RTCComm Warning: The value should be adjusted to communication speed in High speed mode. If it is too low, only one byte from a message will be transmitted. (Polling trials)

[jamesskurski]

Using CW 10.6, I am getting this RTCComm component warning but am unsure how to resolve.

"RTCComm Warning: The value should be adjusted to communication speed in High speed mode. If it is too low, only one byte from a message will be transmitted. (Polling trials)"

I changed polling trails to the maximum value (65535) but that did not fix this. Any suggestions on how to resolve this warning?

[jamesskurski]

Does anyone have any suggestions on how to resolve and remove this warning?

[Rick_Li · Accepted solution]

HI James,

I'm sorry for delay replying, we are heavily backlogged!

regarding the problem, could you please let me know which MCU are you using.

could you please upload a screen-shot on that problem for investigating?

[jamesskurski]

Hi Yong,

I am using MCU MCF51JM128VLH. I have also attached a screen shot per your request.

James

[Rick_Li]

I'm sorry for my delay replying again!

regarding this issue: as is said in the help of this component, "Polling trials" , this property determines how many times an interrupt flag is tested during communication started by method SendChar, SendBlock....

in the generated code, the polling trials value is used as the max count number of loop in function MainComm().

if this number set too low, then, the application might not check the IIC flag in time and thus lost the communication.

but 2000 is a bit long. when change it to 20, the communication is OK. so, please just make sure that the IIC module has enough time to process the datas at specific bus frequency and IIC speed.

I hope it helps!

[jamesskurski]

Hi Yong,

I tried adjusting the polling trials to 20 and I still received this warning. I also reduced it to 1 and the warning still remains.

I noticed that if I changed the mode from Master to Master-Slave then back to Master then more PE code is generated and this warning goes away.

Most likely due to the fact that MainComm() no longer exists along with the variable Tr.

Do you have any other suggestions that I may try to eliminate this warning?

Thank you,

James

[Rick_Li]

Hi James,

I will contact the PE development team for further assistance on this issue and will keep you informed!

[jamesskurski]

Thank you, Yong.

I look forward to your response on how to resolve this issue.

Thank you,

James

[Rick_Li]

Hi James,

just got feedback from the Processor Expert development team regarding the warning message:

see it below:

The warning cannot be removed, except by enabling interrupts. The point of the warning is to caution the user about the constraints for setting the number of polling trials. The way polling mode works is that communication is started and then the I2C Status Register is read inside a loop to check if the current character transmission or reception is complete. The "Polling trials" property sets the maximum number of iterations of that loop. If this number is too small then the loop will end before the first character transmission/reception is completed and the function will return an error. If this number is too big then communication errors will take a long time to be reported and the cpu will be bloked inside the loop all this time.

    Processor expert can't compute the right number of polling trials because the execution time of one iteration of the loop is compiler-dependant. Also, the transmission/reception time of the character can be longer than configured (if the slave or another master slows the clock). So it is up to the user to configure the polling trials properly, Processor Expert only issues this warning to highlight the problem.

[jamesskurski]

Thank you, Yong and the PE team for the response. We'll decide whether to live with the warning or live with the unused generated slave code.

Thanks again,

James

[jamesskurski]

Hi Yong,

Was the information and screen capture I provided helpful to you in determining why this warning is occurring?

Thank you,

James