Now I'm using cpu315-2pn/dp() and fm3501(6ES7 350).before starting counting work, I preset the load value to 0, the comparison value 1 to 5000.

When count work started, a few time later OB40 trigged.In OB40, I read the actual value DBD34(created from UDT2).The DBD34 value is smller than 5000,Maybe 4995 or 4998.The value changed with the encoder's frequency.
I also test using without FC program.In OB40,I read the actual count value using PID260(the FM350-1 starting address is 256). The value PID260 is larger than 5000.

So read the count value in ob40 is not equal to the comparison value when comparison value reached?

Anybody can help me and give some explaination?
Thank you very much!

viralpatel

hi,
can you give more detail

have you made program in OB40 or OB1 ,for comparison?
and you mean to say your counter not reset to 0 after comaprison?
viral patel

baobo

viralpatel

hi,
can you give more detail

have you made program in OB40 or OB1 ,for comparison?
and you mean to say your counter not reset to 0 after comaprison?
viral patel

Yes ,I have programed in OB40 and OB1.
The attachment is my project!
In OB40, I read the actual count value and found that the actual count value was not equal to the current compapsion value which trigged the OB40 hardware interrupt.
In OB40, I also load next comparison value with the sum of current compasion value and 10000.

For OB 32 you configured a scan time of 1000ms, this scan time is not as accurate, like it should be for frequency measurment/calculation, there will be always a little jitter. Then we have the communicatuion from the CPU to the FM, which has also a little jitter, then we have the refresh time of the Interface of 0,5ms,also with a little jitter.

Now you supplying a frequence of 1000Increments/s (which may have a little jitter as well) to the FM Module, count the signals, and in OB32 you compare the differnce between to counts.

If due the combination of the small jitter and every part now just 1increment is missing, you will recieve as result 999, and since that their is no increment loss, you will another time then 1001.This is typcialy for this kind of frecuency measurment/calculation. The measurement error will in this kind of measurement always be 1 Inkrement!

To minimise this measurement error you may increase the frequency supplied to the FM or increase the Scan time of OB 32.

I'm afraid OB3x cyclic time is longer than 2ms!
In my project, I should ues DI module SM321's positive edge hardware interrupt. So OB3x's cyclic time
is not constant.
Maybe I should change my confiruration and use isochronous mode!