Ok, I just succeed to run the code with no errors for more than 24hours now. I disabled all the interrupts, and handled the interrupt routines in the main loop.
Since this device is much more faster than the old one (16MIPS vs 70MIPS), it almost isn't necessarily that the interrupt is used anymore for my application, but unfortunately I still need interrupts since I`m missing sometimes some serial characters due buffer overrun of the UART.
However, this leads me to the problem that has to be caused by using interrupts and the mainloop code. I`m kind of worried about this. All interrupts have the auto_psv enabled and don't share in an unsafe way variables with the mainloop. I always disable all interrupts if I need to read data that comes from an interrupt.
Also, I used the same code on the PIC24FJ without any problems, it has run for years with over 2000 quantities.
I am afraid it is some errata in the chip that I`m unaware of. I`m waiting for my RealIce so I can enable jump tracing. Maybe this gives me more information how and where the problem is.
Since this device is much more faster than the old one (16MIPS vs 70MIPS), it almost isn't necessarily that the interrupt is used anymore for my application, but unfortunately I still need interrupts since I`m missing sometimes some serial characters due buffer overrun of the UART.
However, this leads me to the problem that has to be caused by using interrupts and the mainloop code. I`m kind of worried about this. All interrupts have the auto_psv enabled and don't share in an unsafe way variables with the mainloop. I always disable all interrupts if I need to read data that comes from an interrupt.
Also, I used the same code on the PIC24FJ without any problems, it has run for years with over 2000 quantities.
I am afraid it is some errata in the chip that I`m unaware of. I`m waiting for my RealIce so I can enable jump tracing. Maybe this gives me more information how and where the problem is.