INA226: Address error

Hello,

Address 0x40 is correct for the configuration seen in your circuit. Please see table below.

Please refer to the datasheet pages 19 and 20 for correct i2c communication with the INA226 part.

What speed are you operating the i2c communication?

Regards,

Cas

Hi Castrense,

Would you please help me to check the schematic? i2c communication are as follow

Thanks a lot.

Hello,

The INA226 is ack on the ninth clock pulse. Your code does not seem to be doing anything but repeating the sending of the slave address.

So the first frame (frame 1) is the slave address being sent with a write (0 on pulse 8). The INA gives an ack (0 on pulse 9). Then you must keep I2C communication and send the register you want to write data to. See table 4 in the INA226 datasheet. The next two frames will be the data written to the register set in frame 2. After the 4th frame I2C communication can end. This is for a write to a register. 

The start/restart should not happen between frames when trying to write to the INA. 

Below is the timing diagram for writing to the INA226. 

I am guessing you will also try to read from the INA226, if you have any questions about that or what is written above please reply.

Regards,

Cas

Hi，Castrense

         I am the question of this address, and now the situation is as follows: regardless of the read and write timing configuration of the subsequent data, we write 0x40 into the address. According to the oscillograph, the ninth position is low, but in fact, we print the ACK value of this block as 1, which is supposed to be just picked up the rising edge of SDL. Because our code directly reported an error in the first step, it is impossible for the problem to appear later, because we have not gone to that step. I would like to ask you that you have excluded the problem of the wrong device address in this case, and what might be the cause?

         Thanks!

Hello,

I would like to better understand if the problem is the code or our device. Are there any other I2C devices on the board you can communicate to and are ACKing correctly? If so what devices are they? Are you using 7 bit addressing when communicating with the INA?

Regards,
Cas

Hi！ Castrense

     I  use 7-bit address for communication , and there are other devices attached to this I2C bus,such as Flex  chip -BMR485,BMR685,they also receive the ACK signal back normally

Hello,

Is it possible for you to produce a scope shot when the error appears in the code? Does your code keep trying to speak to the INA if it receives a 1 for the ACK? If the code receives a 0 for the ACK what are the next steps in the code?

Can you replace the INA226 and see if the problem persists?

Regards,

Cas

Hello!

     A scope shot ?  When the received IACK signal is 1 during INA communication in the code, it will directly goto CLOSE_EXIT and report error until the next I2C communication. When the received ACK signal is 0, it will prove that there is a reply signal and it will start to send the register address.  After receiving ACK response, a START signal will be generated again. I2C starts the cycle of reading a byte sequence or writing a byte sequence until the completion. and When I replaced to a full machine, the problem persisted.

   Thanks!

Hello,

Sorry I would like to see the INA not ACK on the oscillograph. If you can get a picture of the oscillograph when the program reads a not ACK that will determine if it is the code or the INA. You mentioned the oscillograph picture above is when the code already threw an error and the problem cannot happen again. I would like to see when the problem happens what the SDA and SCL lines are actually. 

I will also speak to another member of the team to see if they have seen this before.

Regards,

Cas

Hello,

I should have asked about anything else connected to the SDA/SCL lines that could hamper the communication sorry about that. I saw there is a p2p replacement in the ISO1640/1641 but they still have the diode with 0.65V will this be enough to consider a low? Is it necessary to have an isolation barrier in the i2c IC?

Regards,

Cas

     Hi!

    We must have this I2C isolation circuit device, because the grounding of the two sides is different, so is there any good solution to this problem?Iso1640/1641 Low level 0.65V is still high.

    Thanks.

Hello,

I will move this thread to the team that supports the i2c isolation devices. They will be better help than I can be.

Regards,

Cas