Siemens S7300 FB41 PID help

[metinerim]

FB41 out is always zero

Hi,

I made a sample program for using FB41 in OB35 as PV_PER=PIW252 and PQW272, PV_FAC=1, PV_OFF=0 and PVPER_ON=1



Can anybody have an idea why it is so?

 

[boneless]

What are your other values?

Please post a SS of the DB used when online.

 

[metinerim]

What are your other values?

Please post a SS of the DB used when online.

When I sent the PV_PER (PIW252) input as 65.0, the LMN_PER (PQW272) value goes to zero and remains zero whatever the input value is.

 

[boneless]

That is wonderful, but without knowing your other inputs/setpoints this will be a guessing game.

Most important:
cycle (and in which OB is your FB called)
SP
Man_ON
Gain
Ti
Td
Deadband

 

[metinerim]

Here are the two pictures of FB41 when online. As can be seen from the pictures , the output value goes to zero

 

[metinerim]

and OB35 cycle time is 100ms

 

[boneless]

A 1s cycle seems wrong. What OB are you calling the block in?
We usually use 100ms cycle in combination with OB35 (100ms)

Your gain is tiny, making the error value very small.
No deadband declared, so it is likely 1 as initial value.
Your SP seems to be (close to) 0.

Maybe you can explain some more on what you are trying to do?

What is your analog input connected to? What expected value? As in real world value..
What is your analog output connected to?

 

[metinerim]

Boneless,
Thanks for replying. I've fixed the problem by setting the I and D selection as 1 and then the PID block starts to run.

 

[metinerim]

FB41 goes to zero

Hi everybody,

The function block FB41 output LMN_PER is going to zero regardless of PV_PER input.
Can anybody have a comment??

 

[Thomas_v2]

You have a setpoint of 7*10^-44, thats:
0.00000000000000000000000000000000000000000007

Why should the output go up, when your setpoint is about zero, gain is positive and the process value above the setpoint?

 

[boneless]

I have given you several pointers, so far you seem to have followed about 0 off those. Check the comments above with your program and try again...

 

[Peter Nachtwey]

Forget about the FB41 for now. The most important part is understanding your system.
dPressure/dt=PressureRate=BulkModulusOfOil*Flow(t)/VolumeOfOilUnderCompression
The BulkModulusOfOil is about 200,000 psi.
Flow(t)=RatedFlow*Voltage(t)/10*sqrt((SupplyPressure-Pressure)/(2*RatedPressure))
Voltage(t) = Kp*(SetPointPressure-Pressure)

If there is no leakage in the system, and most test systems don't leak, then pressure is an integrating process. You can probably get by with just a P gain that is easy enough to do without the FB41. The problem is how fast will the pressure change each scan or 10 milliseconds. If the volume of oil being compressed is large then the rate of change will not be high. However, if you are testing short pieces of hose the volume of oil will be high and the pressure rates may be too fast for the PLC to handle.