MOSFET burns - 220V inverter for induction motor speed control using Arduino.

Thread Starter

Claudez

Joined Mar 24, 2019
28
Hello,
I have been working on my project for quite a time on controlling the speed of a 220V, 60Hz, 0.35 induction motor (domestic fan), I use Arduino to generate 60Hz with a switching frequency of 10KHz, when I try my circuit with a 20V motor, It works well and I can even vary the frequency.
but, once I connect to 220V the MOSFETs suddenly burn (explode) and the gate driver burns too. I came to understand that I don't have much knowledge of what is going on back there.

components used:
OPTO-ISOLATOR - 4N35
MOSFET -IRF830
GATE DRIVER - IR2113
BOOTSTRAP CAP (MLCC) - 47uF
BOOTSTRAP DIODE - SCHOTTKY DIODE (400V)

I have tried to achieve deadband too but still, the same issue. I have converted AC to DC using a bridge rectifier, inductor (15mH, 4A), and DC link, (400V, 2200uF), then I fed this to my inverter,

if I miss any details or the knowledge, I would appreciate it, Thank you.
here is my attached inverter circuit design.
 

Attachments

LowQCab

Joined Nov 6, 2012
1,922
Your "IRF-830" FETs are barely adequate for the job at hand.
What is the "Locked-Rotor" Current of your Fan-Motor ?

The IR2112 FET-Driver is a discontinued part and can only supply ~200ma. worth of Gate-Drive-Current,
You need a Driver in a Package that can dissipate some Heat,
and deliver at least a couple of Amps, like a TO-220.

A single Gate-Resistor is all that should be required, ( and may not be required at all ),
let the FET-Driver create the appropriate "Dead-Time".

There are Fully-Isolated FET-Drivers with Opto-Isolation built-in.

You also are not doing anything to control the Flyback-Spikes generated by your Fan-Motor-Windings,
this may require a few Capacitors, and/or, some fast TVS Diodes.

Is that 311-Volts shown coming from a stout, well filtered, Power-Supply ?

Are You Modulating a proper 50 or 60hz Sine-Wave using a Frequency of 10khz ?
.
.
.
 

Thread Starter

Claudez

Joined Mar 24, 2019
28
The fan current is .35A and the fets are rated at 4.5A
Is it neccesary to switch to IGBT instead? I will consider such gate drivers, and for tthe output. Before connecting to the fan,
I have a common mode choke of 2A 12mH and a capacitor of 2.5uF in series with a 10Watt 50Ohm resistor which gives a cut-off frequency of 915Hz.

for the input of 311V, i have rectified 220V with an inductor of 4A 15mH and DC link cap 2200uF to give me the input to my inverter.
 

Thread Starter

Claudez

Joined Mar 24, 2019
28
Yes, i think it is well with the 10kHz, i have used SPWM with Arduino to generate 60Hz, and i intend to vary the speed of the motor later, so i made in range of 0 ~ 60Hz
 

Thread Starter

Claudez

Joined Mar 24, 2019
28
Your "IRF-830" FETs are barely adequate for the job at hand.
What is the "Locked-Rotor" Current of your Fan-Motor ?

The IR2112 FET-Driver is a discontinued part and can only supply ~200ma. worth of Gate-Drive-Current,
You need a Driver in a Package that can dissipate some Heat,
and deliver at least a couple of Amps, like a TO-220.

A single Gate-Resistor is all that should be required, ( and may not be required at all ),
let the FET-Driver create the appropriate "Dead-Time".

There are Fully-Isolated FET-Drivers with Opto-Isolation built-in.

You also are not doing anything to control the Flyback-Spikes generated by your Fan-Motor-Windings,
this may require a few Capacitors, and/or, some fast TVS Diodes.

Is that 311-Volts shown coming from a stout, well filtered, Power-Supply ?

Are You Modulating a proper 50 or 60hz Sine-Wave using a Frequency of 10khz ?
.
.
.

I have never scope the output of filter cap from the inverter because I’m not sure, my scope is 300V rated, i heard because my probes are 1000V rated is ok to measure over 300V, how is this?
This is my probe photo?
 

Attachments

Irving

Joined Jan 30, 2016
2,541
You don't need to consider IGBTs for this relatively low-power load.

I have a common mode choke of 2A 12mH and a capacitor of 2.5uF in series with a 10Watt 50Ohm resistor which gives a cut-off frequency of 915Hz.
Can you explain this further please... did you mean this:
1638302172807.png

There's no need to 'filter' the pseudo-sinewave from the bridge, the motor will do that for itself. Also this will create massive back-emf spikes that could also be the cause of your burn-ups.

Notwithstanding the above, your problem is possibly due to the IRF830 having an Rds(on) of 1.5ohm, Coupled with an inductive load the MOSFET and its intrinsic diode are dissipating around 11W (peak switching pulse of around 200W). Those heatsinks are wholly inadequate as they have very little thermal mass. Also your layout, with long gate and source connections, is likely to be introducing parasitics, causing ringing and will further increase switching losses. Ringing will potentially causing gate punch through which would be a reason for your IRF2112 failures.

Also, when reducing speed are you reducing your mean voltage? Remember that revs = volts x Kv, but back-emf is also a function of speed so as you reduce frequency you must reduce volts else the current in the windings increases significantly... which could also be a factor in your burn-ups.

Motor equivalent circuit:
1638303143875.png
 

Thread Starter

Claudez

Joined Mar 24, 2019
28
You don't need to consider IGBTs for this relatively low-power load.


Can you explain this further please... did you mean this:
View attachment 253906

There's no need to 'filter' the pseudo-sinewave from the bridge, the motor will do that for itself. Also this will create massive back-emf spikes that could also be the cause of your burn-ups.

Notwithstanding the above, your problem is possibly due to the IRF830 having an Rds(on) of 1.5ohm, Coupled with an inductive load the MOSFET and its intrinsic diode are dissipating around 11W (peak switching pulse of around 200W). Those heatsinks are wholly inadequate as they have very little thermal mass. Also your layout, with long gate and source connections, is likely to be introducing parasitics, causing ringing and will further increase switching losses. Ringing will potentially causing gate punch through which would be a reason for your IRF2112 failures.

Also, when reducing speed are you reducing your mean voltage? Remember that revs = volts x Kv, but back-emf is also a function of speed so as you reduce frequency you must reduce volts else the current in the windings increases significantly... which could also be a factor in your burn-ups.

Motor equivalent circuit:
View attachment 253908
the filter circuit looks like,
1638328024823.png

i have read from one article that we can filter the spwm signal into sine wave using this kind of circuit to relax the motor from high dv/dt.
 

Thread Starter

Claudez

Joined Mar 24, 2019
28
Also, when reducing speed are you reducing your mean voltage? Remember that revs = volts x Kv, but back-emf is also a function of speed so as you reduce frequency you must reduce volts else the current in the windings increases significantly... which could also be a factor in your burn-ups.
in the code, I have designed it in such a way I reduce the frequency with the voltage, following the v/Hz ratio.
but currently, I'm trying with just a fixed 60Hz frequency.
 

Thread Starter

Claudez

Joined Mar 24, 2019
28
Notwithstanding the above, your problem is possibly due to the IRF830 having an Rds(on) of 1.5ohm, Coupled with an inductive load the MOSFET and its intrinsic diode are dissipating around 11W (peak switching pulse of around 200W). Those heatsinks are wholly inadequate as they have very little thermal mass. Also your layout, with long gate and source connections, is likely to be introducing parasitics, causing ringing and will further increase switching losses. Ringing will potentially causing gate punch through which would be a reason for your IRF2112 failures.
I think this is an important message to consider, what kind of heatsink are adequate for such circuit?
do you mean long gate resistors from the driver? and source connection from the low fets back to ground?
finally, about ringing and parasitic conditions, would them be reduced by shortening the wire distance?
Thank you so much for your effort.
 

shortbus

Joined Sep 30, 2009
9,326
I have been working on my project for quite a time on controlling the speed of a 220V, 60Hz, 0.35 induction motor (domestic fan),
So in effect your trying to make a single phase VFD? There is a reason that they are few and far between and are expensive if you find one. Google this board number and you can find the manual for it,with schematics of it. Then compare what they have to what your doing. https://www.ebay.com/itm/292316764444?ssPageName=STRK:MEBIDX:IT&_trksid=p2060353.m1438.l2649
 

Thread Starter

Claudez

Joined Mar 24, 2019
28

Thread Starter

Claudez

Joined Mar 24, 2019
28
is it possible to make a VFD with this breadboard?
i want to try on it then, later after I found my way then migrate to PCB board, but I would like to know about this.
 

Ian0

Joined Aug 7, 2020
4,860
the filter circuit looks like,
View attachment 253920

i have read from one article that we can filter the spwm signal into sine wave using this kind of circuit to relax the motor from high dv/dt.
You can't put the two halves of a common mode choke in parallel. The common mode choke requires current flowing in the opposite direction in one winding to prevent it from saturating. If you put two halves in parallel the same current flows in both windings and it will saturate immediately.
Use an inductor with a single winding that was designed for the job. It will have an iron-powder core or a ferrite core with a gap.
 

Thread Starter

Claudez

Joined Mar 24, 2019
28
I understand, Thank you for your information. I had a hard time finding an inductor with such value, can you please share any info on how I can find it?
actually, I didn't also work with inductors so much so I'm thankful for your correction. @Ian0
 
Top