{"id":4460,"date":"2020-12-23T03:59:44","date_gmt":"2020-12-23T03:59:44","guid":{"rendered":"http:\/\/fastbitlab.com\/?p=4460"},"modified":"2022-11-22T14:27:50","modified_gmt":"2022-11-22T08:57:50","slug":"stm32-i2c-lecture-37-exercise-testing-repeated-start","status":"publish","type":"post","link":"https:\/\/fastbitlab.com\/blog\/stm32-i2c-lecture-37-exercise-testing-repeated-start\/","title":{"rendered":"STM32 I2C Lecture 37 -Exercise : Testing repeated start"},"content":{"rendered":"
\n
\n
\n
\n

 <\/p>\n<\/div>\n

\n

Exercise : Testing repeated start<\/span><\/h1>\n
\n
\n
\n

 <\/p>\n

<\/div>\n

 <\/p>\n<\/div>\n<\/div>\n<\/div>\n

\n
\n<\/div>\n

In this article, let\u2019s understand the concept of repeated start. Why repeated start is used and how to change our APIs to make use of repeated start all these things we are going to learn now.<\/span><\/p>\n

The waveform obtained from the master receive data exercise is, as shown in Figure 1.<\/span><\/p>\n

\"\"
Figure 1. Trace of master receive data exercise.<\/span><\/figcaption><\/figure>\n

 <\/p>\n

If you analyze the waveform in Figure 1, the first I2C transaction is master write, which starts from the first green circle and ends at the first red circle, as shown in Figure 2. <\/span><\/p>\n

The stop condition is generated at the end to stop (or end) the first I2C transaction. The stop condition in Figure 2 is not required. Instead of raising the stop condition at the end of the first I2C transaction, you can go ahead and generate one more start condition in order to begin the next I2C transaction.<\/span><\/p>\n

<\/div>\n
\"\"
Figure 2. Trace for first I2C transaction (master write).<\/span><\/figcaption><\/figure>\n

 <\/p>\n

Look at Figure 3. The green circle pointed with the cursor is the beginning of the second I2C transaction, and the red circle is the end of the first I2C transaction.<\/span><\/p>\n

\"\"
Figure 3. Trace that marks the beginning of the second I2C transaction and end of the first I2C transaction.<\/span><\/figcaption><\/figure>\n

 <\/p>\n

Observe Figure 3, the bus is released for 5 microseconds, and both the SCL and SDA lines are also released, allowing another master to claim the bus. Therefore, the stop condition is generated at the end of the first I2C transaction because you may not be able to complete all other transactions on time since the SCL and SDA lines are released. <\/span><\/p>\n

Generating a stop means giving a chance for another master to claim the bus.<\/span><\/p>\n

If you think your data transaction is critical and you have to complete all the I2C transactions on time, then go ahead and generate one more start condition instead of generating a stop at the end of the first transaction. When you generate one more start condition, that will be called a repeated start. <\/span><\/p>\n

Repeated start means generating start condition without generating stop condition for the previous I2C transaction.<\/span><\/p>\n

 <\/p>\n

Now let\u2019s make the required changes to the code to include repeated start:<\/span><\/p>\n