CD4047 ,Mosfet IRFP150 Inverter

Thread Starter

Feenz

Joined Apr 9, 2022
15
Hello members..
I have done a square wave inverter as the bellow image
My question is :
Is the design good ?
Is there a better design ?
I have a problem which is : mosfets get damaged every time I switch it on , one of the A733 transistors get damage as well
It worked before with the same design , ,, any suggestion ?Mos Drive2.jpg
 
Last edited:

LowQCab

Joined Nov 6, 2012
4,023
Both the A733, and C945 Transistors are rated for 150ma MAX,
you are putting a ~500ma spike through them when switching the FETs.

I just have to ask ......
Why are You trying to do this with discrete parts when there are so many excellent
"all-in-one" Chips available to do this ?
.
.
.
 

Thread Starter

Feenz

Joined Apr 9, 2022
15
First , I should change the design to these chips ,, keeping the IRFP150N
I just wonder is there a good method to input the signal to the (5 x FETs) on each side ?
In other words , are these 150 ohm resistors do the job in delivering the signal to the all 5 FETs ?
Or there is a better sort ?

Second , if I replace the C945 , A733 with a little higher power transistors , does that resolve the problem ?
 
Last edited:

LowQCab

Joined Nov 6, 2012
4,023
Those FETs are rather heavy-duty, and therefore have almost 10nf of Input-Capacitance combined.
The amount of Gate-Current required to switch these FETs will depend on
the Frequency that You need to switch them.

What Frequency would You like to use ?
( higher Frequency = higher switching Current )

You will very likely need a dedicated Gate-Driver-Chip to handle the required Current.
.
.
.
 

Thread Starter

Feenz

Joined Apr 9, 2022
15
Dedicated Gate-Driver-Chip to handle the required Current
For each single FET ?
If no , then how to connect that driver to all 5 FETs (The better method)?
 

ci139

Joined Jul 11, 2016
1,898
/!\ if you get this thing into "the full blast" and there is no significant load at the output - it may exceed the voltage tolerance of the II-ry /!\

  1. the 7812 input should be better decoupled from the +24v
  2. the bjt drivers should be moved to the 24V side or supplied with their own 14V or with their own 12V if you like
  3. /!\ if the TF can't handle your input current (saturates) - your mosfets get killed or your TF will melt /!\

your circuit has a high power capability = if you miss to predict a catastrophic failure - it will happen in a fragment of the flash
=
you can't rush with the overall design if your subsystems are not confirmed standalone + in conjunction with the entire system !!!!!!!!!!!!!!!!!
 

Thread Starter

Feenz

Joined Apr 9, 2022
15
Really nice information , but this inverter is aimed to get it's supply from a two 12 volts batteries
I can't figure how to use extra supplies sources , that's why i'v used the 7512 .
 

Papabravo

Joined Feb 24, 2006
21,159
I think you need to allow for some dead-time between the switching edges. You certainly don't want to have both banks of FETs on at the same time.
 

ci139

Joined Jul 11, 2016
1,898
a speculative target app - a simulation // simplified illustrative principial - for a shown non-reactive target load (you can study the possible effects of shorting the II-ry , "open circuit" the II-ry ((also removing the 10nF 10uF 100mH from the circuit)) , setting inductive/capacitive load on the II-ry , ...)
 
Last edited:

LowQCab

Joined Nov 6, 2012
4,023
Stop everything ........................
First You need to explain exactly what You are trying to accomplish with this project.

So far, it would appear that You are trying to build a 24-Volt to 120/240-Volt Inverter.
If this is true, then what are your Output requirements ?
Square-Wave,
Modified-Square/Sine-Wave,
True low distortion Sine-Wave,

What are your requirements for Voltage-Regulation ?

There are, so far, no provisions for Current-Limiting.
This is a recipe for smoked parts or even a fire.

What type of Transformer are You using ?
What is it's VA-Rating ?
Is it designed for Inverter-Service, or
are You using a Step-Down Transformer in reverse ?
.
.
.
 

Audioguru again

Joined Oct 21, 2019
6,672
Many modern products do not operate properly when fed from a squarewave. An old incandescent lightbulb or electric heater that has no electronics will work fine.
 

DickCappels

Joined Aug 21, 2008
10,153
Running a such a low switching frequency you can increase the gate resistors in order to reduce the turn-on current spike, and thus continue to use the existing transistors. An alternative is to switch to the 2N2222 and 2N2907 complimentary pair which can handle the 500 ma or so spike.
 

Ian0

Joined Aug 7, 2020
9,667
/!\ if you get this thing into "the full blast" and there is no significant load at the output - it may exceed the voltage tolerance of the II-ry /!\

No it won't - the voltage on the secondary will be the voltage on the primary multiplied by the turns ratio

  1. /!\ if the TF can't handle your input current (saturates) - your mosfets get killed or your TF will melt /!\
      1. It's voltage not current that saturates the transformer.
 

Alec_t

Joined Sep 17, 2013
14,280
According to the CD4047 datasheet the duty cycle of the Q output is not guaranteed to be 50%. That may lead to nett DC current in the transformer primary and hence core saturation (unless the transformer is designed to handle this).
The CD4047 doesn't provide any dead-time.
 

Ian0

Joined Aug 7, 2020
9,667
According to the CD4047 datasheet the duty cycle of the Q output is not guaranteed to be 50%.
I saw that, and wondered how, as the output comes from a divide-by-2 circuit, it could possibly be anything other than 50%.

Also worth bearing in mind that a 24V squarewave has a greater V.t product than a 24V rms sinewave, so the transformer may will saturate unless the frequency is increased by 5%, and the terminal voltage of a 24V battery may be considerably more than 24V, if it is on charge, which would also cause transformer saturation.

if the transformer is a toroid, then soft-start (which the SG3525 can do) is essential to avoid saturation at the end of the first cycle after switch-on.
 

DickCappels

Joined Aug 21, 2008
10,153
It is not a divide by two output. The CD4047 is a one-shot. Maybe a better choice would be to use a CD4060 which is an oscillator (crystal, RC or what have you :) followed by a string of counters. It will give you a good 50% duty cycle pulse.
 
Top