Hi,
My issues is when the control (IN) is changed, the output (COM) sometimes goes into oscillation. In some cases, the oscillation is so fast that the part will no longer function.
IN is driven with a TTL or CMOS driver to both states (~3V to ~0V).
IN is not being toggled quickly (like seconds between toggles).
My application now experienced device stop working many times which may be due to unwanted oscillation. When the device stopped working, COM is connected to a load with greater than 40 kOhm input resistance. Device is no longer functional after this occurrence.
When I have no load on COM I observe oscillation in my setup just by setting IN high and then touching my scope probe to IN.
The datasheet states bypass capacitors are recommended though not required to reduce noise; when adding the capacitor between pins 3 and 4 the oscillation stops. It is only recommended to improve noise margin and reduce the amount of switching noise put back onto V+. With IN driven to both states (~3V to ~0V), it seems there should be plenty of noise margin.
Why is this capacitor helping to stop the oscillation as opposed to just reducing switching noise?
Are there use cases which necessitate a decoupling capacitor?
Are there any other insights to share regarding the above scenario?
Thank you,
HSG