I have an existing design with an MPM3506A on it, but have found I may need to briefly exceed its constant-current capabilities of 600mA.
Looking through the datasheet I’m uncertain if brief over-current situations are possible, without immediately triggering the shutdown or hiccup modes.
Can anyone advise on what parameter to look at in the datasheet, or offer advice/experience on whether I could draw (say) 750mA for 200ms, before returning to a baseline of 500mA?
Unfortunately this is on an existing product, so am unable to change to the more capable MPM3510A.
Well you could try it. Looking at the datasheet it looks like you are assured that the peak inductor current will be at least 2A ( it is dependent on Vin/Vout / duty cycle) I suspect the 600mA steady state is more set by thermal i.e slow considerations. I suspect you can get 1.5A for seconds with no issue.
I’m indeed going to try it, but am wary of ending up with a solution that looks like it works (but may be unreliable), vs. really works.
I did see the values in the datasheet about peak “high side” currents, with a 20% duty cycle, but elsewhere it talks about “cycle-by-cycle peak current limit protection” (the 1.15MHz switching oscillator?), that implies I might only get a microsecond of overcurrent.
After experimentation, the MPM3506A will enter its hiccup mode (cutting output power and resetting my device) within 10-60ms of a 750mA over-current happening. This time seems to be variable, so may be due to luck of where it falls within whatever is happening internally within the part.
Recurring over-current spikes of 1 millisecond seem to be okay, but anything significantly longer isn’t possible.
Attached is a plot of our device’s current vs. time. The mspaint edits of values/scale are due to me being only able to measure on Vin side, but then scaling to what the Vout current would approximately be.
