I am buildinga circuit for charging of my 12 V sealed lead acid battery
using UC3906. The circuit behaves differently than it is described in the
datasheet. It is said in slua115.pdf (U-104 application note) page 3-81
that "In the over-charge state, the charger will continue to supply the
maximum current." The same behavior is shown in Figure 3 of the datasheet
(slus186C-September 1996-Revised May 2005). When the charger switches to
the "over-charge" mode (point C, charge voltage = V12), the charging
current equals Imax. When the charge voltage reaches Voc (point D), the
current limiting circuit starts to reduce the charging current.
This is theory. In my circuit, when I connect partially discharged battery
to the charger, it switches to the "over-charge" mode (Pins 9,10 are low).
The battery terminal voltage grows slowly, but AT THE SAME TIME the
charging current drops gradually. This is not the behavior which is described
in the datasheet.
Here I provide details of my circuit.
Schematics: Fig. 13, U-104 Application Note (slua115.pdf)
This schematics does not have resistor names, so I use Fig.8 of BQ24450
datasheet for reference. The only difference of my schematics from
Fig.8 of BQ24450 datasheet is that in Fig.8 the external transistor is
connected using quasi-darlington scheme, I use common-emitter pnp scheme.
Later, I wired LED + 4.7k resistor between Vin and Pin 9 to detect over-
charge state visually without measuring voltages, but this does not change
the behavior of the charger.
SLA battery: Tenergy 12V, 7.2A (TB1272)
Vboost = 14.4 - 15.0 V
Vfloat = 13.5 - 13.8 V
Ich max <= 2.16 A
External power (bench voltage source):
Vin = 18.0 V
Resistors:
Ra = 210.2 k
Rb = 18.47 k
Rc = 45.5 k
Rd = 519 k
Rs = 0.16
Rp = 1 k
Resistors were measured "in site" with the charger disconnected both from
battery and input voltage, and uc3906 is removed from the socket. I used
combination of "fixed" 1/4W resistors (2W for Rs) and potentiometers.
Compensation capacitor (from Pin 14 to GND):
0.1 uF, ceramic
Threshold voltages (assuming Vref=2.264V):
Vboost = 14.64 V
Vfloat = 13.64 V
Vth = 10.15 V
I measured Vref from Vfloat, when the charger switched to "float charge"
mode (see description below). It was difficult to measure Vref voltage
directly from Pin 13, because the attempt to measure voltage at Pin 13
in "over-charge" mode often caused switching the charger to "fload charge"
mode.
External transistor KSB1151YS, PNP, hFE=100-200, Ic=5A (with heatsink).
External diode 1N5406, Imax=3A, Vdrop=0.744V @ 1.5A (measured).
The charger was connected to the partially discharged battery,
then the external voltage Vin was applied. Charger switched
to the "over-charge" mode (Pins 9,10 low, LED ON). In the beginning, the
charging current was about Ich = 1.6A, Vbat = 14.21V. Ich was gradually
decreasing, Vbat was gradually increasing. This behavior is different from
that described in the datasheet.
When the voltage Vbat = 14.58V and Ic was about 160 mA (was difficult to
catch the exact value), the charger switched to the "float charge" mode
(Pins 9,10 high). Vbat = 13.64 V, Ich = 110 mA deacreasing to 80 mA
after 3-5 min (and very slowly decreasing).
From Vloat = 13.64 V and measured resistor values, I got Vref = 2.264 V.
Using this value, Vboost = 14.64 V. This means, that the current limiting
scheme started to reduce current (from 1.6A to about 160 mA) BEFORE Vbat
reaches Vboost, and actually, the transition from "over-charge" to
"float charge" mode happened even before Vbat reached Vboost.
I have 2 uc3906. Both show the same behavior.
Is something wrong here or this is normal behavior?
Thank you.
Konstantin
