Hello,
I'm trying to use a arduino due to turn on a n-channel MOSFET (IRF520N) to switch on a voltage of 12V for a boost converter. the arduino due io is 3.3v.
I have attached the schema of my boost converter.
but when I run it something went wrong. I belief that the 2N7000 are burned.
how to correct the driver of the mosfet ?
Best regards.
First of all, add a pull-down resistor at the gate of Q1. When nothing is connected, or when the Arduino's output pin is in hi-z mode, a potential will build up at the gate (which is basically a capacitor) that will turn it on. This means that Q3 will be on as well!
Actually, I would get rid of Q1 and Q2 altogether, and use an optocoupler instead, because it's simpler, but your setup should work as well ...
And I would use a boost converter IC, why do you want to use an Arduino for this?
thank you PieterP.
I tried to use a optocoupler (4n53), but it didn't work !
"I would get rid of Q1 and Q2 altogether"
I need to fire the irf520N mosfet from arduino due (3.3V at frequency 10kHz)
"And I would use a boost converter IC"
I need arduino due to control the boost converter for some future tests.
do you have any suggestion to drive irf520N mosfet using 4n53 optocoupler or irf520N mosfet ?
Remove Q1, and place your optocoupler on the high side, like in this schematic:
What is the switching frequency.
The 1N4004 is not fast enough for a switching supply.
Use a schottky diode.
Gate drive might not be fast enough with resistors and optos.
The mosfet will get hot from the switching losses.
You have designed a car without a brake.
Supplies need secondary voltage feedback.
Leo..
Hmmmmm...
What voltage are you trying to generate? I see you have a 22 ohm load, which would take more than 1/2 amp even at 12v in with no switching at all. ( 0.54A ignoriig the the diode drop for now
As Leo says , the 14002 isnt quick enough. And given my sums below, perhaps totally inadequate in current rating.
And you really need a complementary drive for the output device. A typical gate capacitance is ~3nF, which gives a 6uS drive risetime which is slow, and could give bad switching losses at high drive frequencies . And there's no mechanism for controlling the output voltage.
And the capacitors look far too small - depends on you switching frequency.
THE BASICS :
If you're driving this with an arduino PWM output the frfequency would be about 1kHz.
With a 50% duty cycle the current at the end of the 'on' part would be given by :
. V = L di/dt..
The on time = (1/2)/1000 => 500uS
so di/dt is 12/330u => ~ 36,000 v/s => about 18 Amps
and the energy stored
E = 1/2 L I^2 => 53mJ. At 1kHz frequency this gives 53j/s or 53 watts.
With a 22 ohm load this would mean the max voltage you'd reach is about 25 volts.
With a 99% duty cycle you get
Max current 35 amps, max power 105 watts, max voltage 35 volts.
A 35 amp 330u inductor is a pretty juicy lump ! If you're not using a big enough inductor it would saturate in less than 500uS. When this happens trhe inductance drops to effectively zero, and the rate at which the current rises is limited only by the circuit resistances = > a LOT! of current.
You can buy excellent switchmode power supplies from many people - either as completed units or as clever ICs from many people... unless your requirements are really weird I'd go that way.
I built a capacitor discharge ignition unit once, and got about 250V out of such a circuit - that used a 555 timer as drive to the switching FET and a feedback circuit to control the output voltage - that IS pretty weird!
What's your application.. ?
Look at LT's excellent application notes for a much more detailed analysis of switchmode power circuits.
regards
Allan.
Wawa:
What is the switching frequency.
Hello Wawa,
the switching frequency is 10kHz.
Wawa:
Gate drive might not be fast enough with resistors and optos.
but if i removed the resistors and optos, how to drive the mosfet and limit the gate current ?
PieterP:
Remove Q1, and place your optocoupler on the high side, like in this schematic:
Thank you PieterP,
i will test it.
allanhurst:
What voltage are you trying to generate?
Hello Allan,
and thank you for the detailed explanation.
I'm trying to generate voltage under 60V with max current of 4A. the PWM fequency is 10kHz.
allanhurst:
I see you have a 22 ohm load, which would take more than 1/2 amp even at 12v in with no switching at all.
you are correct. any suggestion how to fix it?
allanhurst:
A typical gate capacitance is ~3nF, which gives a 6uS drive risetime which is slow,
you are correct again, the risetime is slow. any suggestion how to make the switching driver better?
i need to drive the mosfet from arduino due (3.3V).
allanhurst:
THE BASICS :
thank you. i really needed that. the PWM fequency is 10kHz, I will make the new calculation.
allanhurst:
You can buy excellent switchmode power supplies... unless your requirements are really weird.
you are right the requirements are weird. i need this circuit for a future upgrade, where i will add another mosfet in parallel with the previous one. the two mosfet's will be connected together.
allanhurst:
I built a capacitor discharge ignition unit once, and got about 250V out of such a circuit ... that IS pretty weird!
well, it's electrical engineering, where the weird stuff are made and found !
allanhurst:
And you really need a complementary drive for the output device. A typical gate capacitance is ~3nF, which gives a 6uS drive risetime which is slow, and could give bad switching losses at high drive frequencies .
Consider what Allan says first: he's right. The gate acts as a capacitor, meaning that it takes some time to charge or discharge it with a resistor (RC time).
While this is happening, the MOSFET is not fully open, but has a higher drain-source resistance, which causes it to get very hot.
That's why a complementary or push-pull driver is a better idea than a passive one (using a pull-down resistor instead of actively driving down the gate).
A complementary driver is a complementary transistor pair (NPN and PNP): when the input is low, the PNP conducts, driving the gate down, when the input is highn the NPN conducts, it drives the gate high. Resistance between the gate and the power lines can be much lower, resulting in higher currents, so faster (dis)charge times.
bilux:
I'm trying to generate voltage under 60V with max current of 4A.
Sounds like you need a supply for a car booster.
+30volt and -30volt.
This is an example of how complex 60volt/4Amp looks like.
Leo..
PieterP:
Consider what Allan says first: he's right. The gate acts as a capacitor, meaning that it takes some time to charge or discharge it with a resistor (RC time).
While this is happening, the MOSFET is not fully open, but has a higher drain-source resistance, which causes it to get very hot.That's why a complementary or push-pull driver is a better idea than a passive one (using a pull-down resistor instead of actively driving down the gate).
A complementary driver is a complementary transistor pair (NPN and PNP): when the input is low, the PNP conducts, driving the gate down, when the input is highn the NPN conducts, it drives the gate high. Resistance between the gate and the power lines can be much lower, resulting in higher currents, so faster (dis)charge times.
Thank you Pieter,
the risetime is somewhere between 6uS ~ 9uS, even when i used the optocoupler.
I will try to use a complementary or push-pull driver (hope it work).
Wawa:
Sounds like you need a supply for a car booster.
+30volt and -30volt.
This is an example of how complex 60volt/4Amp looks like.
Leo..
Hello Leo.
It looks like the DC-DC converter i'm trying to make (from 12V to 60V maximum with 4Amp). but i don't need to connected to a motor or a car, i just need to demonstrate it work.
i'm studying the faults in DC-DC converter.
in a future upgrade, i will add another mosfet in parallel with the previous one. the two mosfet's will be connected together.
and you are right, it is complex. but i need to make it.
thank you.
Hi Bilux...
Suggest you start off with a much smaller inductor at a much higher frequency - some 10's of kHz and maybe 10-50uH - use the sums above such that your wanted peak power and voltage occurs at less than 80% on time of the switch . You could also use a variable frequency, or a mixture of PWM and frequency.
And check for extremes - what if the load is switched off?
You can buy gate drivers for the MOSFETs.
Look at TI, ST, LT etc etc for excellent chips which do pretty much everything, and carefully read all the appplication notes.
Switchmode PSU's aren't easy - you have be careful with components ( eg low ESR for capacitors) and switching transients trying to blow things up (=> snubber components) . I reckon you'll be releasing a fair bit of magic smoke in this project............
good luck!
Allan.
bilux:
Hello,I'm trying to use a arduino due to turn on a n-channel MOSFET (IRF520N) to switch on a voltage of 12V for a boost converter. the arduino due io is 3.3v.
I have attached the schema of my boost converter.
Start again, use a MOSFET driver chip that can actually do the job. Your gate drive to the main
MOSFET is orders of magnitude too slow.
Also you are doomed to failure without over-current protection - unless you make your own large
air-cored inductor you will saturate your inductor and blow up the MOSFET. All SMPS designed use
a fast current sensing circuit to switch off the MOSFET before/if the inductor saturates, and
it is necessary - your inductor will transition from 330uH to 0.01 ohms on a timescale of microseconds
if it saturates (it is just a short piece of wire after all)
The IRF520 is a hopelessly dated device, pick something 'performant' instead, and choose a low
voltage device for a start, high voltage MOSFETs are much higher on-resistance.
For 10kHz switching I suggest going for < 250ns switching time, so nothing but a proper gate driver
chip makes any sense. There are thousands and thousands of gate driver chips all purpose designed
for driving MOSFET gates and many have useful protection circuitry and frankly all are 100
times better than a 2k passive pull-up gate drive.
I would also recommend reading a lot about power electronics before trying to design your own
switching converter, its definitely not trivial.
allanhurst:
Hi Bilux...Suggest you start off with a much smaller inductor at a much higher frequency - some 10's of kHz and maybe 10-50uH - use the sums above such that your wanted peak power and voltage occurs at less than 80% on time of the switch . You could also use a variable frequency, or a mixture of PWM and frequency.
Thank you Allan.
MarkT:
Start again, use a MOSFET driver chip that can actually do the job...
Hello MarkT,
and thank you,
MarkT:
The IRF520 is a hopelessly dated device, pick something 'performant' instead, and choose a low
voltage device for a start, high voltage MOSFETs are much higher on-resistance.
I'm new in practical DC/DC converters.
it would be helpful you have any suggestion of better MOSFETs ?
MarkT:
There are thousands and thousands of gate driver chips all purpose designed for driving MOSFET gates and many have useful protection circuitry
like ir2110 or IR2109?
Just for curiosity, I attach my pspice simulations for the capacitor discharge ignition.
Note this isn't exactly how I made it - and many of the components aren't what they say!
regards
Allan
Schematics_Samples.zip (7.54 KB)
allanhurst:
Just for curiosity, I attach my pspice simulations for the capacitor discharge ignition.Note this isn't exactly how I made it - and many of the components aren't what they say!
regards
Allan
Thank you Allan,
the simulations will be very helpful.
and thank you for your advices along the way.
bilux:
Hello MarkT,
and thank you,I'm new in practical DC/DC converters.
it would be helpful you have any suggestion of better MOSFETs ?
like ir2110 or IR2109?
Goto an electronics stockist and search - you will find thousands(*), you can't recommend anything
without some idea of the requirements.
(*) if you don't find 1000's you aren't on the website of an electronics stockist.
