What am i missing here?

I got into electronics while messing around with a java based circuit sim, now the need has risen to create a 300w ~60v PSU. I got a fancier simulator and went to work and this is what i ended up with:


Pretty self explanatory, the values i had for the sim are
AC source: 230v 50hz
T1: 1:5 ratio
D1-D4: MUR640
R3, R4: 10k
C3: 10uF
C1: 100mF
Q2: NDF10N62ZG (would be replaced with something cheap and available around here)

The sim says it works, output from C1 would yield 55v @ 400W with 400mV ripple @ 200hz. I got some ideas on how to bring the ripple down and reduce the voltage drop on load, what im more interested in is are there any really obvious mistakes that dont show up on the simulation? Or is it really that simple?

hazarada said:

The sim says it works

Click to expand...

The sim's misleading you, it won't work.

hazarada said:

are there any really obvious mistakes

Click to expand...

Yes, the whole circuit is wrong and won't work.

The main thing you're missing is that the mains voltage isn't floating, as in your sim. The neutral wire is close to ground potential. Try connecting one side of your AC source to ground and see what happens.

took me a while to figure that out, but then it hit me! I'm misusing the ground, the ground in the schematic isnt meant to be the actual ground but rather imagine that all the wires ending with the ground marker are simply connected. I've been using the ground as a teleportation pad to make neater looking circuits :/ that begs the question however, what is the ground for? As far as i know its just a wire that leads into.. the ground, how weird is that.

EDIT: revised version!

I don't know how that is supposed to work. For a start, the top winding of the transformer has a few hundred volts DC across it, so very high current will flow and it will burn.

the working principle is that the transformer would provide a reference voltage which would be accumulated in C3 (which is very small, according to the sim the average current across the transformer primary is ~0.0016A). If the voltage in C3 is bigger than in C1 then it creates a voltage at the mosfet gate so C1 will be filled up. I also added a diode between Q2 and R5 since even though i didnt see it in the sim, i think there could be a backflow from C1 to C3 causing a runaway chargeup should the voltage in C1 ever be higher than C3.

little update



shifted things around a bit, i did discover a giant flaw though. The efficiency is like 20%, i suspect its due to the mosfet being permanently half open, if C1 gets drained in pulses the efficiency improves dramatically but im looking into ways to fix it within the psu.

hazarada said:

the working principle is that the transformer would provide a reference voltage which would be accumulated in C3

Click to expand...

Transformers don't pass DC, the output voltage from the secondary is always AC (or nothing).

hazarada said:

according to the sim the average current across the transformer primary is ~0.0016A

Click to expand...

That's impossible. There's a couple of hundred volts of DC across the primary. The resistance of the primary would have to be about 100K for the current to be so low.

Seriously, that simulator is misleading you. I guess the transformer model it uses is flawed, or oversimplified.

Yes well i discovered the old design is total crap as it functions as a linear resistor.



here is my new design, trying to get into the switching action where the main transistor would squirt electricity into the main capacitor in short bursts, its sorta doing that but still kinda sucks.
I replaced the AC input + rectifier with a dc source for the sake of simplicity for now
D2 is a 60v zener
R7 is a dummy 10 ohm load
R1, R2 is 10k
R8, R3 is 100k
C2 is 10mF
C3 is 500nF

under a virtual oscilloscope it looks like this:

where the blue line is Q4 gate and red line is C2 voltage

right now the "switching" is caused by C3 that delays the gate opening/closing to make it wavy (thinking of better ways of doing that)
as you can see the energy transfer is still being done while the gate is just beginning to open
the other problem is that even if there was enough time, it only opens up to around 160v despite being driven by ~300v which i dont understand why.