[Pascone10]

So the defaults are set for 1 and 4. Its said if I recall that the (I believe) interval setting must be a multiple of the program refresh rate. How do I find what that is? Anyone familiar with that? Also is the int in seconds?

I have a MAT setup and the dampers are going crazy to maintain with these setpoint..

Thanks guys

 

[#1servicetech]

The ratio is 4 to 1. so if the prop band is 4 the integral is 1. I usually put a division block in and divide by 4 so that I just need to adjust the prop band.

 

[stability]

What do you regulate? For an ambiant T° I recommend: P-Band about 24 K (means an amplification of 4 %/K) and an integration time of about 600 seconds. With these values the dampers won't go crazy.

 

[scrooloose]

What is your P, I and PID run frequency?

 

[stability]

Sorry but I don't know at all Rover regulators and the calculating of the numbers which represent the PID parameters. Can you give me a hint where to find in the internet an installation manual of this device please? If the program refresh rate can be selected then let the device do it as often as possible with a minimum of once each five seconds. Every temporization is deadful for the stability of the regulation. Imagine you regulate the pression of a ventilation and you scan the value once each minute...

 

[refrige-nate]

The ratio is 4 to 1. so if the prop band is 4 the integral is 1. I usually put a division block in and divide by 4 so that I just need to adjust the prop band.

I agree, This is Exactly what Trane recommends. Use the Plot function in TGP to help you, also add the feedback input into the PID.

 

[Pascone10]

I agree, This is Exactly what Trane recommends. Use the Plot function in TGP to help you, also add the feedback input into the PID.

So that is a interval of 1 second?

 

[refrige-nate]

Integral you mean? Just Take whatever your "P" is (hopefully an AV) and run that into a divide block, Divide by 4. Take the output of the math block and run it into the "I". Then adjust your "P" as necessary. to add the feedback input open the PID block properties and check the "Allow Feedback" box. This will open a new slot on the PID called Feedback. Take an Output Status block of whatever AO your using and link that into the feedback.

 

[refrige-nate]

.

 

[Pascone10]

Integral you mean? Just Take whatever your "P" is (hopefully an AV) and run that into a divide block, Divide by 4. Take the output of the math block and run it into the "I". Then adjust your "P" as necessary. to add the feedback input open the PID block properties and check the "Allow Feedback" box. This will open a new slot on the PID called Feedback. Take an Output Status block of whatever AO your using and link that into the feedback.

Ok i got you there, but yes my question is "is the integral in seconds?"
Also what is the advantage of having the feedback input ?

 

[scrooloose]

no integral is not in seconds.

Advantage of feedback is the PID knows last value. So if you have another program controlling the value based on something (freezestat, low limits, etc..) it knows where to start from.

 

[Pascone10]

OK. In the tridium stuff the integral is in seconds so I was wondering what trane used. Thank you guys for the help. I'll use that divide block. Thats a neat trick. Any recommendation on a good starting point?

 

[scrooloose]

P of 2 and 20 sec PID runtime to start be worth a shot.

 

[tech45]

Could someone explain the " Fail " connection on the PID block please ? I often see a " fail block " connected here and wired to the measured variable. Is this checking for an analog value of " 0 " ?

 

[scrooloose]

Its checking for a value out of range. So if your input is setup as 4-20ma, you will get a fail if it is below 4 or above 20. Thermistors out of there temperature range, etc..

When the fail is true the PID goes to the fail output that is setup in the PID properties. It stays out this output long as the fail point is true.

 

[tech45]

Its checking for a value out of range. So if your input is setup as 4-20ma, you will get a fail if it is below 4 or above 20. Thermistors out of there temperature range, etc..

When the fail is true the PID goes to the fail output that is setup in the PID properties. It stays out this output long as the fail point is true.

Thanks. I see this used often enough so I'm going to assume that the fail occurs somewhere towards the end of the program run time ( or enough time for a measured variable to be produced ) ? I am thinking like a fan or pump starting from a stopped condition with a measured variable of " 0 " .

 

[tech45]

Anytime the Fail input port receives a true value, the output goes to its fail-safe position.- TGP Help

Maybe it is best to not start the PID UNTIL a measured variable is present for something like a fan or pump. I see this a lot too ( factory programs ) with a status input as the enable.

 

[refrige-nate]

Make sure that in Rover configuration your Supply temperature is indeed setup to "fail at end of range". There is a little checkbox in the config that needs to be checked otherwise the fail logic in TGP wont work.