They AZ1117 is getting to much heat, why?

Hi falks,

I have made a circuit. I do not know wheather it is a problem or not, but the AZ1117 3.3V ist getting to much heat. I even can not touch it beacause of its heat.

My question: is this normal ? If not: how i can coll it down ?

I thought, that stepping down from 12V to 3.3V is to much for it.

This depends on the load, but with a voltage drop of nearly 9 volts, the current must be quite low for this to work.
Your schematic shows a 5 volt step down (buck converter?) so why not use the output of that to feed the 3.3 volt regulator instead of driving it directly from 12 volts ?

It's shown as an L7805, which is the low-power version of the 7805 in TO92 package. Given the fact that the 1117 also overheats, connecting the 1117 to the 7805 will likely move the overheating problem to the 7805.

@absoltue_beginner get yourself a DC-DC step down ('buck') converter. They handle higher current levels much better than the linear regulators you're using now.

HI @absoltue_beginner

What happens is a question of power dissipated in the regulator AZ1117.
The input has ~= 11.3 V (12 V - V over diode D1), and the output has 3.3 V. So 8V "stays" over the regulator.
If the circuit after the regulator consumes, for example, 100 mA,
the dissipated power on the regulator is set at 0.8W, if it is 200 mA the dissipated power will be 1.6W and so on ( W = V * I) and in your case V = 8V.
The higher W the greater the heat on the regulator.

RV mineirin

Hi, @absoltue_beginner
What is your 3V3 supply providing current for?
What is the regulators load?

Thanks.. Tom... :smiley: :+1: :coffee: :australia:

Is this possible ?

By the way, the load is 900mA

The current is 900mA 12V DC.

Maybe I should connect the 3.3V regulator, directly behind the 5V regulator ?

That is too much. The circuit would have to dissipate over 7watts as heat.
I made a mistake in assuming your 5v step down device was a buck converter. This is what you should use.

Load 900mA, volt drop across the 1117 is 12 - 3.3 = 8.7V

P= V * I = 8.7 * 0.9 = 7.83W !!!!! Heat Dissipated in the 1117.

What is the rated power disipation of the 1117?

Thanks.. Tom... :smiley: :+1: :coffee: :australia:

Hey guys,

first of all: Thanks for all your answers.

I will provide u later the full project details. So please wait for me before u answer further.

Right bow, I am on the way home through the highway :slight_smile:
As soon as I arrive home, I'm gonna post u the full details of my project.

1 Like

Probably 500-750mW absolute maximum, with a more realistic upper limit of around 350mW :wink:

1 Like

Hi, @absoltue_beginner
What are you supplying 3V3 too?
What is the load?

Thanks.. Tom.. :smiley: :+1: :coffee: :australia:

800mW is the absolute maximum at 25°C if it's in a SOT223 package.

The formula for calculating the junction temperature at a given power dissipation is:

Tj = Pd * Rja + Ta

Where Tj is the junction temperature
Pd is the power dissipation
Rja is the thermal resistance of junction to ambient.
Ta is the ambient temperature

We can rearrange that formula to express the power dissipation:

Tj = Pd * Rja + Ta <=> (Tj - Ta) / Rja = Pd

Putting in the values:

Thermal resistance junction to ambient (Rja) is 125°C/W
Maximum operation junction temperature is 125°C
Ambient temperature (Ta) is 25°C

(125 - 25) / 125 = 0,8

Obviously this does not include any safety margin.

I used the values from the datasheet found here:

Thanks, man, I will read it, but birst, here is my project in detail (see Picture).

What kind of 3.3V microcontroller are we talking about here and what other peripherals are attached to it, if any?

Obviously, the 0.9-1A current draw from the motor doesn't overheat the 1117 since they're not connected to each other. This also means that putting the 7805 in-between the 12V input and the 1117 isn't going to help, and will in fact be counterproductive. But I already pointed that out earlier. Your schematic only gives more reasons not to think about that option.

Anyway, if the 1117 currently overheats, this is because the microcontroller draws more current than the 1117's power rating allows. From the 1117's viewpoint, it would help to connect it to the 7805 instead of directly to +12, BUT as I pointed out above you're already stressing that poor little 78L05 way beyond what is reasonable.

Long story short, you're expecting too much from the voltage regulators you're using here. Get some DC-DC step down converters. They're cheap.

It is a ESP8266 E-12 Controller.

Ah yes, especially if you also use WiFi, it will draw quite a bit of current. But even with no WiFi enabled, the datasheet of the ESP8266 suggests that the average current draw of the processor is around 80mA. WiFi adds up to 170mA on top of this.

Going with 80mA for the moment, that would mean (12 - 3.3)V * 0.080A = 0.696W of dissipation in the 1117. As @nicolajna showed you, this is quite close to the theoretical upper limit. So it's logical you witness overheating of the 1117 in your setup.

How's the search for the DC-DC buck converters going?

Can you show me one on eBay or amazon ?

okay, thanks electronical friend !! :smiley: