SCL totalizer does not work

hello folks

I have written a totalizer using SCL. it totalize flow and also gives the amount of elapsed day,hr,min and sec. everything is working fine until it passes about 24 days and all of sudden it gets halted.

the code is like this and the HMI is attached:

Hello JabbarKamali;

What data type are you using for "my_time"?

If it is an IEC time (from SFB4 "TON", for example) you are limited to about 24 days. Coincidence?

The screenshot below is from the Programming with Step 7 manual:

https://support.industry.siemens.com/cs/ww/en/view/109751825

One possible alternative to  limited timers time span is to create your own, using the Clock byte to generate a 1-second pulse and ADDing 1 to a DINT register every second., for example.

Hope this helps,

Daniel Chartier

dchartier

If it is an IEC time (from SFB4 "TON", for example) you are limited to about 24 days. Coincidence?

Looking at the "TRUNC" he is rather using REAL. Then 2097152 can be not a coincidence as well.

But let him speak for himself.

jacek d

 

dchartier

If it is an IEC time (from SFB4 "TON", for example) you are limited to about 24 days. Coincidence?

Looking at the "TRUNC" he is rather using REAL. Then 2097152 can be not a coincidence as well.

But let him speak for himself.

 you are right. I am using ob35 and hence it makes sample_time 0.1. I tried to add counter(my_time) with a constant i.e 0.1, however it halts at the same number. one more time I make my constant 10, such that in each cycle it counts 100 times. with this condition when it reaches 2097152 it doesn't get halt and continue counting. 

JabbarKamali

 

jacek d

Looking at the "TRUNC" he is rather using REAL. Then 2097152 can be not a coincidence as well.

But let him speak for himself.

 you are right. I am using ob35 and hence it makes sample_time 0.1. I tried to add counter(my_time) with a constant i.e 0.1, however it halts at the same number. one more time I make my constant 10, such that in each cycle it counts 100 times. with this condition when it reaches 2097152 it doesn't get halt and continue counting. 

Be patient. It will stop. In ca.  6 years and 9 months.

But in a meantime you may manage to get familiar with IEEE754 and its weaknesses - specifically with finite resolution of floating point numbers.

You can find a lot of topics on the Forum starting "totalizer not working ".  Find them.

jacek d

But in a meantime you may manage to get familiar with IEEE754 and its weaknesses - specifically with finite resolution of floating point numbers.

You are the REAL^* champion!    

* floating-point-numbers

qwazee

 

jacek d

But in a meantime you may manage to get familiar with IEEE754 and its weaknesses - specifically with finite resolution of floating point numbers.

You are the REAL^* champion!    

* floating-point-numbers

At least I've learnt designation of the standard by heart. 

jacek d

 

JabbarKamali

 

jacek d

Looking at the "TRUNC" he is rather using REAL. Then 2097152 can be not a coincidence as well.

But let him speak for himself.

 you are right. I am using ob35 and hence it makes sample_time 0.1. I tried to add counter(my_time) with a constant i.e 0.1, however it halts at the same number. one more time I make my constant 10, such that in each cycle it counts 100 times. with this condition when it reaches 2097152 it doesn't get halt and continue counting. 

Be patient. It will stop. In ca.  6 years and 9 months.

But in a meantime you may manage to get familiar with IEEE754 and its weaknesses - specifically with finite resolution of floating point numbers.

You can find a lot of topics on the Forum starting "totalizer not working ".  Find them.

 oh man, you hit the spot. you have really absorbed these concepts. Thanks Mr Big Brain.