Is it safe to use regular AA batteries in the solar lights? I wonder, because when the batteries in the solar lights go bad, my dad put Sunbeam AA batteries in them. Reason being, because they're way cheaper than rechargeable solar AA batteries. Plus I bought two cheap solar lights at Walmart that are Mainstays brand but when I opened them they have mini AAA Westinghouse brand batteries in them.

But the Sunbeam batteries do work for quite a while in the solar lights even though they're not rechargeable, they must be recharging somehow.

I’ve always put rechargeables back in them. I find that AA / AAA rechargeable 1.2v camera batteries (if your batteries are the same over the pond ofc!) do an even better job than the actual solar light batteries them selves. So I’ve been buying those I usually get 3-4 years out of them before they start to go bad and not charge as well. 

I’ve never tried regular batteries in them with the fear they may leak and corrode the contacts or worse explode,

Alkaline batteries sprays their electrolyte when charged with a NiMH charger.

I doubt this will happen with the tiny charging current they are getting in those things..

Alkalines do accept some charging, the problems is that 1. They are not properly recovering like proper rechargeable batteries, there are irreversible reactions happening in them, so don't expect them to last long, 2. They have no means of dissipating the heat and gas pressure that is generated by charging, so might explode - but this probably happens only at higher charging currents

Solar lights need low capacity batteries that can fully charge with the limited energy they are getting during the day (a higher capacity battery won't get fully recharged every day, and will deteriorate from this if the charging balance over 24 hours is not positive - i.e. the full night LED use + self discharge exceed the charging), and one that can handle fairly high temperatures, which again means fairly low capacity battery for its size

alkaleaks can infact be recharged.
it must be done at low current and with top voltage limited to around 1.65v/cell.
and the ir rises with every cycle to the point of uselessness.and the probability of leakage increases too.
i now have 17 recharges on a duracell 9v.
it runs a homebrew emergency light consisting of 2 white leds and 100 ohm resistor on a snap salvaged from a dead 9v.
btw nicads are the choice for solar lights due to ability to survive overcharging and regular very deep discharges.
nimh dont do well with either condition.

Interesting,  causebIve been using NIMH  2400’mA batteries for many years in my solar yard light
My problem  low  capicity  nicad  solar yard lights
Developed memory charge problems
And also at same it seems to not have enough
Capacity and go dead by midnight.  I can get
A week of no full sunny days. It have enough
Battery capicity to handle the low charge rates
And still keep up.  Some of batteries are now
10 years old starting to act up.

There is no problem of not fully charging the cells with Nixx, the only important thing is to prevent reverse charge of any electrode when discharging, so prevent the voltage going below about 0.6V (that is when one electrode is discharged, while the other still holds something; it is the polarity reversal, what starts the corrosion and so the dendride growth on the electrode).
Because the voltage booster chip usually use Si bipolar process, which inherently looses functionality below about 0.8V, the reverse charge is inherently prevented.
It is even better for the battery life, if the charge balance is negative. Of course, long time the battery can not give more charge than it gets during charging (minus losses), the system settles so the light dies when the battery gets fully discharged, so when all the energy supplied over the previous day was spent from the cell.
The way how this helps is, when you size the cell so its capacity is larger than the solar cell may deliver, the cell will never overcharge, so will never be exposed to elevated pressure.
Well designed cells for this use are able to handle this, but still it wears them out (mainly the seal, when that fails, the cell either leaks, dries out or get poisoned by the atmospheric CO2 neutralizing its electrolyte).

Of course, this strategy means, the thing will work only part of the night, not the whole night. And that somehow limits the usability (good for decorative lights, but not good for some security light or signalling application)

In any way, a good "solar battery" is able to take quite high overcharge rate at high temperature (sunshine on the hot late afternoon). For that, you need very high excess of the water recuperating electrode. And that means for a given size, the available charge capacity of such cell is way lower than it use to be with "standard" cells (about 600mAh for an HR6, instead of the usual 2..3Ah typical for standard consumer HR6).
Of course, this works only when the capacity is low because of the water recovery electrode margin and not because of the can being half empty as typical for the cells of the "Cheepeese" provenience (it is not as uncommon to see these cans having just few turns of the foils around the outer shell and otherwise being empty; normal quality cells are completely filled with the electrode assembly roll)...

a 2400mah cell likely never hits overcharge.
so it avoids the abuse.
so if it never goes completely dead it gets lots of partial cycles.
one of my tricks is to save my "soft" cells which have higher internal resistance for my friends solar lights.
she uses them about 2 years before they short and die.
beats buying cells for 30+ lights.

Interesting,  causebIve been using NIMH  2400’mA batteries for many years in my solar yard light
My problem  low  capicity  nicad  solar yard lights
Developed memory charge problems
And also at same it seems to not have enough
Capacity and go dead by midnight.  I can get
A week of no full sunny days. It have enough
Battery capicity to handle the low charge rates
And still keep up.  Some of batteries are now
10 years old starting to act up.

a 2400mah cell likely never hits overcharge.
so it avoids the abuse.

It is completely wrong to assume the higher nominal capacity cell (of the same size) will be more robust, it is exactly the opposite.
First the internal temperature load is given by the ability of the cell to dissipate heat vs the excessive power it gets. When the cell sizes are the same, so is the power delivered by the solar panel, as well as the ambient temperature from the sun, there can not be any difference at all.
The difference could be in the internal cell construction in how high temperature it may tolerate before starting to deteriorate. And there the most likely situation would be the more robust, higher temperature construction will most likely occupy more space, so for the same can size you may fit only lower capacity electrode roll.
Similar if you want to use larger excess of the negative electrode area, allow for faster and lower pressure water recovery reaction. That means you occupy more of the space by something that does not bring you any charge retention capacity, so less of the energy storage part will fit into the can, hence lower usable cell capacity. This is the main difference between "consumer" and "solar" cells - "consumer" are optimized for maximum usable capacity, assuming only minimal overcharge (just for the cell balancing, so some 10% or so), yielding capacities in 2500mAh and more.
On the other hand the "solar" cells are expected to be heavily overcharged, so the volume is nearly full with the "extra anode area", yielding just the 600mAh for an AA size.
And another way to allow for a short time overcharge robustness is to leave a void in the can. That space will become a gas storage buffer for a fast overcharge burst, allowing the gasses to be recuperated later. That means you have less space for the actual active cell assembly. It has an advantage of spreading the heat load over longer time, but it is not able to increase the average overcharge load as the previous. This is very common method for cheap solar cells - as side effect it saves the costly active assembly, it means just leave part of the can empty.


If the solar panel has in average higher power than the device consumes at night, each day will the state of charge be higher. Till it hits the overcharge. But that means, the cell will be handling overcharge for some time. Maybe not when first installed, but after few sunny days. The only thing the higher capacity brings is the longer time it will take to reach that level.
So e.g. with a 50mA average charging current at 12 hour day and 30mA average discharging current at 12hour long "night", that means each day the solar cell delivers 120mAh excess charge, so a 600mAh cell will reach full charge at the first two daya, from the days it will be overcharged by about 120mAh/day, assume about 20% cycle charge inefficiency.
A 2400mAh cell will get fully charged after 20 days, then again it will get 120mAh/day overcharge. The thing is, into a 2400mAh cell the cell maker may fit less of the negative electrode extra capacity (needed to recombine the O2/H2 generated during overcharging) compare to a same size "solar" 600mAh cell, so because the overcharging rate is the same, it needs highre pressure buildup to get sufficient recovery reaction rate. And the extra pressure means more stress mainly for the seal, so it will more likely fail sooner (assume the difference is only in the relative electrode sizing, not in the overall build quality).


The true "memory effect" is not anything permanent, nor real capacity loss. It is just some 100mV voltage droop when youstart to discharge the cell below the minimum limit you were operating the cell in the partial cycles before that. But after this deeper discharge, the "memory" gets totally "erased".
In a real world, the true "memory effect" you will never encounter with reasonably designed devices (to utilize even the alkaline cells more than half of their available capacity, the device has to be able to work below 1V/cell even when the rated voltage is 1.5V/cell; that is way enough to completely swallow even a heavy "memory effect" build up inside a NiCd cells).
What is observed in a real world is mainly the internal resistance increase due to cell wear, mainly caused by high cell loading (mainly ase for the common digital cameras or other similar devices). This is not recoverable, but it is not caused by the cycles being just partial, but by any high current loads (digital cameras,...; anything above about 300mA is high current for an HR6) and high temperatures (overcharging, devices kept inside parked cars) and/or other cell abuse (includes deep discharge with one electrode reversal, so with voltages below 0.6V on any cell, high internal pressures during high overcharge rate,...). Dont forget the separator (so the piece of material between the electrode foils, soaked with the electrolyte) is intentionally designed to close its pores (so break the electrical conduction) at high temperatures as a safety measure (preventing cell bursting when short circuited,...; the associated heat closes the pores, so breaks the circuit inside of the main cell structures). When the cell is exposed for high temperature, the pores start to slowly close down, so gradually increase the cell internal resistance. This is the most common EOL mode for NiMH cells I've ever seen, often misdiagnosed as a "memory effect"...