Temperature and Humidity Control Problems

This post came about when using a pretty standard curing chamber setup, I found I was having some interesting temperature and humidity control issues. Below describes these issues and how I solved them.

Setup

As a curing chamber I modified a fridge following the general guidelines readily available on the web. My specs are as follows:

• ·      Free standing full fridge unit (no freezer compartment): 360 litre
• ·     Humidifier: ultrasonic
• ·     Fan 150mm 220V electrical case fan (big brother to a standard computer    casing fan)
• ·    Hanging rack – wooden. I will later change this material or paint it.
• ·   Controller: Lilytech^® ZL-7801A. This controller has some nice features in that it can control temperature based on heating (connect it to a heat source) or cooling (connect to fridge power) it also has two other timer switched outputs. One is NC (Normally closed) and the other NO (normally open) on the same circuit so the power effectively toggles between them according to how you set the time. I’ll get to this more later. Setpoint adjustment is also really simple with this unit.Out of interest, this controller seems to be used typically in the chicken egg incubation industry where the additional timers are used to turn the eggs over.

Chamber overview

Control Panel Mount

Components

Controller

Ambient Weather

Some further input to understanding my initial temperature control problem. I live in Johannesburg South Africa. It is mid July (middle of winter) typical temperatures are as follows:

Time of day	Outside temps °C	Inside my garage
Daytime high	Varies 14 to 18	10 to 14
Typical overnight	Around 8 to 10	8 to 10
Morning low (the dip as frost melts)	0 to 3	Slight dip to about 6

Humidity is low averaging around 35%RH

When hanging my first batch of Chorizo and Salami together I put them in the chamber with a heater to control around 20°C with humidity set point 90% for fermentation then connected the controller to the fridge and dropped the humidity set point to 80%.

At this time I noticed significant swings in the humidity and with a little help I added an Arduino based temperature and humidity monitor with a history trend into the fridge. For the history trend this is connected to www.blynk.cc and trends live on the web so I can check what’s happening any time from anywhere on a phone app (really cool).

 

Arduino Temp Humid Monitoring Unit connected through wifi

Problem

Adding the trend I found I hit 100% humidity and even adjusting the set point I could not get it down. I was also getting a significant discrepancy between the Arduino temperature probe (even after upgrading to a DH22 probe) and the controllers probe. I did not have the Arduino probe in an enclosure so may have had dewing on the probe itself adding to this problem.

This led me to digging out dew point charts and I found that for the humidity we are running at and the temperatures we control to, we are typically within a few degrees of dew point. The problem I was experiencing was with the dip in temperatures in the early morning along with my humidifier over shooting the set point due to controller lag I was dropping below my dew point and had no control over this as I am only controlling on cooling. I also did not have a fan in the fridge at this point. Fridge was off however temperature was steadily dropping. Here is a trend of that data:

Solution

Firstly I added a fan to the same power circuit as the humidifier so this will run with the humidifier. I also turned down the humidifier to very low. This made little difference but was necessary for bad mold control anyway (airflow).

Secondly I realised I needed to add a heat source into the fridge to consistently pull the temperature above the set point. This way my controller was in charge and not the ambient temperature. I googled around for ideas on this and the easiest inexpensive solution I found was an incandescent light bulb in a tin (you don’t want too much light in the fridge – so I believe; although I am still trying to verify that). For information, a tin with a 100W light bulb, tin foil top and plastic lid will get so hot that within 5 minutes the lamp fitting and lid will have melted. So, back to the drawing board.

After some more fiddling, I have settled on a 60Watt light bulb mounted on a block of wood, wrapped in tin foil and connected to a dimmer unit to turn this down as low as possible. This has solved my problem and I am now controlling perfectly with great correlation between the two units. Less heat would be preferred but this works (maybe a reptile heating pad is the next option)

 

Foil wrapped light heater (foil top helps to dissipate heat)

Resulting control:

 

6 hour trend

Dip when adding heater on 24 hour trend

A further change I then made to increase the amount of air circulating inside the chamber was to connect the fan to the controller timer and run on a 15 minute on, 15 minute off repeating cycle. I also dropped the temperature and humidity set points. This had the following trend effect:

Temp Set Point Drop and Fan Change

Below is a longer term trend of the chamber with a few tweaks to settings. Note, the spikes can be ignored. These are when opening the door and making other mods to hardware (I also put the Arduino DH22 sensor in a pill box with lots of holes drilled in it to prevent the risk of dewing on the electronics – I figure this is like parking your car under a tree to prevent dew on your windscreen in the morning).

1 Week Trend