not an MPS engineer… but I have 30+ years experience in electronics engineering. I have designed probably a thousand boards in that time, including designing the timing cards for the Japanese cellular network, and I built the first router that delivered the internet to NZ (and the modems too)… back when there were 20 odd people on the internet here
Yes I tried cooling, but it is same.
Do they all do it? How many have you built how many have this problem? Is there maybe a mechanical sensitivity? If you flex the board does it happen?
Hi all,
Not sure if this is still a issue for the OP, but we experienced a similar issue on one of our products.
We are using the MPM3650 to generate +5.0 Vdc output from +15.0 Vdc input and noticed that the module would shutdown for about 100ms every now and then.
Both the input and output rails were very stable, no significant transient load currents observed, case temperature stable at +63 °C with a 3.2 Adc output, yet every now and then the PG signal drops to 0V and the output decays. 100ms later, it comes back to life and continues on for another 15mins or so.
What I noticed is a discrepancy in the EVK datasheet regarding the SW pins.
On page 3, the schematic shows all SW pins connected together, yet on page 9, the layout of the top layer shows that the SW pin 23 is not connected to the other SW pins:
After removing the IC, cutting our connecting copper from pin 23 to pins 7, 8, 13 & 14, then replacing the IC we did not see this issue occurring.
So @MPS, please issue a clear update to the documentation stating that this SW pin should NOT be connected to the others, if this is indeed the cause of these random shutdowns.
Thanks.
That is frigging embarrassing. What a good catch. Eval boards very valuable tools.
Hello,
Thank you for bringing this to our attention. Were you seeing this problem occurring with multiple chips or multiple boards? The SW node in this case is just for use as a test point since the inductor is integrated in this module. pin 23 and the other pins fan out from the switch node internally. If you short pin 23 to the other pins externally, it creates a conductive loop that current can couple to and create issues.
In the datasheet for the part it says to float the SW pins in the pin functionality section but there is some ambiguity as the diagram shows all the SW pins as one item. The EVM datasheet, however does have a discrepancy where it shows 23 connected to the other pins externally, meanwhile the actual layout has 23 floated independently. I have already passed this along to the relevant people to get it corrected.
Best,
-Kerr
Hi all,
We had same problem like LordZetskus. We connected pin 23 and other SW node pins externally on PCB. We’ve made about 20 samples and half of them had this issue.
We are planning on PCB revisions.
Thanks for LordZetskus! I really appreciate you cleared our issue!