Help needed in push-pull circuit for Inveter.

[koolguy]

Hi,

I am making/testing small MOS-based Inverter in which the freq is generated from 555 and as per diagram IRF540 is used the output transformer is of 12-0-12V to 220V i found when the one side of mos is connected at certain freq the output of inverter was 130-134V and when other part was connected the output voltage reduce to 40-45V.
so, why the push pull circuit is not giving double voltage as it should be please give me solution of it!


Thanks

 

[crutschow]

I don't understand. Why are you connecting only one side at a time? That will tend to saturate the transformer. You need both connected at the same time to get the push-pull drive.

 

[dr pepper]

Check your connections to the trans, you might have the centre tap incorrect, the centre tap should be the end of one winding joined to the start of the other, you might have the 'phasing' wrong.
Also you'd be better with a psu chip such as a sg3525.

 

[koolguy]

I don't understand. Why are you connecting only one side at a time?

I was testing..

You need both connected at the same time to get the push-pull drive.

I have done that..nothing happen.

Check your connections to the trans, you might have the centre tap incorrect, the centre tap should be the end of one winding joined to the start of the other, you might have the 'phasing' wrong.

I think i have connected center tap the middle one correct, but not sure about phase wrong...

 

[MrAl]

Hi,

You also have to check to make sure your 555 is putting out a near perfect 50 percent duty cycle square wave. If it is not, you will not get good results at all. One side would be driven for a longer period of time than the other and that would cause several problems including low output at the very least.

To test, swap the drives to the two mosfets and see if the problem persists ON THE SAME sides of the transformer. If the problem 'moves' to the other side, then the wave is not 50 percent duty cycle. Alternately of course you can check with a scope.
You might get some current ratcheting anyway though because it's not really possible to produce a perfect 50 percent duty cycle.

 

[koolguy]

Alternately of course you can check with a scope.

I don't have scope..

 

[MrAl]

Hello,


Then swap out the drives to the mosfets and see if the SAME thing exactly happens. If suddenly the 'other' lead has the problem instead, then it could be the 555 is not putting out a 50 percent duty cycle.
To be certain you have a 50 percent duty cycle, you could turn up the frequency of the 555 to twice what it was and use it to drive a T flip flop clock input then take the outputs from the Q and Q' of the flip flop.
There's also the issue of adding some dead time, where the drives would be made asymmetrical.

 

[MikeMl]

It is hard to make a 555 put out a 50-50% duty cycle. You have to include a diode in the timing network to make the charge time and discharge time symmetrical.

Also, what is the logic inverter? If it is CMOS, then the FET gate driven from the 555 has much better drive than the gate driven from the inverter. You could add a second inverter to make both gate drives equally bad...

The right way to do this would be to get a transformer with a third winding (also center-tapped), and use that to drive the gates. That way, the period of the oscillation is determined by the saturation of the transformer core; not by a 555 which knows nothing about what is going on inside the transformer core... The way you are doing it, how do you know that the period of the 555 is right to drive the transformer?

 

[crutschow]

You might try a simple, self-oscillating circuit such as this (it should still work if you substitute MOSFETs for the BJTs except you can use lower wattage resistors). The oscillation frequency is determined from the time it takes to saturate the primary. This saturation cuts off the transistor drive to the conducting transistor, causing its collector (drain) voltage to rise. This turns on the other transistor and reverses the primary current, which continues until the transformer saturates in the opposite direction.

 

[szhighstar]

I guess the transformer maybe is not correct.

 

[audioguru]

You mentioned "at a certain freq". A mains power transformer works properly only at the low mains frequency.
The transformer should be called, "24V center-tapped".

What will you power with the square-wave from the inverter? Incandescent light bulbs or a heater?
Compact fluorescent bulbs, motor speed controls, light dimmers and most electronic products depend on the higher peak voltage of the sine-wave from the mains.

Replace the 555 and inverter with a CD4047 IC that has its own oscillator, a divider that makes both outputs exactly 50% and one of its two outputs is already inverted.

 

[4pyros]

If you have no way to cheak the waveform from the 555 then yes use the CD4047.

 

[koolguy]

Also, what is the logic inverter?

I am using 4081 NAND to NOT gate..!!ok, i will try with 4047

 

[koolguy]

If it is not, you will not get good results at all. One side would be driven for a longer period of time than the other and that would cause several problems including low output at the very least.

HI,

Can you please explain me all problem beacuse i have to explain in my report!!

 

[koolguy]

OK, Today i have connected 4047 but i found the output from it was 5-6V. but Vcc was up to 12V. So, why it is ~5V from 4047 it should be 12V as MOSFET also turn full on at 12V...

 

[audioguru]

OK, Today i have connected 4047 but i found the output from it was 5-6V. but Vcc was up to 12V. So, why it is ~5V from 4047 it should be 12V as MOSFET also turn full on at 12V...

The CD4047 is Cmos so its outputs are rail-to-rail (0V to +12V when its supply is +12V and it has no load current).
Mosfet gates draw no current when they switch slowly at 50Hz so I don't know what is loading your CD4047.

Please post your schematic.

 

[alec_t]

What load do you have connected to the 4047 output? The output voltage will drop by a volt for every ~3mA you draw from it.

 

[koolguy]

The CD4047 is Cmos so its outputs are rail-to-rail (0V to +12V when its supply is +12V and it has no load current).
Mosfet gates draw no current when they switch slowly at 50Hz so I don't know what is loading your CD4047.

The schematic is same as you have given, i have not connected any load except the multimeter itself. ( i have not connected that 100ohm resistance with CD4047)

 

[audioguru]

A multimeter is not an oscilloscope. The 12Vp-p square-wave might measure 6VAC on a cheap multimeter because it simply averages the voltage.

The 100 ohm resistor, 16V zener diode and capacitor protect the CD4047 from voltage spikes.

 

[koolguy]

I also found that when i connect the single MOSFET gate to 4047 i.e. 2nd gate was open the output from transformer was not holding on multimeter rated on 600V AC.
Here i am posting some pic of it!!
as there was change in load and no-load voltage output from transformer ( transformer rating 12-0-12V 500mA for testing purpose only)