#dP

@EXCAVATOR:

It is impossible to make suggestions without seeing the full scope of the code. It is not impossible that the recording is not giving a true view.

Please do what this expert suggested ->

jacek d

Connect 2 counters to the timer output and to the memory bit. If their counts agree then everything is OK.

 I have put 3 counters in to the code (attached). I'll count "timer Q contact", "timer output itself" and "the coil". I'll do the test in couple of hours. I hope to find result there.

Looks like an old story. In fact, you call the timer TWICE. AFAIR, accessing outputs (Q, ET) of an IEC timer on S7-1200/1500 causes its instance to execute.

So, following your last screenshot. If it happens that the timer block (the second network) sets the output then accessing the output (the first one) will clear it and the memory bit "loses" this pulse.

BTW, what is the logic at the input of the timer block?

More readings on the topic here and here.

jacek d

Looks like an old story. In fact, you call the timer TWICE. AFAIR, accessing outputs (Q, ET) of an IEC timer on S7-1200/1500 causes its instance to execute.

So, following your last screenshot. If it happens that the timer block (the second network) sets the output then accessing the output (the first one) will clear it and the memory bit "loses" this pulse.

BTW, what is the logic at the input of the timer block?

More readings on the topic here and here.

 The problem has been solved. Thanks to everybody helping out. Especially these last 2 topics pointed out the correct source of the problem. I have used a coil and temporary variable right after the timer and used this variable throughout the logic. This time it didn't fire timer every time.