really weird Li-Ion charging phenomenon... please advise

Ok.... so first a list of things I have here that I have become fairly familiar with:

6 17500 protected cells, 4 AWs "new" and 2 DLG brand.
5 17670 protected cells, 3 AWs "new" and 2 AWs "old blue"
2 16340 protected cells, 2 AWs "newest new" high current
2 18650 protected cells, 2 AWs "newest new" high current

I've used the DSD and Ultrafire(AWs version with solder tabs) chargers a lot now, both chargers have performed near flawless performances under supervision. I always get cells ~4.15-4.23V. Smaller cells tend to terminate the charge closer to the 4.15 while larger cells tend to terminate closer to 4.20, My DLG brand cells occasionally come off as high as 4.23 for some reason. this doesn't bother me though, seems pretty reasonable...

My problem is the 18650s.. has anyone else seen this phenomenon:
I use the cell or cells in a flashlight, works beautifully. I put the cell on the ultrafire charger and it begins charging. When it terminates (well, goes green that is) I pull the cell and test voltage, 4.00. but the charger says it is done, and If I slide it back on the charger, it immediately goes green. Remove cell, still tests 4.00V...

Here's where things get really weird.. So I take the cell that is testing 4.0V, and put it on the DSD charger, it turns green immediately upon insertion. Remove cell, Retest voltage, now the cell reads a perfect 4.20V.

What's really weird about this is that I can replicate this on both cells. It doesn't seem to be a fluke of any kind.

could this have something to do with the protection circuit not "tripping" closed circuit all the way because there is no significant load when under a voltage test.

so far as I can tell, the cells are fine. The charger is fine... any ideas?

update:

it also "jumps" up to 4.2V (from 4.0) if I use the cell in a flashlight for a moment then retest.

It's probably some type of open circuit protection for the Portection circuit. Thats just a guess.

I don't have the DSD, but doesn't it put the cell under an initial slight load to test its condition/state before charging? The cell is fully charged to 4.2v from the Ultrafire but for want of a better metaphor it has gone to protected sleep at 4.0v and the DMM doesn't wake it up properly as it isn't putting a load on it. The load from the DSD "activates" the cell and it then gives its true reading. Don't know why this doesn't happen with the other cells though. Maybe being smaller your DMM gives them just enough of a jolt. If it's not that, I don't know!

I could think of quite a few things but I believe the bottom line is the protection circuit is causing this and as DM51 said it is probably having a slight load put on it that brings it up to 4.2 volt and your DMM does not provide any load or an extremely small load (usually muliti-MegOhm so almost nothing). Seems just a bit of an oddity of the protection circuit and probably nothing to be concerned about. I don't have any (but may soon) or I'd check it out.

hmmm... well, for now, I'm going to assume the cells are safe to use... and bring them caving tomorrow. Hopefully all goes well... Being used in a single cell application, so there shouldn't be much risk if any.

well, I brought them caving, used 1 of the cells for about 20-30 minutes worth and it worked fine.

any other thoughts on this matter?

They must be fine. It has to be something in the protection circuit of the 18650s that just needs a bit more of a load on it to show the true voltage. You could ask AW about it.

Thanks for Eric for pointing me to this thread. There are a few things you can do to test the protection circuit to see if it is functioning like it should.

1) run down the cell until protection cut off and check the voltage;
2) ' flash - amp ' the cell with a multi-meter to see if you can get a 0.00V reading;
3) use a power supply set to CV ( 4.4V ) to see if the cell will cut off at around 4.3V.
The three basic functions of overdischarge / short circuit / overcharge can all be tested using the above methods. I haven't seen anything like what is described in the 1st post. However, if you can perform all the basic function test and they turn out okay, I don't think you have anything to worry about.

wow, thankyou for the *VERY* fast reply!

running tests now... not sure if i will have any way to test overcharge (and don't really want to force a high state of charge if the other protection checks out fine).

so far, the flash amp test went perfect, dropped to 0V 0A within a fraction of a second. waiting on a discharge....

discharged till light went out, open circuit was 3.04V after discharge.
it's charging back up now without a problem. (cell is not hot during charge, steadily climbing voltage)
everything seems to be fine. I'm not going to worry about this.