Electronic – TP4056 lipo charger output voltage

The circuit you show is essentially OK in general operation but has the significant disadvantage that, as described, it deep cycles both batteries and reduces batterry life, even if solar energy available > required load energy. You can largely overcome this by cycling when Vbat falls only slightly. The changeover switch can be automated using 2 x MOSFETs and a comparator.

If switched manually you will get a changeover "blip" which the 50 uF will not do a lot to limit. 50 uF will drop 1 volt in 50 microseconds at 1A so in 25 uS at 2A. Your switch would need to be "rather fast" to achieve this. The comparator plus two MOSFETs as switches solution 'fixes' this with a suitably fast changeover. However, there are potentially better ways.

If you connect the battery to the load via an on off switch and the charger to the battery directly (assuming an internal diode or equivalent) the circuit will work OK in most cases. Problems might occur if you some advanced charging system (eg MPPT of some flavours) but in most cases there should be no problem. There may be some interesting "boundary conditions" (see below).

Consider:

Assume:

• Charger capable of > 2A - say 2.5A.
• Battery max Icharge = 2.5A (set by desihn by charger)
• Load = 2A.
• Battery say 50% charged.
• Charger capable of proper CC/CV LiPo charging with Vmax = say 4.2V and tail to say 20% of Imax = 20% x 2.5A = 500 mA.
• The solar charger should be able to charge a worst case discharged battery safely - but that is really outside the scope of the question.

Charger sees load of semi-charged-battery + 2.5A = > 2.5A. Charger "does what it can" and supplies 2.5A.
Battery charges in CC mode at 2.5-5 = 0.5A.
System voltage = battery voltage will rise as battery charges.
When /if battery reaches max voltage (ie Vbattery = say 4.2V) it will revert to CV tail (4.2V)

Boundary condition: As mentioned above, the presence of the load current hides the battery "tail current" state from the charger. As follows:
Charger now sees 2A load current + battery CV tail current.
As the load current swamps the battery tail current the charger will never "trip" the battery charging and charging will continue indefinitely as long as battery tail current + load current > 500 mA. If you charged the battery "all day long, everyday" in this mode and I load always > 500 mA then the battery would get damaged. But if load is occasionally removed or reduced to<< 500 mA the charger will trip off.

How much this matters depends on load and charging characteristics and a look at the typical load vs time situations will allow you to assess what should be done.

There are various "work arounds". One easy one is to stop charging at Vmax with bo CV current tail. This reduces available battery capacity t about 80%-90% of what you'd otherwise get - and usefully extends battery cycle life.

The switch changeover system is not without its bad effects on the batteries.
If you discharge to say 3.3V you are effectively doing multiple deep discharge cycles and battery cycle life will be low.

If Iload < Icharger_available then you COULD run the load on the solar charger with the bttery uninvolved. However, the changeover switch system does not account for this and cycles the battery unnecessarily and reduces its life.

The following is an "idea starter" only.
When Vp[v is high enough it will supply the load directly.
If Vpv is high enough and if the battery requires charging it will also charge.
Diode D1 is shown as an 1n5819 but ideally would be a MOSFet arranged as a "zero voltage drop" diode. If panel does not provide enough power for load battery will controbute.
This circuit "needs work" but has the makings of a useful system.

^{simulate this circuit – Schematic created using CircuitLab}

.

Protection circuits are usually distinct from charging circuits. Many battery packs are designed with the intention of being charged by a dedicated unit that will control the charging process.

The charging process may involve cell balancing, if the pack contains a large number of cells in series, generally 4+ cells in series (4S, 14.4V) nominal will require balancing, 3S and lower it's also a good idea to balance for the health and longevity of your battery, but not necessarily required. Balancing circuits can get complicated and typically involve a BMS (battery management system) composed of a dedicated IC and multiple external mosfets, There's a project on github that seeks to create an openBMS system. This might be a good resource if you are looking for more information.

The charging cycle for lithium ion batteries can be quite complex, especially in the case of multiple cells in series, but typically involves 4 basic steps:

• Read voltage, if lower than a certain value (typically 2.8V or so for Li based cells) then begin trickle charge until cell reaches safe charging level, doing this avoids damaging the cell.
• constant current charging: the cell is charged at a constant current, typically .5C-1C for normal charging, e.g. for 1000mAh battery, charge at between 500mA-1000mA.
• constant voltage charging: once battery reaches a certain point (typically about 60% of total charge (3.8V or so) begin charging at the target final voltage (4.2V for normal 1000 charge cycle expected life) you can go higher, and it will give you more battery life, but it will reduce the life of the battery.
• Maintenance charging: The battery has a natural discharge rate in the neighborhood of 8% at 21 °C per month, when the battery drops below 10% of full charge, recharge to target voltage using constant voltage charging. This is configurable depending on application.

Notes: there are ICs that will handle the bulk majority of this for you, otherwise you will need to resort to designing an MCU controlled circuit with external boost/buck converter, or linear regulator.

The protection circuit (PCM) is fairly simple and often times integrated right into the individual cells, these cells are typically labeled: protected, or unprotected. The PCM will monitor things like: input voltage, output current, cell voltage, temperature, etc. They often times are not that robust and should be considered a last resort in a critical system. Alarms should be going off if the PCM is ever triggered.

To answer your specific question: the DW01 is optimized to favor the charger in the case of overcharge, so the charger would remain connected to the circuit, suppling nessesary voltage, while the battery is disconnected, TP4046 looks like it is designed to handle up to 8V and as a linear charger it will be dissipating excess voltage as heat. These ICs are often thermally protected and will automatically shut off in the case of excess current or voltage. If there is an overcharge, it means the circuit is likely destroyed, thus it is nice to try to disconnect the battery in this case, as it would would pose a significant safety risk.

Also, remember that mosfets contain internal diodes, so even if charge protection mosfet were to be turned off, the battery would still be attached to the circuit as long as the voltage on the drain side of the mosfet is below a certain voltage.