Building a chicken incubator! Moderators: adafruit_support_bill, adafruit

[robodude666]

Introduction

Hey guys,

I'm building a chicken egg incubator! Long story short I moved from NYC to Idaho for a gal. Her father is a farmer, and is in need of a large chicken incubator. The size of incubator he's interested in sell usually for $900 to $2,000 new and are way out of his budget. Seeing as I love electronics, I offered to help build one. Because minds work better in groups, I'm looking for some feedback on my ideas/design.

Overview

For those of you who are not familiar with the requirements of incubating a chicken egg, here are the basics:

- Keep temperature between 99.5 to 100.0f (99.9f is best).
- Temperature fluctuation of +/- 0.5f is tolerable over short periods of time, however fluctuation of +/- 1.0f over long periods of time can be fatal.
- Humidity should be kept at 58 - 60%. During the last two to three days, while hatching, humidity should be increased to 65%+.
- Eggs must be "turned" roughly +/- 30 degrees five times a day, except for the last two to three days during hatching.

(The exact temperature/humidity required is also dependent on the breed, however the above is generally true for all chicken eggs.)

The incubator needs to be designed to hold ~200 to 300 eggs, and because of this it will be a multi-level incubator, much like this one:



Each shelf will hold ~90 - 108 (maybe less or more, not sure yet) eggs. The multi-level design is required to be a forced-air incubator, to make sure air is moving around and the temperature is relatively even on all levels so fans will be incorporated into the design.

That's a general overview of the project. Onto the specific:

Temperature

From my research, many one-level incubators simply use one or two 60 - 75w light bulbs, that are connected to a commercial temperature-controlled relay. This type of heating system will not work very efficiently for a multi-level incubator as the area is larger. The light bulbs also take a period of time to heat up/cool down, which will affect the ability to adjust temperature. Because of this, I'm considering using a couple 40 - 100w heating element/coil. There will be one near the top and bottom, or one near each level. In addition, there will also be a fan blowing air past the heating element to distribute the heat.

Humidity

Most large commercial incubators I've seen have a water tray on the top/bottom of the incubator. Humidity is dependent on the surface area of the exposed water from my understanding. I'm currently considering having a water container (like a hamster's water bottle) hanging outside of the incubator with an electronically-controlled valve on the inside that feeds water into a tray to increase humidity. Decreasing humidity can be accomplished (I hope) but simply increasing the fan's speed and moving the humid air out.

Electronics

The entire system will be controlled by a LeafLabs Maple Mini because I love LeafLabs and the STM32 (plus it's size and cost works in the project's favor)... and I happen to have a few.

For temperature/humidity monitoring, I will be using the DHT22 temperature/humidity sensor, as for the cost it provides both sensors with relatively good accuracy. I plan to have 3 or 4 scattered around the incubator, probably one on each level.

I'll be including a DS1307 RTC, to keep track of what day in the cycle (to stop turning and increase humidity 2-3 days prior to the end of the 21-day cycle). Because I'm primarily care about the day, I'm not to concerned about the accuracy. I may even go with a software+EEPROM-based RTC to save some money.

And of course an LCD to view data on. I may or may not include an SD/SPI-Flash logging capability. That's dependent on how things go.

As far as temperature/humidity control goes, I'll be implementing a PID controller. Each heating element will be turned on/off dependent on the temperature sensors. If it becomes too hot, the fans will help remove the warmer air. Humidity control will probably not run on a PID controller, or may with a larger window as I don't want to A) charge humidity rapidly and B) use up all the water in the container.

Questions For You

Any suggestions regarding the temperature/humidity control? Thoughts on using heat elements vs light bulbs? Anyone know a good source for heating elements by the way (mcmaster's are too expensive and not the right kind)? Are the ones out of old coffee machines or water heaters too powerful (they're like 500 - 1500w)? I'm assuming weaker 40w units can simply be controlled with an SSR and a 12v supply or something along that line, correct? Or are they all 125 VAC? Anyone have a datasheet on one?

Any help would be appreciated, thanks!

-robodude666

[franklin97355]

Interesting project. The picture you included looks good and I would use it as a start. The mechanicals are straight forward and your choice of sn=ensors seem reasonable.

[jigsawnz]

I build a small scale incubator that I try to improve in stages. It can hold around 50 eggs but you have to rotate them manually. The one thing I found out is that the lightbulbs were a really bad heat source in my small scale project. The bulbs gave a big temperature blast on the top of the eggs. The Sensor would read a sudden increase and turn it off again and so on. For my next run I got a 20 Watt heat plate for a more steady temperature. The DHT22 sensor measures temperature and humidity and adafruit provides a library for arduino.

https://picasaweb.google.com/lh/photo/g ... directlink

Here is a photo of my old setup.
Arduino, MicroSD shield, Chronodot RTC, VFD display, DHT22, Powerswitch Tail

[robodude666]

jigsawnz,

Where did you get the 20W plate? Where inside is it installed? How quickly does it heat up?

-robodude666

[franklin97355]

How quickly does it heat up?

You don't need it to heat quickly, in fact that would be counterproductive. It needs to have enough power to maintain the desired temperature within 0.5 degrees but if it acts too fast you will find the temperature cycling like in jigsawnz case.

[jigsawnz]

I bought the heating plate on ebay. Haven't gotten around trying it out yet. When I install it, it will be in the bottom. Warm air rises up. A fan will blow it down again to have some circulation going.
I need two parts to redo the project properly but there out of stock. It's not easy sourcing elektronics in New Zealand.

[lyndon]

When you say "X is too expensive or doesn't work for my application" it's good to be specific so we can help.
How much do you expect to spend on a heater? What size do you want? I looked at the cartridge heaters that McMaster-Carr carries and they look fine.

FWIW, I have never built an incubator, but I have raised chickens from day-old chicks many times (have a batch at 8 weeks old right now...) and heat lamps/lightbulbs work fine for that purpose. I don't see why they wouldn't work for an incubator, it boils down to good air circulation and proper placement of the sensor.

[robodude666]

@franklin97355 - True. You don't want it to heat up quickly. Temperature changes should be made over a 4 - 6 hour period.

@jigsawnz - Do you have the auction listing? I'm interested in knowing what to look for.

@lyndon - Sorry. The ones I saw on McMaster were $20 - 80 a piece. I will be needing three or four of them, and I may be making more than one of these units. I am looking at the $5 to $10 range for the heating elements. The McMaster $17 Cartridge Heaters looks interesting, but it's a little expensive. Is it meant to be inserted into a metal coil that distributes heat around an area?

The type of element I'm thinking of is this:

http://www.fullersupplyco.com/incubator-heating-element

And it's within the price range I was thinking about. It's sold by a company I'm not familiar with, with no datasheet or any information on it... Looking to maybe finding something that has a stat or datasheet.

Heating lamps are good for keeping baby chicks warm. They work for small incubators, but become less efficient for larger incubators.

-robodude666

[jigsawnz]

http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... OC:NL:3160

This is the Heat source I will use next. Now I have it at home it still seems overpriced for the quality you get. I will see if I can turn it on someday soon. The most important thing might not be the heat source but how well you insulate the incubator. Insulation is a huge factor in keeping the temperature constant.

[adafruit_support_mike]

The basic differential equations for heat transfer provide two easy suggestions for making a system stable: insulate it well and make sure it has plenty of thermal mass. The insulation part is easy and obvious.. the harder it is for heat to leave the enclosure, the slower it will change and the more easily you can adjust it.

Thermal mass isn't quite as obvious, but it is easy: fill a few 2L bottles with saturated salt brine, glue the lids on with epoxy, and shove them in the enclosure along the air path from the circulating fans. It takes a lot of hot/cold air to change a gallon of brine's temperature significantly, so it will resist sudden changes in temperature. You can do the same for humidity with a large wad of something absorbent like wood shavings, cloth, or paper.

You might look at moving the heat/humidity treatment out of the enclosure to separate chambers. Make a 'hot box' that stays a few degrees warmer than the target temperature, and a 'wet box' that's more humid than the incubator needs to be. You can build controls to keep both of those boxes stable, then adjust conditions in the incubator by pumping in an appropriate amount of heat energy from the hot box and an appropriate amount of water vapor from the wet box. Between adjustments, the two boxes can work to bring new air up to the correct temperature/humidity. Since the boxes aren't connected to the incubator per se, you get the luxury of being able to use coarser adjustments and larger value swings.

[adafruit_support_rick]

I'm still confused from the original post - did you move to Idaho for a gal, or for a chick?

[lyndon]

Yes, cartridge heaters are designed to be embedded into a substrate.

Didn't occur to me before, but a bunch of chassis-mount power resistors should work: http://www.digikey.com/product-detail/en/AHA10AJB-5R are $1.95 each and rated at 10W to temperatures well above cooked chicken! These are 1% resistors, so if you shop around you should find cheaper 5% or 10% ones available.

Drive them with a cheap PC power supply and you're all set. These particular resistors are rated to about 150degC at full power rating. You only need to get up to about 38C, so you're OK in that regard. So if you took 4 of them in parallel, you can dissipate 40W. Normally I'd derate a component, but it's only a heater operating well within its specs, so let's run it at full power. To get 10W out of each resistor using a 5V PC power supply means that each resistor has a value of 2.5ohms. 40W from a power supply rated to provide 100W at the 5V output provides plenty of safety margin for the power supply.

[robodude666]

Great idea lyndon!


Would a fan blowing across the power resistor be able to distribute the heat around the incubator sufficiently?

Looking at the datasheet, the resistor outputs a certain amount of heat based on the % load. What would be an effective way of controlling the load digitally, so I can adjust the load till I get my required temperature? Would a MOSFET be sufficient, or would I need an SCR?

Also, would I really need 40 W to get my ~38 C? Wouldn't a single 10 W be sufficient based on the curve in the datasheet (page 2, temperature rise)?

-robodude666

[lyndon]

A fan would probably be needed for good temperature distribution. A FET or even a relay will be fine to control it: it's not that much power. I'm willing to bet that most low-end commercial incubators are using bimetallic strip thermostats that are far less accurate than what you're doing.

As far as how much power you need, that's beyond my area of expertise. It depends on the room ambient temperature, the volume of the space to be heated, the amount of insulation used, etc.
I'd start with E= m*c*deltaT and go from there. Look at the third post here http://www.physicsforums.com/showthread.php?t=216184 for a simplified formula.

[john444]

Hi Robodude666,

Would a fan blowing across the power resistor be able to distribute the heat around the incubator sufficiently?

It seems to me that Lyndon is on the right track with a bank of resistors (to prevent a localized hot spot), a fan to distribute the heat and a FET to pulse-width-modulate (PWM) the power sent to the resistors.

Wouldn't a single 10 W be sufficient based on the curve in the datasheet (page 2, temperature rise)?

You are a little off in your understanding of the datasheet. It is referring to the power dissipation and temperature rise of the resistor itself - not the air or objects around it. Consider lots of low power resistors that add up to the maximum power you need. For example ½-W resistors cost about $5.00 per 100. That would give you a 50-W (maximum power dissipation) and a large surface area (so the surface temperature is not high). You can just as easily have one power resistor mounted on a large heat-sink to spread the heat. Or some number of resistors in between. Net result is the same.

Lyndon expressed a little differently but, the temperature inside the cabinet is related to the power dissipated by the resistors inside the cabinet as well as the temperature outside the cabinet in a complex and time dependent relationship. It may be simpler and more accurate to build the cabinet and try a convenient value of power dissipation in the resistor bank and see what happens. If it heats up to quickly, add mass. If it uses to much power, insulate. You should be able to focus on what works well pretty quickly.
My experience has been the heat / temp control is the easy part. Building the cabinet without dead-air pockets and mechanically rocking the shelves will take more time than the heater & temp controller.

Good luck and let us know how it goes.
The lurkers want to know.
John