Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

PID method control on Heater using microcontroller

Status
Not open for further replies.
B

Bus Master

Full Member level 3
Joined
Dec 9, 2001
Messages
186
Helped
9
Reputation
18
Reaction score
3
Trophy points
1,298
Location
Just behind You ;-)
Activity points
1,455
pid heater control

Hi all,

I'm implementing a PID control algorithm on a heater using microcontroller. I intend to use Ziegler-Nichols tuning method, which needs a clear Step-response graph for determining the PID equation parameters.

my question is:
do i need to apply a 100% power to the heater and record system response until it reaches a steady-state output? i.e., does the Step input has to be a 100% power? or it can be of any value, just an instant change in the heater power input?

please advise as this is my first time to implement a PID control method.

Thank you in advance for your time.

Yours,
 

pid heater controller

Assuming a linear system behaviour (very likely) the step can have any amount that gives an observable response.
 

    B

    Bus Master

    Points: 2
    Helpful Answer Positive Rating
pid microcontroller

No need 100%.
U can set 10% of Full scale to heater, study the respond, then correct the PID Par. Then give 10% again, If OK, give 20%..Do the same to 50% if system is linear or up to 100% if non-linear (Depending on what kind of Heater)
 

pid for heater control

I recommend you to set the power (open loop fashion) for reach the steady atate temperature that you want to hold in closed loop. To say it, you need about 50 % pwm for a temperature of 50ºC. When the steady state has been reached you apply a low level step change (say a change from 50% to 60%) . The level change of amplitude depends on the signal / noise ratio.

When the reaction curve has been obtained you can use a method called AMIGO which is too better for PID sinthonizing.

Finally I recommend you the following code for PID implementation. This code has antiwindup protection and derivative filtering.

**broken link removed**

please see the chapter 10 of this excelent book

If you have more question do not hesitate to contact me.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top