Incubator build advice needed.

[Pommie]

I'm currently building an egg incubator for my daughter or rather, her chickens! My current setup is two 92mm fans with 6 x 6R8 20W resistors as heating elements. This is all powered by a 300W 12V SMPS. This is my current prototype,

The various parts are switched using,
The fan speed by CCP1.
The servo for the egg rolling by CCP2. (not shown)
The heating elements by a 1mS interrupt that turns them on for between 0 and 100mS every 100mS (0-100%). To switch these resistors I used a IRL540 (100V 28A MOSFET). My thinking was that with a slow switching speed and only 10A it wouldn't need a big heatsink. I was wrong, at 90% duty cycle it very quickly gets to near 200C (None contact thermometer so maybe inaccurate). The MOSFET can be seen on the vero board in the middle of the fans.

So, my questions, how would you solve the MOSFET problem? I'm considering simply using one for each bank of resistors therefore reducing the current to 5A. This would mean that both gates will be driven by 1 pic pin!! Will this be a problem? Do I need to buffer this pin? Note that to remake the veroboard circuit is easy to do but I don't really want to redo the PCB.

My second question needs a little more ingenuity. The other thing I have to control is humidity. It needs to be around 40% for 18 days and then 65-70% for the remaining days. I don't see the 40% being a problem as ventilation should keep it low (ventilation is required due to CO2 buildup). To get to 70% I was considering using another 20W resistor with a wick in a water bath. However I can't seem to find a suitable way to achieve this. Anyone any suggestions?

Anyway, sorry for such a long post and thanks for any suggestions.

Mike.

 

[ClydeCrashKop]

Maybe put the MOSFETs in front of a fan and consider then heating elements?

 

[bryan1]

G'day Mike,
Well if the fet is getting well over 100C how about a small water tube where drops are introduced along with some some airflow. With the tube being part of the heatsink the water will flash and provide a crude form of humidity control.

 

[alec_t]

There looks to be a lot of wire (hence inductance) between the pic and the FET gate. That's a recipe for oscillation, which would indeed cause heating. Do you have the customary 100 Ohm (or thereabouts) resistor on the gate pin, very close to the gate, to damp oscillation (and also limit current which the pic is trying to drive through the gate)?

 

[dr pepper]

If your using 5v to drive the gates of the fets that might not be enough, with irf540's you might not be turning on the fets fully, even at 10a they shouldnt be that hot, another thing that can cause overheat is switching times, a fet gate driver would improve things.
You could switch 12v to the fet gates using 5v from the pic and level shifting transistors, or you could do it winding some gate drive transformers.

Nice project by the way.

 

[Pommie]

Looks like the simplest solution is to use two MOSFETs, I just disconnected one bank of resistors and it all runs cool now (40C). Due to the fact that I'm only switching at 10Hz I didn't worry about switching times or oscillations and my scope doesn't detect any. For anyone interested, at 100% the resistors get to 160C even though they are in the fan air stream.

As for humidity, I think I'm going with a model airplane tank that sticks out the front so it can be refilled and has a wick (piece of shamwow) inside that's over a resistor. I'll let you know how that works out.

Mike.