Sending a break signal on SCI (UART)

Jude,

See if this is what you are looking for:  (Section 13.3.4.1 in case the copy is not clear enough)

13.3.4.1  Idle-Line Multiprocessor Modes

In Idle-line multiprocessor mode, a frame that is preceded by an idle period (10 or more idle bits) is an address frame. A frame that is preceded by fewer than 10 idle bits is a data frame.

There are two ways to transmit an address frame using idle-line mode:

Method 1:I n software, deliberately leave an idle period between the transmission of the last data frame of the previous block and the address frame of the new block.

Method 2: Configure the SCI to automatically send an idle period between the last data frame of the previous block and the address frame of the new block. Although Method 1 is only accomplished by a delay loop in software, Method 2 can be implemented by using the transmit buffer and the TXWAKE bit in the following manner:

Step1 : Write a 1 to the TXWAKE bit.

Step2 : Write a dummy data value to the SCITD register. This triggers the SCI to begin the idle period as soon as the transmitter shift register is empty.

Step3 : Wait for the SCI to clear the TXWAKE flag.

Step4 : Write the address value to SCITD.

Kevin,

I think it might work. Just need to get my way around it I guess.

The reason I was after the break detect in the first place was to trigger a DMA receive upon a "created" break. So typically, the sender would issue a break and follow that by data which the DMA would receive and interrupt the CPU after the DMA buffer is full.

How would I use the idle-mode to do this? Can the idle phase give me the interrupt which the DMA can rely upon to trigger?

Jude,

If you are trying to detect a BREAK, why not directly use the BREAK DETECT interrupt?

Or if your ultimate goal is to use the DMA to transfer received data, set the RX DMA bit (by writing a 1 to SET RX DMA). This will send a DMA request signal when the receive data is ready. You can configure the DMA to interrupt the CPU after some number of transfers...

Best Regards,

Forum Support

Kevin,

The issue I was trying to get at was to _create_ a deliberate break, detecting the break wouldn't be an issue. Having said that, your suggested approach to the DMA sounds like the way I need to approach the setup I've got. I'll go and give that a try.

Cheers,

Jude.

Got the DMA working with SCI. This issue can be resolved.

Thanks!