L9823 - Output not working when connected to Fan ( Random behavior)

[Harirajan]

Hello team,

I'm using this L9823 for driving a 24v, 0.24A DC motor, but sometimes when i try to turn on the Fan it is not working and after some 3rd attempt it is turning on. The output is capable of driving upto 2A right. I monitored the current and during the turn on time i get a current spike of 1.3A for 80us. Does the cureent spike making the output to turn off? I also observe the same spike during the fan working condition also , but the duration of  spike is less. Can anyone give a solution.

[Harirajan]

Can anyone support?

[Harirajan]

@Peter BENSCH Can you support on this topic.

[Peter BENSCH]

You have not mentioned whether your motor is connected to a single channel of the L9823 or to several channels in parallel. In any case, the data sheet states

• table 3, absolute maximum ratings: output current up to 2A
• table 5, overcurrent shutoff threshold: min 0.5A, typ 1.6A, max 2.5A

Since you have also not mentioned what your schematics look like, I assume that pin SFPD is not connected in your case. In this case, the internal pull-down is active there, which is intended for loads without inrush current (data sheet, section 2.4). So if you now have a currenst spike of 1.3A when switching on and the overcurrent threshold for this example is below this value, the output is logically switched off.

You could connect SFPD directly or with a pull-up of max. 18kohms to VDD so that the output does not switch off immediately but is limiting. This can also be found in section 2.4:

Output short circuit protection SFPD = High (dedicated for loads with inrush current, as lamps): when the load current would exceed the short circuit limit value, the corresponding output goes in a current regulation mode. The output current is determined by the output characteristics and the output voltage depends on the load resistance. In this mode high power is dissipated in the output transistor and its temperature increases rapidly. When the power transistor temperature exceeds the thermal shutdown threshold, the overload latch is set and the corresponding output switched off.

As long as the overcurrent pulses remain short enough and do not cause the temperature of the chip to rise too much, your motor should be able to run continuously.

Hope that helps?

Regards
/Peter

[Harirajan]

Hello @Peter BENSCH,

Thanks for your clear clarification, I would like to mention some few points for better understanding.

1.The motor is connected to single channel, and in other channels are also occupied. 

2.The SFPD pin is connected to GND in my case because of safety reasons, if the case of short circuit the channel will be fully turned off.

3. Is the output current limit can be programmed and can be checked via SPI?

4. What happens if i connect a load of 0.65A? and SFPD is connected to GND. 

5.  I have attached waveform images for your reference

     Image1:  Motor not turning on, current spike observed for 80us,1.3A max

     Image2: Motor turning on after some attempts, but current spile also observed, but duration of peak current is less. This confuses me.

Now i need to change the fan from 0.24A to lower ratings or can i configure in the program itself?

Worst case: Can i add some external circuit with mosfet with slow turn on circuit? will that work? like a load switch or soft start , will that work?

Thank you

[Peter BENSCH]

Firstly, I am still confused by the current spike because the motor should be working inductively and its current should increase with the curve, as can be seen on the right side in image 2. Is it possible that the motor contains electronics that cause this spike?

Before you change the motor or consider external circuits: you should also be able to switch the SFPD pin with a GPIO from the controlling MCU:

• with SFPD = H to start the motor
• with SFPD = to L for full short-circuit protection after a few milliseconds

Regards
/Peter

[Harirajan]

Yes, even I'm also confused with the first waveform, but i suspect it due to the "tdly SCB" , because that 80us in the image 1 is constant in all the observation.  It might be because of the program.

and due to old design consideration, the pin 21 (NC) and was connected to 24v, will that cause any issues? in datasheet nothing mentioned. 

And you gave me a very good answer, by controlling the SFPD pin with a GPIO or MCU, thanks

[Peter BENSCH]

I have learnt internally (internal case # 00223168) that pin 21 is not connected to the die, so it doesn't matter if an (almost) arbitrary voltage will be applied there.

Regards
/Peter