I have a serious problem with the BQ24401
1. low PWM frequency
From 12 identically assembled prototypes only 1 reaches the 300 kHz PWM frequency, 2 have abt 130 kHz and the rest are below 60 kHz.
The components determinating PWM frequency have their exact values on all the boards.
Replacing the BQ24401 on the 300 kHz board with a new one made no difference.
The -50 mV trip voltage at the SNS input is verified at the 300 kHz board but has about twice its value at the others.
Any idea ?
2. no dT/dt-detection
Due to power dissipation in a compact design I had to reduce the charge current from C/2 to C/4 in order to prevent early dT/dt detection.
The BQ24401 takes revenges on this with no dT/dt detection anymore.
The only stop criteria still functionable are max time and max temp.
The dT/dt voltage changes measured are still greater than the -Vcc/161 = 31 mV/min or 0,52 mV/s but the BQ24401 ignores this.
The datasheet explaines that the actual temp value will be compared with the value measured 2 samples earlier. So a -8,3 mV change in 16 s should terminate the fast charge. Right ?
What is the internal A/D resolution and are the values filtered or averaged ? I made my measurements with an external A/D and had to filter out the noise in order to get qualified data.
Akkuladung kpl 300mA - Ueberladung.pdf
In order to cancel noise I have added RC filters (100k / 100nF) to both TS and BAT inputs but with no effect.
3. input current regulation instead of output regulation
The datasheet proclaims that "The bq24401 implements a hysteretic control loop that regulates the current being delivered to the battery pack
to a user programmable value that is set by the value of the RSNS resistor."
But in the schematic the sense resistor is implied in the input loop instead in the output loop. Therefore the input current is regulated and the design is useful only with a fixed input voltage.
I understand that - to change this - I have to redesign with a sense resistor in the output loop and - in order to archieve lowest ground resistance in the battery to load connection - to insert this resistor in the highside output of the charger.
4. corresponding current regulation is sensitive to input impedance
Ground in not equal to Ground, especially if there are sense resistors.
With different power supplies as input sources I get up to 20% different input currents.
I assume that this is due to the different ground capacitance of the linear Lab power supply and the Mean Well switching adaptor.
Also high frequency ripple currents of the switch mode supply may make a difference. But additional decoupling on the chargers primary and across the sense resistors also was not worth mentioning with respect to input current regulation.
5. max temperature sensing works great
The charging is therefore only terminated by reaching the max temperature and of course max time - poor batteries ;-(
Any suggestions ?
Thank you
