Posted on February 19, 2012 at 11:56You do not mention what you are doing with the BOOT0 and PB2/BOOT1 pins. If you hope to get repeatable results then these pins should be pulled to the appropriate levels (likely to be booting from System Memory if you intend to use the ST bootloader, or User Flash if you are using e.g. JTAG to load your program; but then for this you have to pull other pins as well).
You say that the reset pin functions as you hope for a few minutes. How far are you able to get in that time? Can you establish contact between your debugging environment and the STM32 in that time? Incidentally, what debugging software are you using, and what programming hardware (e.g JTAG, SWJ).
The STM32 microcontrollers can drive their reset pin low for a pulse of longer than 20 microseconds under certain conditions (for example watchdog reset). Might the 0.6 V you observe actually be the average voltage on the pin, which is actually doing something more complicated? An oscilloscope would be very helpful here.
As to your comment about power-supply-decoupling, in my opinion, the loop from each Vdd pin through the capacitor to the nearest Vss pin that must be kept short, and so must the path from each Vss pin to the ground plane.
If your Vss pins have a very short path to the ground plane (ideally less than 5 mm) then the fact that the decoupling capacitors go to the ground plane on a similarly short path rather than to the nearest Vss is not likely to be a problem.
I am aware how frustrating it can be trying to get a chip up and running. Do you have a commercial demo board to hand so you can confirm that all your other tools are working as you expect?
Hope this helps,
Danish
