Chasing a board ptoblem, Prop resetting constantly

T Chap

I know this is hard to figure out remotely but I need some insight into a problem.

I built a board with a Prop and other devices, with an on board 1.2 amp 3v3 regular powered by another supply that is a switching 5volt.

If the board is doing any amount of work, the Prop resets, even turning on 3 or 4 leds will trip it. When programming the Prop from the PC, the rail goes to 3.25. But after programmed, it drops to 3.0 volts. Now, the VS1033 and Vinculum can be still working in the background playing an mp3 file while I hit program, it just keep running as it doesn't know to stop unless there is a command. Originally I thought the VS1033 mp3 decoder was dropping the rail somehow, causing the reset. However, after testing that the rail jumps to 3.25 when programming, and while the VS1033/Vinculum are still playing, it seems that the Prop is possibly the culprit behind the voltage drop.

I don't have any pins programmed on the Prop except an LCD and Tx Rx to the Vinculum, so I think that rules out some type of output on the Prop getting shorted to GND, but this is exactly what appears to be happening. Seems like a lot of rail to drop after programming is done, considering it is sitting idle after.

Not much to go on here, just looking for any ideas of how to get it solved.

Phil Pilgrim (PhiPi)

Program the Prop to do nothing, i.e. all pins are inputs. Does it still pull Vdd down? If not, test each output pin individually, both high and low, and measure the voltage with a voltmeter. Which pin is the culprit? Is it shorted or driving a very low-impedance load?

If the all-input mode still pulls Vdd low, try programming the Prop to run at a lower frequency. Does Vdd come up a little? If so, you may have a defect in your power distribution traces. Are they fat enough (0.03"/0.75mm or more)? Does your switching regulator have an adequate filter cap, according to its datasheet? Try loading the regulator with a 33-ohm, 1/2W resistor to ground (100mA). Does that pull Vdd down even more? (It shouldn't, if the regulator is rated at 1.2A.)

How about posting a schematic and board layout for us to look at?

-Phil

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Still some PropSTICK Kit bare PCBs left!

T Chap

Well Phil, I wrote a very detailed response noting all the tests you mentioned. Then, as an extra measure, looked back at the data sheet for the 3v3. I knew the part number seemed unfamiliar to me, and how this new part entered Digi order sheet the picture is a mystery, must have been a substitution part due to back order I suppose. But, unfortunately the sub part has reversed pin outs for in/out. I do suspect this discovery will help the situation immensely!

T Chap

Great, that solves that problem. What made this so confusing was that even in backwards, it output 3v3 with no load. Thanks again Phil, that check list you made put m on the path to the power supply, even after I had already tested the supply once.