# Determining the transfer function of a humidity chamber

1. Mar 27, 2012

### Ben Jamin

I have created a mock up of a humidity chamber which i will soon be integrating into a temperature chamber for undergraduate project. I am confused on how i can control the system. I have currently a humidifier with a relay built in, this relay is going to switch on and off from a PWM signal generated in a controller. Also within the chamber i have a mini fan to mix the steam and a analogue voltage output humidity sensor which i can use to get feedback from the system. At the moment all of these are linked up to a computer but i need to design the controller. I was thinking PID.

Im having trouble modelling the system to work everything out. The plan was to introduce a step response in terms of duty cycle so say 20% and then check the corresponding voltage output obtained from the sensor.

Im having trouble working out how i can find the transfer function of the system and generally modelling the system for control analysis. How can a step response in terms of % create a transfer function with an output in volts.

The humidifier runs at 6 volts so its either switched on (6 volts) or off (0 Volts)

Once iv done this i need to design the controller

Any advice would be great thanks.

2. Mar 27, 2012

### Staff: Mentor

Welcome to the PF.

The humidity chambers that we use at work typically are pretty closed systems, so you don't lose much humidity from just air circulation. So you need to add humidity very slowly, or you will cause condensation, which is a bad thing generally in humidity testing.

So I would focus on how you can diffuse the moisture well in the air stream before it gets into the main chamber, to be sure you don't get condensation. Then you can do some experiments to see what it takes from the humidifier to slowly ramp up the humidity in the chamber. In most chambers that I've worked with, it takes hours to ramp the humidity up and down.

3. Apr 14, 2012

### Ben Jamin

Thanks for your reply, my lecture recommended that i simply insert a generic household or office humidifier into a cardboard box and introduce a step response in the duty cycle (It runs off a PWM signal). Then he recombined i develop a controller with the ability to alter the humidity given a set point.

From your response it seems my main concern is mixing the steam into the airstream, the problem is the humidifier runs off a system where its switched on, or off. Theres no control of how much steam it produces. How would you recommend me ramping up the humidity just for the purpose of developing the controller. In the main chamber it will be introduced through a piping system which circulates the air using a fan and will be coupled with a temperature system. Hopefully the controller can be developed to decouple the two by introducing a step response to each system and finding the inverse of these to alter the transfer function.