Problem with AW C size Li-ion cells

My AW C size Li-ion cells do not seem to have any capacity left.
They used to work well in my Mag 85 and 1274 but since I tried them with a AW soft start switch and an Osram 12V 35W IRC (64432) (I used 5 of them to drive the bulb at 18V and about 4A) they do not work any longer with these set up.
When fully charged , the cells cut off after about 2 minutes runtime.

I have tested few things to try to understand more about the problem:
I have measured the current drawn at 18.5V and it is just under 3.95A for this bulb. This also match Luxor bulb testing, he found 4A at 18V.
I have done further testing with the cells and a WA1274 bulb fitted in a standard mag switch. Starting voltage of the cells 2x 4.1V = 8.2V (No load).
I measure 2.75A current at the begining of the test, it take 3-5 minutes for the cell voltage to drop to bellow 6V, around 5.5V it cuts off. Which I think is normal. I have repeated the test with 2 sets of cells, same results.
After the test when I switch off, the cells get back to more than 4V which indicates that they are still charged?
I have also tested one single cell with a WA1274, the current measure 1.8A to see if at low load things improve. The cell voltage again drop quickly (about 5 minutes) to about 3V at which point I stop the test.
I tried to charge them again using a Ultrafire charger and repeated the tests but same results.
Has anybody had problems with these cells?
Do you think the cells might have been domage? Does somebody knows what could be the problem?
What voltage should a Li-ion cell be when discharged but not loaded? Could my cells not be charging properly?

Hello Luminous,

It sounds like your cells have developed some higher internal resistance...

A Li-Ion cell when drained should see the voltage rise to around 3.5 volts, or at least over 3 volts depending on your low voltage cut off and the current draw. You indicated that your cells are charging up to 4.1 volts, so that's about 90% of full charge, and indicates that your charger is doing its job.

The next step is to check each cell individually to see what voltage they drop to under load. You may just have a couple of cells with higher internal resistance. If they all are about the same, then you should use them in lighter load applications...

At around a 0.5C load, the cells should hold about 3.7 volts.

Tom

The thing I do not understand is that the cells never used to behave that way. Is it normal for a cell to increase its internal resistance over its life? These cell may have had about 20 charges and are about 1.5 year old so they are not that old.
Is there anything I can do to get them back to normal? These cells are suppose to be good up to 5A and now they cannot hold their voltage at 1.8A!