Hi,

Build: Arduino controlled water heating tank using a 48V 12A heating element. Arduino is there to deactivate element once set temperature is reached, and to check there is water in the tank so as not to burn out element.

So far I have been using chatgpt to try figure out what approach to take.
Parts in mind for next test (still need to order):
IRLB4110 Mosfet N channel, 180 A max current, 100V, supposed to take 10V to saturate it fully.
Gate driver: UCC27531

Here are some notes I had from a previous test, this was before I decided to use a 48V 12A element, and was working with a 20V 5A element, which was far too slow at heating for my intended purpose.

**This week I got the system running using a MOSFET for the first time. STP16NF06 60V 16A TO220. It worked, using 2 push buttons I was able to get the arduino to switch the 20v 5A heating elements on and off depending on what I set the target temperature to, which was displayed on the GC9A01 monitor, alongside the ‘live’ temperature as read by the DS18B20 temperature sensor.

The mosfet got really hot. Chat gpt thinks this might be because the 5v from the arduino’s digital pin isn’t able to fully saturate the mosfet. After doing some digging it seems it would be better to use a different mosfet, like the IRF3205ZS, with a gate driver like the TC4420. **

Since making that note I have decided to switch to 48V 12A. So the IRF3205ZS mosfet, capable of 55V, then seemed too low a safety threshold, so now I am thinking of the IRLB4110. I know it might be overkill, and use more power in the circuit. I am more interested in making a robust, durable product than making it as power efficient as possible. But please let me know if you have suggestions otherwise, including heat sink sizing, other components I am not aware I'm missing.

I am new to using mosfets and after reading a few forums here thought it would be a good idea to make a post.

 

if your board is driving mosfet directly, you need mosfet that can be driven by the logic-level (example RFP30N06LE) or you need to use suitable driver circuit.

 

There is no particular compromise with using the IRLB4110 FET, and the TC4420 Gate-Driver,
they will work just fine.

As soon as You remove the FET from it's protective packaging,
in a static-free-environment,
permanently solder a ~10K-Resistor from the Gate-Pin to the Source-Pin,
this will insure that the Gate is not accidentally destroyed by Static-Electricity during handling.
This will also insure that the Gate will be forced to Ground when the Power is removed from the Circuit.

You will need to provide a ~15-Volt Supply for the Gate-Driver.
Supplying the FET's Gate with a solid ~15-Volts will ensure a very-low Rds/On Resistance,
( less than ~0.005-Ohms ), this will minimize Heat-Dissipation from the FET.

I would recommend a very simple High-Voltage-Linear-Regulator like the LM317HV from Mouser,
( DigiKey-Electronics is out of stock for some reason ).

Note, this Regulator requires a ~5mA Minimum-Load,
so connect a "Power-Good LED" to the Regulator's Output, along with the Gate-Driver.
Because the FET's Gate doesn't need any Current except during switching,
and the Quiescent-Current required by
the Gate-Driver-Chip may not be enough Minimum-Current to keep the Regulator stable.

A small Cast-Aluminum-Box should provide adequate Heat-Sinking for all three TO-220-Packages.
( only the FET will dissipate any noticeable amount of Heat ).

If it's any particular advantage to You,
You could easily use higher-frequency PWM control with this setup.
( actually it's already ""PWM"" controlled, it's just PWM at an extremely low frequency )
.
.
.

 

An IRLZ44n should be fine and could be driven directly by a 5V Arduino for on / off low frequency switching.

 

if your board is driving mosfet directly, you need mosfet that can be driven by the logic-level (example RFP30N06LE) or you need to use suitable driver circuit.

Thank you, I'm planning to use a gate driver as I had a problem without one last time.

 

An IRLZ44n should be fine and could be driven directly by a 5V Arduino for on / off low frequency switching.

Thanks for suggesting, I want to use a mosfet capable of higher than 55V to be safe.

 

It can be quite tricky to find a suitable Logic-Level-FET that will reliably
work, and be free of over-heating problems at the same time.
.
.
.

 

There is no particular compromise with using the IRLB4110 FET, and the TC4420 Gate-Driver,
they will work just fine.

As soon as You remove the FET from it's protective packaging,
in a static-free-environment................

Super reply, thanks so much, I appreciate the reassurance and additional tips. Regarding the PWM, I guess you were getting at being able to throttle the heating element's power. I don't need to control it any more than on/off. I am overseas right now so won't get to try this for a few weeks yet, but will update here how it goes.

 

It can be quite tricky to find a suitable Logic-Level-FET that will reliably
work, and be free of over-heating problems at the same time.
.
.
.

Are you responding to Bob, as in, do you mean without using a gate driver?

 

There's no need for a "Logic-Level" FET,
and it's built-in compromises,
if You have access to a ~10 to 15-Volt Supply.

For very low-Power-Circuits,
Logic-Level-FETs are great for simplifying the project overall.

A Gate-Driver,
even though it may add a small amount of complexity and cost,
is always the superior way to "Switch" a FET-Gate.
It is specifically designed to minimize the "Switching-Time",
which is where a lot of the unwanted Heat and stress is generated.

There are a few applications where the FET is used as a "Variable-Resistor",
in which case,
a standard "On-Off" Gate-Driver, or a bare Micro-Controller-Output, will not work at all.
These are most often Linear-Voltage or Current-Regulator-Circuits.
.
.
.

 

There's no need for a "Logic-Level" FET,
and it's built-in compromises,
if You have access to a ~10 to 15-Volt Supply....

Yes, I'm powering the elements with 48V so as you suggested I can step that down to create the 15V required. Thanks again.

 

Thanks for suggesting, I want to use a mosfet capable of higher than 55V to be safe.

Sure, whatever you want. Perhaps your 48V power supply randomly jumps to 57V.

 

Sure, whatever you want. Perhaps your 48V power supply randomly jumps to 57V.

My understanding is that it's never a bad idea to overspec so you aren't pushing the component to its limit.

 

My understanding is that it's never a bad idea to overspec so you aren't pushing the component to its limit.

Much more important for current specs than max voltage rating. Operating a transistor at 48V when it is rated for 55 will not do it any harm.

 

One suggestion I have not seen is to monitor the voltage across your mosfet when it is in the conduction state. That is the sure way to verify that you are providing adequate gate drive. It also works for transistors but it is harder to interpret some times.
One thing missed is that with resistance heating it is much more efficient if the input power can be reduced by duty cycle reduction as the temperature setpoint is approached. Also, with electric heating especially,in addition to the solid state device used for duty cycle control, I always included an electromagnetic relay to assure positive shutoff of the electric heater power.