James Sweet wrote:


Much debate has raged over the years on the topic of NiCD memory
effects, and to this day it remains controversial as to whether the
memory effect exists. My own experimentation suggests that the perceived
memory effect is due to poor quality chargers that severely overcharge
cells in the specified time unless they are discharged first, leading to
a dramatic reduction in lifespan.

I've seen it with NiCads. You charge one and it appears to be
discharged after brief use. Then you put it in a holder with a 2-ohm
shunt, and it keeps producing current, but the voltage is low because
the internal resistance is high. With the high internal resistance, it
may take a long time to discharge completely. Completely discharging
gets rid of the high resistance.

It appears that if a NiCad is left charged or partially charged for a
long time, a chemical change will cause high resistance in the charged
part, but discharging will correct it. Lead-acid seems the opposite; if
it is not routinely fully charged, the discharged part will develop high
resistance, which can be corrected by fully charging, which can take a
long time due to the high resistance.

NASA says NiCads should be drained and shorted if not needed
immediately. When you get ready to use them, the first step is to
overcharge them, 40 hours at C/20.

Sometimes a NiMH may not charge fully. I think that's from gas bubbles
and not memory effect.

On a related note, years ago I got tired of the
backup battery in my digital clocks always being dead whenever I needed
it so I did a simple modification and installed a 1K resistor across the
isolating diode for the backup battery resulting in the power supply in
the clock providing a steady 3-5mA of current to a NiCD "9V" (really
7.2V) battery. This has been working great, and more than 6 years later
the original batteries in the four clocks I modified are still working
great despite being trickle charged steadily for those years.

In my experience, keeping a NiCad charged eventually results in memory
effect, which can be corrected. If you removed one of your batteries, I
wonder if it would provide 8.2V across 150 ohms for a couple of hours.
If it won't, that could be memory effect.

In my experience, slow or trickle charging NiCads causes electrical
leakage and eventually shorting. Internal electrical leakage can be
detected by testing capacity after two weeks off the charger.


In a nutshell, at high enough current level to charge a battery quickly,
180-500mA or more being typical, charge time is very important and once
the cell reaches full charge that energy will go into heating the cell
rather than the chemical reaction that stores energy. If you keep that
up, the electrolyte will vent and permanent damage will occur. On the
other hand, if you charge a cell at a low current of say 5-10mA, you can
charge it indefinitely without damage.

I think the heating comes from the gases recombining. I wonder if the
heat can damage a cell even if there is no venting. (I like to check
temperature with an infrared thermometer.)

They both have a nominal cell voltage of
1.2V, they both have a recommended standard charge current of C/10,
where C is the mAh rating of the cell, and they both need to be charged
by a current limited source with an OCV higher than the nominal cell
voltage, and both can be trickle or float charged indefinitely at C/100
or so. This is based on my education, research, and years of personal
experience. If you insist that this is incorrect, please cite sources.

Are C/10 and trickle charging really recommended for NiMH? In my
experience, both kinds of cells do better if I stick to
microprocessor-controlled fast chargers.