[Elecraft] Gel Cells in the K-2 (and other battery powered rigs)

[James R. Duffey]

Jeff - Your battery failure may be related to undercharging. To float charg=
e
a Gel-Cell, the battery should have 13.6 V to 13.8 V applied, limited to a
current of 0.1 C or less, where C is is the amp hour (AH) rating of the
battery. Many battery manufacturers print the exact float voltage on the
side of the battery. It usually varies from 13.6 to 13.8 Volts. You should
measure the voltage at the battery terminals (not the input to the K-2) and
adjust it to the correct value. Too little voltage is as bad as too much.
Don't take the stock answer of 14.0 V at the input of the K-2 as gospel,
measure it at the battery. The ideal voltage for trickle charging a battery
is that voltage where the current to a fully charged battery is zero. If th=
e
current is positive, the battery is being overcharged, and if it is less
than zero, the battery is being undercharged, that is, the battery is
supplying curent to the charger

Overcharging a battery is as bad as undercharging it. Ideally, one would
limit your charging of the Gel-Cell to  14 Hours at at a time at the 0.1 C
level. If you can't rig up a timer, you might want to plug in the charger
when you operate and unplug it when you stop, or a day later. This will
increase your battery life over leaving it connected all the time.

Fully charging a Gel-Cell is difficult. Triple charging is recommended; the
battery is charged at constant current until an end point voltage of 14.4 V
to 14.8 V is achieved. This can be done safely at 0.1 C, many batteries can
be charged at 0.3 C, some batteries tell what this maximum level is on the
side. After this voltage is reached, one switches to a constant voltage at
14.4 to 14.8 V, depending on the battery (check the side again), and it
continues to charge until the current drops. After the current has dropped,
the conventional trickle charge at 13.6 to 13.8 V kicks in. The
Unitrode=A03906 IC charges batteries this way. A circuit using this triple
charge method was in the handbooks for years. Again, voltages should be
measured at the battery terminals, not at the input to the K-2.

A compromise between the two techniques achieves better charging than the
trickle charge, but is not as complicated as the triple charge. A
conventional voltage regualtor (such as the LM317) can be configured as a
constant current charger by connecting a resistor between the output and th=
e
reference leg. The battery is charged with this constant current source. Th=
e
voltage of the battery, which is roughly the same as the output voltage of
the charger is measured, and when the desired voltage is reached (14.4 to
14.8) an SCR is triggered. The SCR is in series with a potentiometer in the
reference leg to ground. The potentiometer is set so that the voltage from
the charger is the trickle voltage, 13.6 V to 13.8 V.

Either of these two schemes should maintain the battery so that it is usefu=
l
for hundreds of charge/discharge cycles. That is several years of typical
Ham use. If you get less, either you or the manufacturer  are doing
something wrong.

The capacity of Gel-Cells depends on the discharge rate. Usually the rated
capacity is based on a 20 hour diecharge. So the 2.7 AH gel cell in the K-2
will only have the full 2.7 AH of capacity when discharged at an average
current of 135 mA. Depending on the battery and the discharge rate, the
capacity drops to 1/2 to 2/3 of the rated capacity at a 10 Hr discharge
rate, and less at higher discharge rates. So if you draw 270 mA out of the
battery (I seem to recall that this is the average current draw for a K-2
running at 5 W with a 50% duty cycle on CW), you can expect far less than
the 10 hours that the 2.7 AH rating would indicate, perhaps as few as 5
hours or so.  At 15 W, I would expect a very short life indeed for the
internal K-2 gel-cell.

I have gel-cells that are over 5 years old with hundreds of cycles on them
that still take and keep a full charge. You should be able to do as well.

I hope that this helps somebody. Sorry for the rambling. - Dr. Megacycle
KK6MC/5
--=20
James R. Duffey KK6MC/5
Cedar Crest, NM   DM65