Tuning PID - Temperature

[LuizMarques]

---

Hi all i've been woking on a temperature PID controller in order to cool/thermal box a through a fan(controls air flow of a compressor) and a peltier element

Before the actual control my box drives off unwanted moisture and volatiles( "inicio" and "defrost" process)

the peltier element has a fixed power within a temperature range

from -10 up to 0 ---- 2.4V .

From 0 up to 35 --- 5.5V

and above 35 degress --- 8V

i recorded temperature deviation on manual mode setting  99% PWM output at a 445ms per sample . Afterwards i removed the offset data and i used PID tuning . Allthough it's identified a first order model ,PID gains are so low compared to what it has to be. why?

[GerdW]

Hi Luiz,

your two Matlab/Simulink images show a very different time scale: one is in ms, the other in s. When x axis is in s then you get very low Kp and high Ti values…

In your Excel sheet you forgot to include units at all, so it will be of no help either!

On your VI: it greatly improves after using AutoCleanup. Why don't you use Autocleanup?

[LuizMarques]

Hi GerdW! thanks for answering. 

Infact time scale was wrong! Now step response actually meets what was on excel data. 

even so I'm still getting low gain values . I wonder why...

what I am trying to do is run the analysis of the system plan by raising the response curves for PWM 99% 80% and 60% to perform a mean between the plants and verify the best gain for the system. PID controls the fan while peliter maintain its power fixed on different ranges of temperature

 i'll make sure to use autocleaner afterwards... 

[GerdW]

Hi Luiz,

even so I'm still getting low gain values . I wonder why...

Kp is the relation between change of PID output value and corresponding change of pv: both changes are rather small in your image…