Well, before that, I've been trying to revive my second puma (that one was *really* messed up with bridges). I fixed the bridges with the new iron, and loaded the SM. It didn't work, though, and took me a while to figure it out that the crystal had one of its 4 pads without touching the board. Again, too much solder paste in the other crystal's pads made it go up.
/* how to check the crystal */
The PLL VCO voltage should be somewhere close to 1.15v. If you have 2.5 (full VCCPLL) then you have a crystal issue.
After fixing that, I reloaded the SM, and apparently its working (I can debug it and see the instructions that are being executed). However, it is locked in two instructions, not sure what they do. Its like a while (something);
(@FC96):
BRSET 0xCC,#32, *+6 ;abs=0xFC9C
(@FC9A):
BRA *-4 ;abs=0xFC96
I believe it compares the byte at address 32 with 0xCC. Now I have the assembly manual, but still have to lookup what is in the 32th byte,
So, I still don't know why it hangs like that. I wouldn't be surprised if the MCU is dead, at least in part. A pin (PA6) in the working boards shows a square signal, here shows a loud and clear 0v. I tried resetting it (to let power stabilize), no differences.
Anyway, I got tired of that so I put some time to test the oven properly. I hope to assemble something tomorrow (its 4:40am now)
This is how it looks. The rise time is slow, but reaches 230°C in less than 5'. At 4:30' (more or less) the solder starts to reflow, so its acceptable to me. I still have to try one more trick to check if the risetime gets better.
With more °C/min I can try to follow a proper reflow pattern. No point in trying it with this sluggishness.
The weird values appear when I move the probe while opening the oven door to let it cool.
I can't work at beach hours! The workplace is between 34°C and 39°C in the day, I've a reflow oven inside this human oven, its a super efficient setup :pFred wrote:+1 too mucho tiempo en la playa = no bueno ;-)