Exceptions with baudrate 9600

[UlliBien]

Hi, I'm using espsoftwareserial to communicate with a PZEM-004T using library https://github.com/olehs/PZEM004T.

I got a lot of reading errors and mostly an exceptions occured after a few seconds. Somtimes it was stable for nearly a minute, but never longer than two minutes. When I got an exception epc1 mostly pointed to optimistic_yield().

In the interrupt-handler rxRead I found the line WAIT;. This macro expands to #define WAIT { while (ESP.getCycleCount()-start < wait) if (!m_highSpeed) optimistic_yield(1); wait += m_bitTime; }

m_highSpeed is set in begin(): m_highSpeed = speed > 9600;. The baudrate for the PZEM-004T is 9600 bps. So m_highSpeed is set to false and optimistic_yield() is called.

I set m_highSpeed to true for my special case. All the communication errors disappeared. And now more exceptions occured. Currently my software is running stable for more than a hour.

Yous should revise the definition of WAIT. In the code for WAIT in the original library is #define WAIT { while (ESP.getCycleCount()-start < wait); wait += m_bitTime; }

Kind regards UlliBien

[TD-er]

Indeed, using 9600 baud will result in timeouts and corrupted messages. Simple change in begin(): old: m_highSpeed = speed > 9600; new: m_highSpeed = speed >= 9600; Results in stable reading.

@UlliBien What do you mean with "more exceptions"?

[UlliBien]

Sorry, not "now more", "no more" instead. After making sure m_highSpeed was true, the exceptions disappeared.

[devyte]

@TD-er said:

Simple change in begin(): old: m_highSpeed = speed > 9600; new: m_highSpeed = speed >= 9600;

@plerup there is one user in the Arduino repo who confirms the proposed fix. Without it, behavior is flaky.

[plerup]

As your mileage may vary here I'm going to add a new method where you can specify what "highSpeed" actually is, but the default will be changed as you suggest

[plerup]

Done in latest commit