I have a problem I need a CPF resident charger expert to explain.
MagChargers have a diode on the positve charging ring to prevent shorts of the light housing when in service.
Yup.
I have sucessfully hooked NiMh smart chargers to MagCharger cradle pins and charged MagChargers with appropriate batteries with no problems.
I betcha you just didn't notice the problems, or the "smart" was really pretty dumb. See below.
I recently build one using D lithiums on BatterySpace pcbs. The 7.4 V pack and PCB work as prescribed and it charges on a 7.4V Liion smart Charger in balance like it should with this pack pcb.
The charging cradle pins are hooked directly to the aftermarket chargers. With the BC-6 charger I can charge the battery in this light in the cradle through the + ring which has the diode.
The new Liion smart charger when connected this way shows the appropriate voltage potential at the pins but it will not charge the light in the cradle. The status LED does not turn red and I read no current with the DMM.
WHAT IS HAPPENING WHAT IS THE FIX?
If I remove the diode it will work, but that is short protection. The PCB does protect agains shorts in the light and as always I build fuses into my pack.
Testing a short circuit I found that my fuses blow, I have not determined if the PCB kicks out. I could run it that way. I prefer to keep the diode in the light body.
Youse is outta luck (WRT Lithium chemistry).
Dumb chargers (aka "trickle" chargers) such as the standard MagCharger simply put out 200mA (in the case
of the MagCharger, actually I was seeing more like 220mA), and rely on a combination of "smart" user to pull the
battery off the charger/cradle before it boils away, and/or on being a sufficiently "trickle" ("low") current to simply
leave it indefinitely and rely on the battery to just burn off ("waste") the overcharge (any current after the battery has
had its fill) current. NiCd batteries are good at this (burning off excess trickle charge), NiMH are not, and LiIion
cells/batteries ABSOLUTELY ARE NOT.
Now as to the smart chargers...Heh! For LiIon, most (*ALL* I should hope!) must first sample (measure) the battery
voltage to make sure it is in range (not shorted, not open, blah blah blah). With the blocking diode in place, no voltage,
the battery appears "open" to the charger; No Joy! The (*ALL* I should hope) charger doesn't think it has a battery,
and so won't engage the charge cycle. This is especially true in the case of the "smart" chargers that auto-sense
the number of cells in the battery (e.g., 3.0-4.2V == "one cell battery"; 6.0-8.4V == "two cell battery"; and so on) and
select an appropriate charging regimen.
As to the NiMH/NiCd smart chargers, there are still problems (your experience notwithstanding). A first-order approx-
imation to "smart" is simply to measure the voltage applied by the charger and declare victory (battery is charged) at
an appropriate point (e.g., say 1.5V per cell, or 7.5V for a 5-cell MagCharger NiCd/NiMH battery). This would rely on
the "smart" charger outputting a fairly small charging current (I would *guess* 0.2C or so as the max), so no "1 hour"
(let alone "15 minute") fast charging here! And as to whether or not the "smart" charger is smart enough to STOP charging
at that point, or simply light the "done" LED and rely on the smart wetware element (that's "you") to pull the battery is
anyone's guess. Hint. Hint.
My guess is that this pseudo-smart charger can probably reliably achieve 95% or better "full charge". I would further
guess the 100% "full charge" is achieved by lighting the "done" LED (green seems to be the usual choice) at, oh, say,
90% charge, and assuming the wetware (again, that's "you") won't notice for awhile, long enough for much of that last
10% to more-or-less trickle-charge to full charge, yet still soon enough to pull the battery before it completely roasts itself.
Voila, "smart" interactive charging (by incorporating you into the charging algorithm, pretty "smart", huh?) Now, with the
blocking diode in place, this "algorithm" is offset by .5-.75V (voltage drop across the diode), but on a 5-cell battery that's
only 150mV (per cell), which is probably bordering on the negligible (would you really reliably notice you only got 87%
instead of 92% "full" charge??), and so would probably work "pretty well', all things considered.
On the other hand,"smart" (really, smart!) NiMH/NiCd chargers rely on actively *AND ACCURATELY* measuring the
voltage of the battery being charged, looking for that famous 3-5mV Negative Delta Vee I'm Now Totally Charged indi-
cator. No Joy! with a blocking diode, fer sur!
So, whether for NiCd/NiMH or Lithium-chemistry *SMART* charging, the blocking diode is a deal-breaker. The charger
simply *MUST* have unfettered access to the battery in order to properly gauge the battery's condition and not over-
charge (overcook) the battery.
Now . . . one possibility (AND I STRESS POSSIBILITY -- this is just a thought experiment) that just might work well
enough (reasonable approximation of "full charge" AND SAFELY avoid overcharge conditions) would be to bypass the
blocking diode with a resistor (wire the resistor in parallel with the diode). The resistor should be sufficiently large (high
resistance) so that shorting the charging rings presents no "danger", yet still small enough (with respect to the charger's
voltage-measuring input impedance) to present a negligible to the charger's voltage-measuring circuitry (i.e., to not fool
the charger into dangerously under-reading the battery's voltage and proceding to dangerously overcharge the battery).
A 1K resistor would limit the short-circuit current to the 10mA neighborhood (dissipating 100mWt across the resistor)
which is Daily Life Safe territory (assuming you could stuff a MagCharger into your pants pocket, where it shorted the
charge rings against keys/change/whatall, your family jewels would remain safe). Now, assuming the charger's input
impedance is 100K (I would guess it's MUCH higher than that, but that's only a guess), a 1K series-resistor voltage
drop is again negligible (1K/100K is 1% tolerance, which is almost certainly much tighter than the tolerance of the
charger's internal circuitry).
Do let us know if you survive the experiment! (Personally, I'd start with a larger resistor, say 47K - 100K value, and
monitor the charger's applied voltage very carefully to see if it works.)
