Since a long time, we have some issues with the programming of the MSP430F133.
The programming is done with the JTAG port - TCK, TMS, TDI, TDO, RST, Vcc, GND. The MCU is programmed at the beginning of the test sequence with a test software
At the end of the test sequence, the application software is loaded and the offset data are written @1080
The issue is following, randomly, we get the error message "JTAG FUSE BLOWN" -> after this message , the processor is not writeable anymore
The flasher device is a FLASH Pro 430 Elprotronic , the processor is a MSP430F133.The flasher is isolated with relays (between flasher and acces point)
The wire length total is less than 40cm. The DUT is powered from the flasher during the programming (3.3V)
The programming is done using the MSP430FPA1.dll (from Elprotronic)
The devices under test are "power supply board" with different configuration but always with the same daughter board (where the MSP430 is located)
The DUT is tested in version 75W, 150W, 300W, 600W in range 12V, 24V, 48V
The first programming (with a fresh board or a board not powered since a long time) is always pass
Then the test process switch the main power ON and OFF many times during the test.
At the end of the test, the main power is disconnected from the DUT, then the programming relays are switch ON
If the programming is launch immediatly (or with a small delay..ie 5s) then the programming is failed with the message "FUSE BLOWN"
So the question is :
What could generate the fuse blown state during the programming
We monitor the 3V3 during the programming (with oscilloscope) and we don't see any transient phenomenon (ok we don't have a high performance scope), it stays stable at 3V3
For sure we do not see voltage over the 3.6V, we check the grounding also
If we wait more than 30..50s, then the programming is OK and we can restart the test process.
What should we check and monitor to have an idea about the destruction of the JTAG during the programming
The only way to avoid the fuse blown, is to wait a long time (at least more than 30s..60s), behaviour looks like a cap discharge but we also did a discharge of the primary capacitor (charge at 400 V)
and the secondary capacitor