MOSFET Burning Problem in UPS

irf3205 use for ups

I want to discuss about the problem which i am facing...........

I have a Local Microcontroller based UPS in which MOSFETs are used as switching Inverter. Total six MOSFETs are used in 2 Sets of three parallel MOSFETs.

The MOSFETs used here are IRF3205.This UPS has disadvantage that the MOSFETs used here are burning so frequently, (The other UPS of same Model design have same problem).

I tried my best to find out the reason of MOSFET burning but found so many difficulties.....Mostly these MOSFETs burned during Battery Charging process and rarely it got burned during Inverting process. This UPS have a built in circuit of overload detection and it works fine if overload occur. Battery charging is also controlled and protect battery from overcharge.

Last day i resold new MOSFET and start checking its functions of Charging and Inverting. But after few minutes of Inverting Three MOSFETs of same side got burned one after another at the time of NO Load (As i know some UPS are not designed for No Load). After re-soldering more three new MOSFETs its work fine for one whole day (Charging and Inverting). But after one whole day same sides MOSFETs got burned out during Inverting process.

Guys, i am unable to find out the reason of these MOSFET Burning and need your help to troubleshoot this issue. May be this is not a good approach of MOSFET usage. So, for your reference i am uploading the schematic diagram (which i made by tracking of UPS PCB).

Thanks and Regards,
masterleous

irfp150n inverter schematics

The MOSFET polarity is drawn incorrect, but the real circuit is obvious. I wonder, how overload protection or charge control is achieved? Also it would be interesting,how the MOSFET gates are driven. Cause they have unusual high ohmic resistors (assuming the values in your schematic are correct), I guess they are driven by the uP directly. This may be part of the problem.

The MOSFET switching is rather slow, it's probably O.K. for a square wave inverter, but would cause inacceptable losses in PWM operation.

Generally, I assume that the UPS is simply bad designed, or at least not rugged enough for your kind of usage. Improving it requires possibly similar knowledge as designing a new one from the scratch.

ırfp 460 mos fet diagram

Thanks for your reply,

You are right that this is a very basic UPS (Square wave). MOSFETs are switching directly by Microcontroller through an optocouplor. Charging of batteries is not so much controlled, They are charging their batteries through these MOSFET and at fixed voltage, but the only overcharge controlling is a relay used in-between Battery and Transformer which cut the supply to the batteries on observing battery full voltage (via LM324 Op-Amps IC).

Anyway, is their any suggestion regarding the burning of MOSFETS, which i can apply directly / or indirectly to save MOSFET.

or give me an idea about safe working principle of MOSFET.

Thanks and Regards,
masterleous

mospet in ups

try a 15k resister between gate and source ( same point of 12v zener ) .
there may be some driving circuit microcontroller cannot drive 12v to gate directly.

which microcontroller is used. in some design microcontroller is used to drive pwm3525 IC

reasons for mosfet failure in ups

Microcontroller AT89C2051 is driving an optocouplour which is further driving MOSFETs,

Can you give the example or easiest way to drive SG3525 IC via Microcontroller. and is this IC is suitable to further generate such PWM Signal that is required to obtain Pure sine wave after filtration?

how use fet in ups

check the link sine wave generation
**broken link removed**

and 3525 microcontroller is only used to shutdown the PWM IC (Enable/Disable)

h igh frequency inverter irfp150n

Maybe the MOSFET's are not sharing current equally which will depend on the temperature of each device.

What type number of MOSFET are you using so we can see the current ratings. From this you can pick a suitable balancing resistor in each source leg.

mosfet burn reasons

The MOSFET type has been previously reported. MOSFET are self-balancing by it's PTC RON, so high current applications don't use balancing resistors normally, cause they would considerably deteriorate the efficiency.

I don't think, there's a fair chance to understand the reason for spontanous failure of your UPS from a distance. There are many ways to achieve this by simple design faults. It can be e. g. a matter of a faulty drive pattern from the uP. Also the high gate series resistors may enable destructive self oscillations of the power stage in some situations.

If the design is faulty (to my opinion, this is almost proven by the failure of multiple exemplars of the same instrument, assuming you have been keeping the allowed operating conditions), it needs a redesign. This can hardly work without respective knowledge and development tools.

P.S.: PWM operated inverter with sine output can't work with a slow gate drive, that is possibly acceptable for a low frequency square wave type. Switching losses can't be no longer ignored then.

blown mosfet ups

FvM said:

The MOSFET type has been previously reported. MOSFET are self-balancing by it's PTC RON, so high current applications don't use balancing resistors normally, cause they would considerably deteriorate the efficiency.

Click to expand...

You are right, but that requires proper gate drive design and parameter screening of devices from the same batch.
http://www.irf.com/technical-info/appnotes/para.pdf

Our lab once had a problem with a high power 5kW/50V DC load for testing power supplies. They had 2 banks of about 10 or more per bank MOSFETS mounted on a force cooled heatsink. This thing kept blowing MOSFETS now and then. Our service dept would then just purchase and replace the blown parts until another failure occur some time later.

We then advised them to fit a very low value source resistor for each MOSFET. That was the end of that problem.

He could try it as a test and see if that stops the devices from blowing up. At least he will know then that there is a possible problem with the gate drive causing a balancing problem.

problem in ups

Thank you for the explanation. For practical reasons, it's not always possible to guarantee usage of same batch devices. I think, that it's not absolutely neccessary, if there's some margin in power rating. It would be interesting, to know if the device failure, you reported, have been caused by current inbalance only or by parasitic oscillations. Source resistors can be expected to reduce also the latter danger.

I also agree, that weak gate drive and no source resistors is possibly a bad combination. Some kind of redesign is obviously required.

irf3205 mosfet u p s usage

There is no pulldown for gate drive.

A micro cannot pull it down to zero as there is a resistor inbetween.

A low ohmic resistor from gate to ground should work perfectly

Nandhu

irf 3205 u p s circuits

Thanks guys,

you have discussed my problem in a very helpful manner. Now i am reading these stuff, and will post the circuit which i'll make after reading application notes.

Thanks and Regards,
masterleous

irf3205 application notes

As i seen the out put stage it include a capacitor

of 33UF with 3.9Ohm resistor in sieres is the main loade on the MOSFETS It should be

.33UF to stop switching noise (it's a zobel net work)

site:www.edaboard.com mosfet+burn

ziaulislam said:

As i seen the out put stage it include a capacitor

of 33UF with 3.9Ohm resistor in sieres is the main loade on the MOSFETS It should be

.33UF to stop switching noise (it's a zobel net work)

Click to expand...

Thanks,

0.33UF instead of 33UF

What is Zobel Net Work? Can you explain a little

irf3205 app notes

A VERY SIMPLE PWM INVERTER TO INCRES

WATTAGE YOU CAN ADD MORE MOSFET.

irfp150n failure

Sorry i think it is not very clear i am posting in

RAR

irfp150n diagram

ZOBEL NET NET WORK USELLY USE TO STOP HIGH FREQUENCY OSSILATIONS & SEPRESSING SWITCHING NOISE.

pin1 of irf3205

Hi All,
According to my view there is a fail in zener diode , you have to remove both zeners totaly and replace them by 27k ohm resistors \]

burned ups

check this out

what is the problem of ups

ziaulislam said:

Sorry i think it is not very clear i am posting in

RAR

Click to expand...

Thanks for your circuit diagram