• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

inverter mosfet burning hot

Not open for further replies.


New Member
here i attached the diagram i have build.

what i'm sure are.
Occilator works OK
Mosfets are OK ( i Have tested each one seperately ) but only one of them which drived by NPN transistor gets hot in a short time ( 1-2 minutes then after burns ) both are fully turns on and off.

i've got to mention a thing is that as for testing i have connected 2 leds
and lit them like a flipflop i have observed that the one fed via NPN transistor lights brighter then fed by PNP transistor. assuming PNP transistor causing more voltage drop then NPN ones.

Since Mosfet drived by PNP transistor works grate i just think ( it might be wrong ) may be i should put grater resistor to the gate of the mosfet which
fed via NPN transistor ( Burning one ). i'm not sure that this will be the solution. i'll give it a try. or i should use another occilator

any suggestion please

thanh you.



New Member
i got a question, why do you use the BD139 and 140 transistors?
the 555 can drive the power tr mosfets by itself. and why did you use such power transistors(BD....)?


New Member
bogdanfirst said:
i got a question, why do you use the BD139 and 140 transistors?
the 555 can drive the power tr mosfets by itself. and why did you use such power transistors(BD....)?
i had no special reason for using BD type transistor i was just intendet to use NPN and PNP transistor. and may be i thought it would be better to use little powerfull transistor to drive mosfets. if this is wrong i can try less power transistor like BC type

555 should be able to drive mosfets by itself. but it can triger N channel mosfet at positif edge of Square wave. so i should have also use P channel Mosfet for negative edge of the SW. and i have decided to use transistor. I've got use one or another. isn't it ok...???


Active Member
I think the 555 supply is 15V(datasheet).In this case the PNP transistor always conduct. See this circuit.



Active Member
I think Sebi has it there.
You have a potential from the 555.
Try a decoupling capacitor,
maybe around five to ten MFDs



Active Member
The capacitor distorted the signal rising an falling edge, in this case bad choice: if the FET-s become slowly rising signal, also burned.
See my small modifyed circuit: here You can run the 555 from 5...15V, while the output sage working on 24V. In the original circuit was a problem with PNP stage: it can never closed, because 555 out in HI only 15V and in this case always remain 9V for open the PNP transistor.(it open 24V minus 0.7V!)


Active Member
Hi Sebi,

yes, you are assuming he is running the 555 from 15v,
which would mean that its output couldnt go over 15v.

As T2 would need its base to go to 24v to shut it,
you assume that it would always be open, and you
may be correct.

I think that the resistor chain R3,R1,R2,R4, thats
10K,20K,20K,10K, is wrongly valued.
Consider the voltages, of the chain without T1 and
T2 connected.
That would be 4v,8v,8v,4v, across the resistors.

I would expect that T1 and T2 would be well into
current with that chain applied.

This wouldn't actually matter if the drive to them
was enough to fully close each in turn, which i dont
think it would be from a 555.

I would be inclined to alter that biasing setup
until both T1 and T2 were closed, and include base
stoppers like the output stage, with two capacitors
from the 555.

If the 555 is run from a lower supply rail,
and i suspect that it is,
then the drivers should also be run from that rail.

If the 555 is being run at 24v it may be alright
for a while, but you should not leave it like that.

Here's a suggested alteration, i didn't really
get what Sebi meant about the capacitors,
i guess about five to ten MFDs, maybe a bit less.
On the diagram, ive put min current, for here that
should be only a bit away from conducting, i'm sure
you know that.

Best of luck with it.



New Member
John1 and the other guys, you realy are surprised me with your great efforts for sorting the problem out. i realy love this forum, i learn lots of thing.

i have sorted out the heating problem ( i have built another oscillator with IC3524 ) the invertor is now working without problem. last night , i have tested the invertor by using continuously for 1 hour, without any problem at all. Now i have 220 volts 50 hz voltage generated from 24 volt battery.

Now i am planning to change this inverter design into a UPS, can anybody teach me how to switch the line voltage to invertor voltage to supply the same load without any interruption.

what i am planning is to use tristor or triyak at output of the transformer to switch the line to invertor or vice versa. But i presume there will be very short interruption. I don't know how to get this rid off. I guess a large capacitor should be connected to supply enough current for a very short time ( 2-4 milisecond ) but WHERE SHOULD IT BE CONNECTED

Can anybody direct me on this regard too.


Active Member
Many UPS work with relay switch, and the charging via output transormer.In this case You need additional (precise) detector for exact switching. I recommend a continous charging for accu without any switching circuits.
Yes, the 3524 is a best choice, developed for this application.It contains also dead-time control.


Active Member
Thank you Sarac for your appreciation,

They dont usually switch the output from a UPS.
The usual arrangement is to always run the 'special load'
through the UPS, with the main supply keeping the
batteries charged, in the event of a main supply failure,
the UPS just carries on with no interruption.

Purpose built commercial units are usually done like that.

Switching the output from the UPS to restore supply to a
'special load' has other problems.
That sort of thing is done but care has to be taken with
the relay(s) or circuit breaker(s) to make sure that the
supply cannot feed backwards down the line.

I would suggest you don't switch the output,
use it as Sebi says, under continuous load and charge.

Regards, John


Active Member
Thanks John for "translating". Sometimes i just hope to understand my words,because i never learn english...


Active Member
You're welcome Sebi,

You speak English better than i speak your language.

Regards, John


New Member
john1 said:
Thank you Sarac for your appreciation,

They dont usually switch the output from a UPS.
The usual arrangement is to always run the 'special load'
through the UPS, with the main supply keeping the
batteries charged, in the event of a main supply failure,
the UPS just carries on with no interruption.

I would suggest you don't switch the output,
use it as Sebi says, under continuous load and charge.

Regards, John

You & Sebi are all talking abt ONLINE UPS. I'm aware of this, but my initial intention was build up an OFF LINE UPS bcs i thought that keeping the batteries on charge all the time, would effect it badly. for instance at max load ( 1 KW ) aprx.40A would be drawn from the batteries and i will have to use another transformer for keeping the batteries charged.
this charging current will not be less then 40A or may be it should be double ( load current + charge current ) thus i should use 2 KW transformer for input voltage.

I also thought that the electricity cost would be doubled bcs i will use 1kw
but pay for 2 kw

But on the other hand OFFLINE UPS's are using the batteries only when the blackout or brawnout Occurs and battery life would be longer. also i dont have to use second and bigger transformer then save money :)

Just thinking above i have decided for OFFLINE UPS. i might be wrong. If you could, please tell me what is wrong and what is right for above written thoughts by considering that i am a learner.

Regards to you all.


Active Member
Hi Sarac,

Yes, by having the UPS unit 'on-line' during use the feed to
charge the batteries would have to be sufficient
to run the load.

The extra current to keep the batteries charged however,
would not be much more, maybe not even noticeable.

I agree that converting to low voltage,
then back to 230 volts,
is wasteful, but its not as bad as double.

By having the UPS unit only nearby, not 'on-line' you would
be using it as a standby supply, a supply that you could use
if the mains were to fail.

The usual method in this case is to mount a socket(s) on the
UPS which is the output from the UPS. Then in the event of a
main supply failure, the appliance can be unplugged from the
main supply and plugged into the supply from the inverter,
in this way it ceases to be an "Uninterruptable Power Supply"
and becomes a standby supply.

It is tempting to think a set of relay contacts could easily
be arranged to switch the supply, and do this automatically,
but the electrical supply companies are usually very strict
about this sort of thing, and such a switching unit would
usually require written permission from them before they
allow its use, this would involve it being examined and
passed as fit for use by their representatives.

Such permission is rare for 230 volt systems.

I do not know of a 'fail-safe' system that would satisfy
the examiners, there may be one, but i dont know it.

In my experience every suggestion will get a rebuttal.
It is easy to think that a simple set of relay contacts
would do this, but the examiners will say that contacts
could stick, we may have people working along the line.

You may suggest an arrangement where two relays are used,
so the contacts are not shared, and 'sticking' would not
transfer any supply backwards.
An answer could be: what if this coil fails in use?
The examiners are not easy to satisfy.

The only system of transferring supply that i know of
which is acceptable (at 230 volts) for a domestic user
is where a different socket has the independant supply,
and there is no possibility of 'cross-connection'

Apart from the restrictions imposed by the electricity
authorities there are also practical difficulties to
be addressed when switching AC supplies, the problems
of voltage and sychronising.
The problem of voltage is not too difficult, but the
problem of synchronising is quite difficult to get
workable, although its not difficult to understand.

The additional problem of making the voltage and the
sychronisation automatic for the purposes of switching
is not something that even an experienced electrical
engineer would undertake to build from scratch without
considerable thought.

Another problem arises if the main supply resumes after
a short interval, the synchronising and voltage would
have to be matched before the standby supply could
switch the load back to the main supply.

It is not impossible,
but it is not easy,
and it is not a method that i have ever seen (for 230v)
and it would require permission to use,
which in my opinion would not be obtainable. (from
the electricity authorities)

However, if you use your own generator(s) or for some
reason you are independant of the main electricity
suppliers, then you can presumably do as you wish,
you would still have to meet the statuary household
safety regulations for domestic supplies,
earthing, insulation, etc.

Just give it a socket of its own.
Thats the easy way.
Or use it as a UPS, under load and charge.

Let us know how you get on, and which way you choose
to go, lots of us have battery run power supplies.

Regards, John


New Member

you talk such thing that i never thought of...

I totaly agree with you regarding safety problem, that indeed, in case of failure on switching, might be dangerious stiuation occur. I do not think anyone can guarantee that the system would be % 100 fail safe.

considering your words, i should concentrate for building up an ONLINE system.

But Just for experimental purposes i still would like to try OFF line ( Self Switching ) system then after change it to Online ( i 've got to save money for another Transformer :( )

thanks for your valuable support

regards, sarac
Not open for further replies.

Latest threads

EE World Online Articles