Using the class D amp driver chip IRS20214. When operating at a duty cycle of 3% and above everything is as clean as a whistle. Same at 1% unless I drive it with an LTC6992 but that's another story. Anyhow, at around 2% is seems to generate a secondary spike at its input when the bootstrap node flies. I can muffle it with a cap across the pin but since it apparently "strikes" inside this does not prevent double tripping. I am operating it around 150V supply, well within its means.
Tried shielding the IC itself and so on, nothing helped.
Has anyone used this and ran into similar stuff? Any idea why this doesn't happen at other duty cycles?
I am also wondering how anyone could use IRS as the suffix for an IC :-)