Other Parts Discussed in Thread: C2000WARE
C2000 Team,
I am attempting to run the ADC on F28377D silicon. If I use rev.A or rev.B silicon, everything works. If I use rev.C silicon, it does not. I have tried several different programs, including one from C2000Ware. Details for this particular test setup are as follows:
Board is TI FAE board with socket, F28377DPTP silicon. I have rev.A and rev.C silicon for this board. Program is C:\TI\c2000\C2000Ware_1_00_00_00\device_support\f2837xd\examples\cpu1\adc_soc_software\cpu01. This program converts 2 SOCS on each of ADCA and ADCB, giving results AdcaResult0, AdcaResult1, AdcbResult0, and AdcbResult1.
I jumpered 2.5V to each of the ADC input pins (the board has provisions for making this connection). When I run this with rev.C silicon, I get 0 for all results. When I use rev.A silicon in the socket, I get correct results. Same program. I don't even quit CCS. I just turn off the board, swap the silicon, re-power, and then reload the program into RAM. For rev.C silicon, the program runs. I am just getting zero for the ADC results. Note that the board uses 3.3V references for the ADCs.
I've also tried my own code, which is ADC interrupt driven. Here, I get the value 0x0F37 for all 2.5V conversions on rev.C silicon. This seems like a saturated value, as there is NEVER any variation in the value due to noise. As before, if I use rev.A silicon, I get correct results.
I've tried multiple rev.C chips, including a brand new device just removed from a vacuum seal bag. The problem is not damaged silicon.
Any idea what is going on here?
Regards,
David
