they do work and they are a lot cheaper but if you don't know what you are doing you can also make mistakes which would not be fun to deal with. Saying that I believe all incubators under $100 are complete rubbish so it could be worse.
You have to try to obey a few principles. Ventilation is important but not too much. Insulating helps to lower the heating requirement, a polystyrene shipping box is good.
The thermostat should be very sensitive. I use reptile thermostats I buy on
ebay. The cheapest chinese ones I can find to be honest but I find them to be excellent, been using them for years.
Heating is generally done with a heat bulb. I use a 25watt one in a 1ftx1ftx1ft insulated box.
I can highly recommend a govee thermometer which gives accurate readings to check temp and humidity.
You can go still air or forced air but forced air is generally more reliable. I prefer forced air so I will detail that.
The most important is the layout of your incubator. Adding a cup of water can alter the air flow for example and raise or lower temperatures. A higher humidity can also cause higher evaporative cooling on the eggs thus lowering their internal temperature slightly. People often mention they had poor hatch rates with high humidity with chicks drowning. I personally can't see how it is possible for chicks to drown in a pool of embryonic fluid when they break the internal membrane - they would instantly have the whole air cell space to breathe so there would have to be a huge amount of liquid for the chicks to "drown". To me I know sensors are not the most reliable and most people don't take into account that they should constantly be checking their temperature, a higher humidity or an added obstruction, even a little draft in the room can affect temperature in the incubator thus lowering hatch rates. It is well known that temperatures are the most important thing to maintain steady. If humidity goes too high and affects temperature stability or even lowers temperature by just a fraction, it might lead to underdeveloped chicks that are then too weak to break out of their membranes and it is understandable for a lot to blindly state they "drowned" but maybe they didn't develop properly and thus were too weak and died instead without ever having even taking a breath of air or embryonic liquid. Unless a proper autopsy is performed the proper cause of death can only be assumed yet so many times I have read it as absolute fact on why things went wrong when it doesn't physically make sense to me. Granted I have never had what I would call a drowned chick or maybe I have. If it never makes it then it never made it whether it drowned or not is kind of irrelevant to me at that point but I find it interesting that humidity is usually the main culprit blamed.
Anyway I am rambling sorry about that, so back to the design. I find a round shape is best. This deosn't work with the insulated boxes too well as they are square, but you can just place bowls in there to create a round shape that is insulated.
If you go for a round shape distributing the air is then super easy. All you do is attach the fan to the top so it is raised off the surface about 1cm and blows the air out onto the lid. Like this the air runs evenly all around the incubator, travels along the floor and gets sucked back into the fan above. You want to place the heat source just in front of the fan so the fan is constantly spreading the heat. I use usb powered router fans, no wiring to deal with.
If you stick to the square shape which is possible - it just adds more uneven temperature spots then I find enclosing the heat source in a separate insulated box works well, with my 25 watt heat lamp for example it would raise temperatures near the lamp too much. But with it being insulated from the incubator and a fan blowing across it at all times the heat is only spread via the fan. What I am saying might not make too much sense now but when you have one side of the incubator runnnig a degree hotter than the other side it can become problematic. It might take a few tries to get things right and you always have to let the sensors adjust for an hour first before getting a reliable reading.
One problem with a home built design is that it usually does not include an egg turner which is quite a useful feature.
