Interesting. The photo was of the old buck converter. This is a whole new board with LDO. I soldered the first one in my reflow oven, but the replacement one I did by hand. I'll take a look at the ground and report back. Thanks.

Don

 

I don't know...the ground pin looks pretty solid.

 

I added some solder to the ground pin to make sure it's a good connection but I still have the problem. It stays at 5.02V until i get up to about 7.1V input, then it starts climbing. I made it up to about 5.8V output before I chickened out this time.

The ground is connected and is true earth ground. I checked all grounding for continuity. I'm running out of ideas.

Don

 

Where are your meter connection points??Have you measured the regulator output voltage AT THE PINS OF THE REGULATOR?? The symptom that I see is a voltage drop in the common side some lace between the regulator common terminal and the negative meter connection point. That almost drove one tech "rather close to the edge" when a voltage regulator was not holding te specification as the current increased. Presently the symptom points in that direction. UNLESS the regulator is oscillating.

 

The soldering on all three terminals on that chip looks suspect. Put some flux on them and remelt the solder until it flows. It should never look jagged as it does in your picture.

 

OK, looks like it's solved. I took the regulator off and measured the output pad on the PCB with input voltage applied and got a voltage (!). Hmm. I had an extra, unpopulated PCB so I started checking continuities. I stumbled upon a connection between +15 and the wrong side of a power transistor (gate leg). There was a bad (extra) trace in my design. The +15 was leaking back through a gate resistor (82 ohms) and then on to the MCU output pin (hard telling yet what I did to it).

Anyway, I cut that trace, soldered the regulator back on, and now it works. Thanks for all the tips and help!

Don

 

It HAD TO BE a connection issue! Not what I had imagined, though. GOOD JOB catching it.