High Voltage Ripple across ESP32

[naitik]

Hello, We are designing the product based on esp32. The prototype PCB We have converts 230v > 12v > 5v > 3v3 starting from 12 Volts stage we having ripples of 14 Volts which propagated forward on Each stage of Regulator. eventually at 3v3 Stage we are getting ripple of 4.6 Volts at every 40 uS.
My question is that, are these short pulses of 4.6 v is tolerable ? And can anyone specify what are ways this ripples can be reduced?
thank you.

[tommeyers]

Please post your circuit diagram with part#s

[ESP_Sprite]

And to officially answer your question: 3.3V + 4.6V exceed the specs as indicated in the datasheet. It's impossible to say how the ESP32 reacts to that, but it's not officially supported and may lead to undefined behaviour.

On a sidenote: Are you sure these ripples aren't because EMI is getting into your measuring probe via e.g. the ground lead loop?

[becorey]

And can anyone specify what are ways this ripples can be reduced?

You could use an LDO or TVS diode to reduce the spikes. But better to look first at your 3.3v regulator, why is it passing such spikes through. Is the feedback loop node receiving noise? Is the output capacitance too low? Need to see schematic and pcb.

[naitik]

Here's Schematic :

There is 12v power supply which takes input from UK Plug 230v. the 12v power rails goes to 5v regulator then 5v goes to 3.3v then 3.3v regulator goes to 2v regulator. At each stage of conversion the Ripple voltages are propagated forward as well.

[naitik]

Also we are sure that Ripples are not recorded due to bad probing. because when we power the board using PC we do not get any significant amount of ripples in the oscilloscope.

[ESP_Sprite]

That doesn't tell you anything. If the board is powered by the PC, presumably the DC-DC converters are off, there's none of their EMC in the air and bad measuring won't pick it up.

[naitik]

yea it makes sense now. i totally missed the EMI noise due to converters probably the Oscilloscope picking up some of that by providing power from pc tells basically nothing