charger and separate 3V linear regulators and a 5V boost regulator.
I also ordered a 2-cell version that was supposed to be a V8, but what I received had no V marking, and didn't look like the listing picture. It has a single mystery chip that does both the charging and the buck conversion. There's no minimum current shutoff, but it does shut off for a couple seconds when you remove the charging power supply. So it's not usable as a UPS.
I think as time goes on all of these "shields" will come to use the single-chip solutions, because they are cheaper, and they are unlikely to be usable for UPS. And I don't think we will be able to hack them so they work right.
The original V3 does seem to work ok, but needs to be modified if the project uses significant current.
The main device on my boards, which I presume are V8, is a TP5906 which is a charge regulator and 5V boost converter in a single package. The 5V output is totally under the control of this device and, sadly, I agree that these boards are too 'advanced' to modify into a UPS for something like a Raspberry Pi.
I watched the link you provided and it seems that the earlier V3 boards with the TP4056 do provide a path to convert into a UPS. Of course, who knows how long V3 boards will remain available!
A quick search located several DIY UPS systems but they typically talk of lead-acid and steer away from using 18650 style batteries presumably due to the problem of protecting against over-discharge.
Interestingly, I do have a couple of solar charger boards which offer charge regulation with battery protection which may well be suitable for a UPS. I'll do some reverse engineering and testing of it over the next week or two.
KiwiGraham:
Interestingly, I do have a couple of solar charger boards which offer charge regulation with battery protection which may well be suitable for a UPS. I'll do some reverse engineering and testing of it over the next week or two.
Do you by chance have a solar charger that uses the MCP73871? That chip supposedly has power sharing built in, and would seem to be ideal for a UPS. But I don't see any modules using it which also have convenient connectors/holders and a boost converter.
Hi again,
As far as I can tell my Solar Charger V1.0 board uses the CH3065 device which is a small 8-pin part.
The MCP73871 looks like a good candidate - an initial look through the datasheet and I suspect it can't disconnect OUT when the cell reaches it's minimum voltage level. It does have a status output to indicate that the battery is in a low charge condition. Section 4.2 talks of allowing the OUT pin to have power even in the condition where the cell is deeply depleted. The device also accepts a thermistor for temperature measurement (useful). The chip has Voltage Proportional Charge Control which allows control of chow the charge progresses. I like this device for many reasons!
AdaFruit have a carrier board with this device which can accept input from USB or solar (usb-dc-and-solar-lipoly-charger.pdf) but this doesn't appear available in New Zealand yet in this form or been made into other boards. Will have to exercise patience!
If you get the chance to run some experiments with it I would be keen to hear your findings.
Adafruit parts are too expensive for me. But there are some MCP73871 modules on Aliexpress and Ebay. However, I believe they don't have connectors built in.
ShermanP:
Do you by chance have a solar charger that uses the MCP73871?
I have a couple of these 18650 modules, for which I have soldered a connector over the 5V "power in" USB and uses many 6V cheap (1$) solar panel 1W (haha, who believes it?) modules in parallel to load.
It works fine, excepted that the cheapo solar panel modules are crap and their resin won't stand the UV irradiation more than one summer.
Hi Guys,
Just noticed that Adafruit have another universal load sharing board available based on the BQ24074 from Texas Instruments. WRT the MCP73871 it has some improvements such as wider input voltage limits and provides a load stabilised output at 4.4V.
The carrier board is $9.95 from AdaFruit.
I also noticed the MCP73871 on AliExpress for $2.55.
