Hi, we have a similar design:
Vth = 10.5V
Vfloat = 13.8V
Vboost = 14.7V
Vin = 16.9V
Imaxchg = 250mA
and we obtained the following results in ohm (all resistors are 1%, with except for Risns that is 5%):
Ra = 214.6k (choosed: 216k = 200k + 16k)
Rb = 17.24k (choosed: 17k = 16k + 1k)
Rc = 46.4k (choosed: 47k)
Rd = 586.8k (choosed: 582k = 5602k + 22k)
Rp = 1143 (choosed: 1100)
Risns = 1
Rt = 406 (choosed: 500)
As transistor we choosed a BCP53.
We choosed to use the configuration with a diode (we choosed an 1N4007) and with Rc tied to /PGOOD.
We cannot achieve the 13.8V float voltage, but we only have 13.2V on battery terminals.
Another question is about Vrch. If we follow the formula for the example given on datasheet, the voltage threshold for the example's battery to be recharged should be
Vrchg = 0.9 * Vref * (Ra + Rb) / Rb = 0.9 * 2.3V * (75k + 16.9k)/16.9k = 11.26V, that is quite over the full battery voltage.
Is this formula correct?
Thanks in advance
Stefano Bettega
I split this thread as the other thread had gotten quite lengthy.
The recharge voltage is 90% of the float voltage. The battery voltage this corresponds to is set by Ra and Rb. This is the voltage at which the charger initiates a new charge cycle.
Based on the resistors you choose, the float voltage should be 13.7V. What temperature is this test run at? Is the IC hot? The reference voltage is designed to change with temperature.
Which transistor topology are you using? Is the battery voltage sensed after the diode?
Can you post a schematic?
Try increasing your input voltage a little bit. You may be in dropout.
Hi Chris,
see attached the currently used circuit topology (please don't take into account the pin labeled 14V1).
We are using a BCP53 PNP transistor and a L4007 diode. Sensing is done on battery, after the diode.
I found I missed something in calculating the correct recharge threshold: the formula on the datasheet is given for the simple charger topology in figure 3, where you have only Ra, Rb and Rc. For a complex topology, involving Rd like the one in figure 5, you have to take care about a third resistance, so your formula may be rewritten taking into account that Ra = Ra+Rb and Rb = Rc. This should fix the calculation, giving Vrch=12.3V.
Well, we tried to discharge the battery until it reached 11.9V, but when charging only 150mA will flow from battery charger to battery.
IC is comfortably warm, only transistor will heat up, but still into a reasonable temperature. Our input voltage is now fixed to 13Vca, thus leading to 18.3Vdc at the charger input pin.
Thank you for posting the schematic. What issues are you still trying to debug?
Where did you measure the 150 mA? As you know, the charging current is set simply by R75. Measure the voltage from pin 2 to pin 3--measure it at the pins of the IC. This voltage should be at 250 mV. If it is and the current is still off, the resistance used for the measurement is wrong--either the sense resistor value is off (check the power capability of its package) or the IC is not using a Kelvin sense connection to the resistor and is seeing additional resistance in traces.
Three main concerns about this schematic are not so clear for us.
An interesting thing could be a way to sense if a battery is really connected to the charger (i.e. the protection fuse is not blown). Doing a logical AND between STAT1 and STAT2 pins, if both are OFF (logical level high) should carry an high logical level only when no battery is connected. Is this correct?
If there is no battery, the IC will regulate the output at Vfloat. This assumes that the load current is less than the maximum charge current, as set by R75.
Both STAT pins will be off if the battery is fully charged and the IC is regulating in float mode. See figure 4.
To set the fast charge current, the IC looks at the voltage between pins 4 and 1, not 2 and 3 as I originally stated. Measure the voltage between pins 1 and 4 and we can see if the IC is regulating the set charge current. If this is substantially less than 250 mV, then it could be tapering the charge current while the battery voltage is being regulated at Vboost. Measure the voltage from pin 13 to pin 6 at this point and see if it is at 2.3V. If it is, then the IC is tapering the charge current. I also recommend looking at the above pins/signals with a scope to make sure they are truly at DC, which they should be.
Another to try might be blowing some air across the IC or using freeze spray to see if the boost voltage changes. Neglecting any component variation/tolerance, to get a 13.2V boost voltage instead of 13.7V (what your resistors set it to) only requires about 24 C of temperature rise. If it is a small board, then this is a very possible amount of rise.
I recommend testing this charger with a lab power supply and resistors instead of a battery. This is a more predictable test setup that can be varied to all the different charge states without waiting for a battery pack to charge/discharge.