Sensing a 'Fall' in Temperature?

[scimdarren]

Hi All

Using Step 7, ladder logic. I need to program the following sequence:

If temperature starts to fall, open internal fuel gate to feed fuel.

So I need to operate a digital output based on a 'decline in temperature', any advice, please, on how to capture a decline in a set value? I guess I am looking for a Math function?

Cheers

Darren

 

[RheinhardtP]

Save in History

Hi scimdarren

If you are controlling around a setpoint you would have to cater for overshooting your setpoint by using a PID loop,

On your question

What if you were to save the value of each previous scan in a History Register of the temp, then in the new scan do a smaller than comparison of the value to the actual value in the scan and set Fuel bit accordingly.


Compares Value of Previous scan to updated current Value
-[ < History.Temp Current.Temp]----------------( Open Fuel)--

Saves the Current value in the History register
-[move Current.Temp History.Temp]----------------------------

Although it sounds like you would have to use a PID loop for this operation.

Do you require a setpoint for the controlled temperature??

Regards

 

[PLucas]

Darren

I do something very similar to this using a series six PLC to detect absolute encoder loss. I haven't done this in Step 7, but I am sure the idea could translate across.

What I do is compare the 'count' on this scan with the count on the last scan, this is acheived by moving the data to another memory area and comparing the moved data with the new data read in.

For example:-

L PIW256 //Load analogue input (temperature probe)
T DB0.DBW0 //Transfer to DB area (set as INT for this example)
		 //This is the temperature this scan.
 
L DB1.DBW0 //Load temp reading from previous scan
L DB0.DBW0 //Load 'current' temperature reading
<I		 //Compare for less than
=#temp_loss //Set flag temperature loss
 
Then at the end of the block....
 
L DB0.DBW0 //This part loads the current temperature reading and
T DB1.DBW0 //transfers it to the 'temperature last scan' area.
	 //As the transfer takes place at the end of the block
	 //the data in DB1.DBW0 does not change until this 
	 //point, thus the earlier compare uses the updated 
	 //data from PIW256 and the 'old' data in DB1.DBW0

Hope this gives you an idea on how to acheive what you want.

Paul

P.S I see RheinhardtP replied while I was typing, he gives an answer similar to what I have said.

 

[JohnW]

Hi Darren,

You may want to slow down the sampling using a timer, every scan is too fast for a temperature input, you will get fluctuations that will trigger a response but it is not really a fall in temperature. The timer interval wants to be longer than the filter time on the analog input.

 

[ayman metwally]

Greetings Darren,
Do you simply mean what is called a "high rate of change" alarm (Negrative change)that will open internal fuel gate to feed fuel.
If so, you have to specify this rate of change
Ex: 10°C\minute
1- Say that you will record a vale of your temp. each 10 sec, each one in a known memory register (you need six)
2- Some Math: If difference between the highest and the lowest value of these readings is more than or equal 10°C (according to our example)then you have a rate of change case (open internal fuel gate to feed fuel).

That is my idea and I need your openions about it.