MSP430 Uart Settings on the MSP430F5437

Jason,

how is the UCS module initialized? Which VCORE level are you using for working with the 12 - 20 MHz?

Your screenshot seems to be the register settings for 12 MHz clock input instead of 18 MHz. 

The User's Guide (SLAU208N) doesn't specify the register value for 18 MHz clock input. Have you tried to compare your settings with the one from the USCI UART Calculator?

http://processors.wiki.ti.com/index.php/USCI_UART_Baud_Rate_Gen_Mode_Selection#USCI_UART_Calculator

We are using the values from the Low Frequency Baud Generation page. I sent you the incorrect values earlier - I was running a 12 MHz version of the firmware at the time and accidently took a snapshot of the register values from this version rather than from the 18 MHz version. 

One point of note is that the example table (34.4 in the data sheet), of baud rate values lists 12, 16 and 20 MHz values, but not 18 MHz. I wonder if there is some reason for this omission.

Any thoughts?

-Jason

Jason,

JasonHaedt said:

One point of note is that the example table (34.4 in the data sheet), of baud rate values lists 12, 16 and 20 MHz values, but not 18 MHz. I wonder if there is some reason for this omission.

I think the datasheet author just took the most common clock settings, so nothing should avoid you to use 18 MHz in this case.

Have you tried to use the USCI calculator/USCI header file in the link I gave before?

http://processors.wiki.ti.com/index.php/USCI_UART_Baud_Rate_Gen_Mode_Selection#USCI_UART_Calculator

Leo,

Thanks, we did use the calculator and see unexpected behavior at 18MHz that we cannot explain with the UART.  I do appreciate your feedback.Any other ideas besides verifying the register settings with the calculator?

-Jason

Jason,

have you tried to output the clock input (SMCLK) to a pin (e.g. P1.6) to see whether the signal is ok?

Anyway, are you using MSP430F5437 or MSP430F5437A?

I did output the clock on an I/O pin and verified that it is operating at the expected frequency. It is very strange that the bad boards works at 12, 16 and 20 MHz, but not at 18, but there must be some explanation. I also just ran a test program with all interrupts disabled, just toggling an output and examining the timing on the scope. At 18 MHz, the bad board ran slower than the good board.

We are using the MSP430F5437A