I’m reopening this thread ( MP24833 strange behaviour ) because after more than a month and a half without a response, and on the very day I received an email saying that the company’s engineers respond to forum posts in less than 24 hours (hilarious), they posted a reply and closed the thread.
I’d like an explanation of how a high inrush current can damage the boards.
Perhaps I didn’t explain it well enough, but these boards function as a kind of “emergency lighting” in an installation with a 48-volt battery bank. The boards are not connected while the system is charging and only connect after the system enters backup mode. That’s why I mentioned that the operating voltage is slightly less than 48 volts.
To prevent over voltages or overshooting from switching and long wiring, I’ve implemented a surge protector on each branch of the lights, consisting of an SMCJ48A TVS diode and a 220uF x 200V electrolytic capacitor.
Verified with two different oscilloscopes, the startup is “smooth” and the over voltage is between 1.5% and 2% after many tries. I had already indicated that in the original post there was a transient of less than 1 volt during connection.
In all cases, I am within the chip’s parameters. Can you explain how connecting many boards, and not just one or two, would destroy it?
I really expect a response within 24 hours, as promised in the email you sent me as an advertisement.
By the way, 40 PCBs with a 4.5uF input capacitor is less than 2000uF. I tested this by placing a 3000uF (low ESR, like the capacitors on each pcb) capacitor at the end of each branch, and not only were there no inrush current, but the transient response improved, and the over voltage was reduced to less than 0.5%.