IRL540 Getting Hot DC Boost circuit

Other Hardware General Electronics
1

Not sure where to post this... this is more of a hardware question. Please keep in mind that I am hobbyist so I will try to provide information to the best of my knowledge. This also means that I am very limited with equipment to trouble shoot.

I am working on a DC boost circuit and my MOSFET is getting pretty hot even with a heat sink.
The heat sink is roughly 50mm x 25mm x 15mm

The L1 and D1 do not get hot. The LED gets very bright and has 35VDC going across the LED with resistor.
the LED is rated 32-35VDC ~700mA
D1 is a rectifier diode rated for 10A
R1 is now 22K
I had a friend measure L1 inductance with an LCR and it was ~25mH

12v supply voltage from a battery with 10A fuse.

Here is my last code. I thought changing the frequency might make a difference but no luck
I have played around with the duty cycle too a little too. Please let me know if there is something wrong with my circuit or code.
thanks

int pwm = 11;

void setup(){
  TCCR2B = TCCR2B & 0b11111000 | 0x06;
  pinMode(pwm,OUTPUT);
}
void loop(){
  analogWrite(pwm,128);

}

boost.png

2

Another boot strap circuit.

boot3.jpg

3

You only need to turn the MOSFET on until the current in the inductor has risen to the value that magnetically saturates the inductor core. Any longer than that and you are wasting energy/loosing efficiency.

It is unlikely that a 50% duty cycle is optimum.

Have a read of this tutorial.

4

thanks JohnLincoln, I will try to play around with the duty cycle.
I will probably add a voltage divider for feedback.

I will post my results when I get to try it again.

5

Did you mean 'mH' as 'milliHenries' or 'microhenries'? I am assuming you meant millihenries, but I have seen it used the other way. If it is 25 microhenries, the frequency is way too low for the inductor, and the MOSFET and inductor will get really hot.

6

Normally you'd have some feedback to automatically adjust the duty cycle to keep the voltage or current constant.

With a high-power LED, you'd normally be monitoring/regulating the current. That means you wouldn't have R2, or R2 would be a much smaller value with just-enough voltage drop to allow you to monitor the current.

With standard low-power LEDs, we usually just use a current-limiting resistor and a constant voltage supply. But, this is inefficient so with high-power LEDs it's standard practice to use a switching constant current (or controlled current) supply. Since you are using an inductor in a switching design anyway, you should go with a constant-current design.

I assume you are powering that thing from 5V? How much power/current does the 5V power supply provide? With a voltage-boost circuit, you feed more current in than you get out. Power = Voltage x Current, and you can't get more power-out than you feed-in.

7

If you inductor has a core you must use a comparator to detect over-current, because
if the core saturates the current will shoot up to 100's of amps very fast indeed.

Basically once you have enough current flowing in the core to saturate it, it then behaves
as if air-cored and the inductance drops to perhaps 1/1000th of its normal value, leading
to catastrophic rises in current in few microseconds, this can make the MOSFET explode
(wear eye protection if working on high power circuitry like this).

All DC-DC converters using cored inductors have protection circuitry and over-current
shutdown because its a necessity, not a luxury.

The MOSFET you are using has quite a high on-resistance - there are much betters one's
available.

8

It's a really late here and my eyes are starting to cross so I apologizes for any mis spelling or wording. Here is some more information about the circuit.

CWashburn - The inductor is a toroid with an iron core and I'm pretty sure it's millihenries The inductor does not generate any noticeable heat

DVDdoug - I now have a 12VDC 19A supply connected to the boost circuit I was using an old lead acid battery and a fused 5VDC 500mA USB connection to the Arduino

MarkT - is the on resistance different than the RDS(on). I chose the Vishay IRL540 because the RDS(on) was 0.077 with VGS = 5.0V and the VGS(th) is 1v to 2v. If you know of a better MOSFET please let me know.

I will try to make some changes and post back. thanks for all the info it has been very helpful.

Thanks for the feedback

9

Thanks everyone for you help. The issue was with my inductor 10mH as too low... I ended up changing the inductor to 9uH and it works great. I will try to upload my current schematic sometime later.