# DC square wave Transformer

simulate this circuit – Schematic created using CircuitLab

I have dc into a bunch of 555 timers to create a square wave single that goes into the gate of the n-channel mosfet. im trying to power a transformer with 60 volts dc power. the resistance on the primary coil of transformer is 4 ohms. why will this not work? i do a open circuit test on transformer. it pulls more than 2 amps with no load. need some advice. i just need this transformer to work

• You need to push and pull, or else you will run your xfmr up to its saturation current, perhaps google "12 volt inverter schematic" for inspiration May 23, 2017 at 23:59
• how can you push pull dc currect
– joe
May 24, 2017 at 0:03
• sorry how can you push pull with a mosfet switch
– joe
May 24, 2017 at 0:05
• Sample of a low cost inverter, electroschematics.com/9155/12v-to-220v-voltage-inverter can you use a transformer with a center tap May 24, 2017 at 0:11
• yes i was just looking at that site. i dont have a center tap on this transformer. i can get one designed tho. my problem is that i have a square wave freq gen from 0-10khz that i must change freq. so im trying to go from 20 volts(pwm) to 3500 volts(out put of transformer) i though i could just push dc and the transformer would work.
– joe
May 24, 2017 at 0:17

First, D1 and D2 in the above diagram serve no purpose. They drop 1.2 volts from the power supply. That's it. I just wanted to get that out of the way.

Transformers are tricky. They don't just give the input voltage times the turns ratio. Transformers can only pass alternating current. Balanced alternating current. If the voltage across them has a DC component, the current on the primary side will increase in the direction of that DC component until the primary resistance cancels it out and/or the transformer saturates. You never want that to happen. For every Voltsecond you apply in one direction you have to apply another voltsecond at a later point in the operating cycle to balance it out. This keeps the current from building up in one direction.

Here's a simplified version of your circuit in my favourite circuit simulator. Notice how the current just keeps on going up and up and up.

What you need is to switch the voltage to the opposite value for at least half the cycle. Here's what that looks like:

Here, we use two diodes and two switches. The diodes let the built up current run backwards, dumping energy back into the power supply. Try playing around with that circuit. You'll find that bringing the duty cycle above 50% causes the same problem as in the previous design, the current just keeps increasing. You're applying 60 volts in one direction for more than half the duty cycle and 60 volts in the other direction for the rest which makes the current build up in one direction as it did before.

# Half bridge and Full bridge topologies

The most used topology in power electronics is the half bridge. You find it everywhere from DC-DC point of load converters to motor drivers to Utility scale HVDC power conversion Put two of them together to get a full bridge.

Here's what your finished circuit would look like. The switches stand in for MOSFETs. I'll get to the practical implementation in a little bit.

In this case, the transformer is the Load the half or full bridge drives. Everything is nice and symmetric. With a half bridge configuration you only drive the transformer with half the input voltage. You also have to make sure the capacitors on the other side don't empty out. Try reducing their size or increasing the resonant frequency of the system and see what happens. (That's why the diodes are there.)

Implementing a half bridge is pretty simple. You need an independent 12-20V power supply sitting referenced to your power voltage. Add a few diodes BJTs and resistors and that's it. Here's the circuit for driving a single half bridge. The independent power supply can be produced by a charge pump. For prototyping it's easier to just use an independent lab supply for both but for the finished product a charge pump is just the circuit above. A 555 timer, two diodes and some passives. All it has to do is supply ~10mA to the upper gate driver.

That's pretty much everything. Here's what you should do:

1 Build the High side gate driver and test that it works.

2 Build the rest of a single half bridge.

3 Test that it switches correctly

4 connect it in the half bridge configuration to your transformer with a split set of electrolytic caps on the other side and try driving something.

Hopefully it works.

• thanks for your comment that showed me a lot of info. first to answer your question i have a pwm variable freq 0-10khz (mosfet gate 20 volts) to step up voltage to 3500 volts to power a series LC circuit. witch is forward voltage lc circuit. my problem is that i have a pwm at 20 volts. i need to get the pwm to 3500 volts. i got a transformer designed and built its a regular transformer that is designed to handle 150-10khz 1 amp. my problem is that i can research different topology but if i have never seen them i cant research them and learn how they work.
– joe
May 24, 2017 at 1:34
• if you dont mind what would you call the topology that you gave in your first example so i can research. also thanks for your time. any info would be greatfull
– joe
May 24, 2017 at 1:34
• The series LC circuit is operating at resonance? That's why you need the variable frequency right? The frequency range matters. There's no way any transformer based circuit can operate at 0Hz. If you can describe your application a bit more that would help. The load on the output for example. May 24, 2017 at 1:47
• again thanks for your input. but what i really dont understand is how would i regulate the curre nt flow. if the primary coil is 4 ohms volts is 60 volt. the current flow will be 15 amps. i need to regulate current and have the still work transformer working
– joe
May 24, 2017 at 1:50
• correct richard it will be at resonance at 3500 volts is the end goal
– joe
May 24, 2017 at 1:50