SG3525 push pull converter issues.

[Mbox314]

I am trying to build a battery powered inverter that takes 13-17volts DC from a lithium battery and outputs 120V modified sine wave AC

I tried building the Inverter using a SG3525 that is featured on Tahmid's blog here

I am having alot of trouble getting it to work right,

The output A doesn't work, there is no signal coming from the pin 11 on the IC. Outbut B from pin 14 works fine.

The duty cycle from output A increases as voltage on the inverting input increases, I don't know why.
The non-inverting input is wired to pin 10 as shown in the instructions.

Any idea what I might be doing wrong?

 

[george_leon]

both - and + input of error amplifier should be less kept below +5v(Vref). So keep NINV IN at +2.5 v and design feedback circuit accordingly.

 

[FvM]

both - and + input of error amplifier should be less kept below +5v(Vref). So keep NINV IN at +2.5 v and design feedback circuit accordingly.

According to be understanding of the datasheet, the error amplifier common mode range does include Vref, in so far there shouldn't be a problem to connect N.I. input to Vref. It's a common usage of SG3525 as far as I'm aware of. Furthermore, the observed problem can't be related to incorrect error amp operation.

There's most likely a different problem, wiring error or possibly damaged SG3525.

 

[Mbox314]

Thank you for the reply. I will try to solder up a new board for the SG3525 and hopefully the problem does not persist. I will let you know what happens on the second attempt.

 

[mrdevil83]

The pin 10 of the 3525 is power down and the only thing in that schematic that is connected to that pin is a resistor from pin 15 witch is Vref 5.1V. Yhe idea is that vhen the switch from pin 10 grounds the chip he turns on. When the switch is turned of and the Vref is on pin 10 3525 is in it's off condition.

study the schematic again and rewire the components.

Is that green thing I see after the etd core the output choke? I think you need atleast 2mh and even if it is a ferrite toroid I think you need atleast 35 turns.

I don't see the 470k and 68nf from pin 9 to pin 1 anywere.
Please post a image from the conections side of the board.

 

[jtrao]

It is not a good idea to test the switch mode power supplies on bread board, unless you are expert in power electronics. High frequency switching currents creates lot of noise and spikes in ground. PCB with good layout is recommended.