LM3478: High ripple at no load

Hello Danilo, 

Thanks for reaching out. What are the Vin, Vout, Iout and switching frequency specs for this design? Does this behavior disappear if you increase the load? How does the scope plot look like if you zoom in one event? I am asking because it might be that you are hitting OVP in no load. I'm looking forward to hearing from you. 

Kind regards

Hi Elisabetta,

This behavior disappear after 70mA output load.

As I already mentioned Vin 10-15V, Vout 22V, Iout max 3A,   Rfa = 26.7K,

Thank you,

Alex.

The frequency is 550KHz, the ripple at 3A is 120mV

The frequency is 550KHz, the ripple at 3A is 120mV

Hello Danilo, Alexandru, 

Thanks for sharing the specs. Since this behavior is disappearing increasing the load, you are hitting OVP. The LM3478 has a minimum on time of 325ns, so at no load this means the the device is still dumping a small current onto the output cap, which will cause the voltage to increase. It will continue to increase until it hits OVP at which point the part will stop switching until the output voltage drops a little below regulation. OVP is reached when the FB pin is 50mV over regulation, for further information you can check page 11 of the datasheet https://www.ti.com/lit/ds/symlink/lm3478.pdf?ts=1620289716536&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FLM3478 

Kind regards, 

Elisabetta 

Hello Elisabetta,

Thanks for the detailed explanations.

Can you suggest a solution considering my card is a battery backup device and the lm3478 has no load most of the time (the main power is on).

Also need to mentioned that the "on battery" minimum output current is 20mA witch cause 40Hz ripple at 1.2V.

Except for a bleeding resistor can you see a solution to solve this problem?

Thank you in advance.

Regards,

Alex.

Hello Alexandru, 

In this case you might want to  try to increase the Cout or add a dummy load. 

Kind regards. 

Hello Elisabetta,

Thank you for the reply.

Dummy load is not an option because is running on battery. Can you be more precise with "increase Cout". 

Kind regards.

Hello Alex, 

Thanks for your reply. Since the dummy load is not an option, you can try for example to have one cap like C59 replaced with 100uF and see if it gets better. Can you also check the Vout in DC coupling please? 

Kind regards

Hello Elisabetta,

I have a 220uF/50V capacitor already (I forget about them when I tell you the Cout).

How about replacing the chip with LM5155? It will solve my problem?

Best regards,

Alex.

Hello EM,

Thank you for your response, we will redesign the board with LM5155.

Regards.

Hello Alex, 

Thanks for you feedback. I will now close this thread. If you have further questions, please feel free to open a new thread. 

Kind regards,

EM