Configuring a Spyder PID to modulate a vfd based on building pressure

[agent4g94]

Anybody have any suggestions for me about tuning a PID loop that looks at building pressure with a setpoint of .005"? Im having trouble slowing down the loop to prevent the fan from ramping up and down too fast. The loop output is then forwarded into a reset block to convert it into fan %. Please help i'm losing hair by the handful!!

[BACnet]

I know this may sound a bit simplistic- but have you tried to simply reduce your P, I and D constants to be smaller? Can you divide them each by 10 for example?

[viceman]

there is a rate limit block (i think thats what its called, sorry i dont do webs anymore) that you can use to slow down your pid output.
it goes inbetween your pid and ao. you can limit the up and down rate.

although the best way is to tune your PID.

[Big Unit]

I would say tune the P.I.D. and it will take a while. If you're not real familiar with it plan on pulling out a lot more hair

[amigo]

Anybody have any suggestions for me about tuning a PID loop that looks at building pressure with a setpoint of .005"? Im having trouble slowing down the loop to prevent the fan from ramping up and down too fast. The loop output is then forwarded into a reset block to convert it into fan %. Please help i'm losing hair by the handful!!

are you sure your setpoint is 0.005 ? that seems awfully low.

I think that should read more like 0.05 for building pressure.

[agent4g94]

ViceMan, you are the man. Duh, rate limiter, i should have known. What an oversight on my part!! I guess i should post on here more, my question was answered in 1 hour. Its too bad my boss paid for three days of PID troubleshooting. haha, sucker

[agent4g94]

Amigo, It's a spec from some dumba$$ engineer. Its too bad too, when i fart in the building it sends my fan for a ride. Ploop, oh wait the fan just went to 50Hz.

[amigo]

Amigo, It's a spec from some dumba$$ engineer. Its too bad too, when i fart in the building it sends my fan for a ride. Ploop, oh wait the fan just went to 50Hz.

Yeah, that will sure keep your fan busy, especially on days when chilli is on special,

[kostasgreece]

Slowing down the PID cycles IMO wont affect significally your fan behaviour, especially when you try to modulate this output based on a process variable on that range with a setpoint somewhere on that range of values.

Why? Because the variation of PV shall exist no matter when you choose from PID settings to see it and act. You would see something like PV cange - no reaction - after a while jump on a reacted output... believe me..

The best way to go is to tune up your PID, and maybe in some cases apply an antispike filter on PV (be careful not to brick the input...)

Cheers, have fun with PID tunning

[agent4g94]

The rate limiter function worked wonders for my program. It's too bad ill never get that hair back!!! Check out this printscreen of the graphic, working well

[gnomethang]

The rate limiter function worked wonders for my program. It's too bad ill never get that hair back!!! Check out this printscreen of the graphic, working well

Glad you got it sorted. What did the screen dump look like 10 seconds later?

Seriously - I always try to filter the PV in any loop. The degree of that is always determined by the application.