Someone did a nice job of reverse engineering ( well copying at least) another student's design from Switzerland, who now works in that field. I believe the drawing is penned by Taner Nedzhip & Davraew_nInBveyr.
WHat concerns me is the lack of attempt to define the specifications of each block in the diagram and the interface to each block and theoretical analysis. Instead a trial & error approach leads to cross your fingers results. Comments like "it works" is perhaps, loosely based on the assumption that the individual actually knows how to test it with a given load & environmental condition and has acceptance criteria.
I suggest each of you review the specifications of a simple Capacitor and pay attention to the details of all the specifications and see how an error in your assumption that this does not matter can affect the outcome. For example;
1. How would performance improve with choosing Thin Film Polypropylene Caps? ~$1
2. What is the impedance or at least the EST of the electrolytic caps used for the battery source and Bridge V+ and the effective impedance of the MOSFET and CAPS?
3. How does that affect the output change under linear and non-linear load conditions?
4. What is the surge level compared to rated level on worst case conditions for each critical part?
5. How does the load R on the current transformer (CT) affect saturation levels? I believe that CT is 1:1000 ratio with a rating of 20Amax with a 1Ω load and 4Amax@ 100Ω.
6. You ought to know by now, the relationship between max flux and electric field { B-H curves } are affected by the load R, which in this case is non-linear function of frequency as the load is the OA output impedance which increases with KHz + the the feedback values 100Ω & 1.5KΩ for each stage.
7. The design should not have undertaken to modify the secondary windings, rather understand the BH loop, test it. get Mfg specs, consult with them or at least reduce the Load R and increase OA gain with another stage if necessary.
8. If I were to evaluate efforts on this Rev. Eng, and I had time, it could have been simplified slightly and improved.
9. I am concerned about the instability of two stage PIC control and would not be surprised there may be problems here under dynamic load, especially loads with high crest factors (Pk/RMS ratio).
10. I expect more from Eng. Students in there efforts to learn.. Start by making a spec and a block diagram with specs for each block to create like a virtual component or block with inputs and outputs and a function curve or min/typ/max tolerance and sources of error.
11. How would you change the design schematic to show functional blocks so it can easily be re-design for any size with Select on Build P/N's Then once that is done, it is scaleable from 1kW, 5kW, 10kW,50kW etc. So that way of thinking gets you to understand how it works and what is affected and not affected to make a universal design..
12. How is harmonic content of the voltage affected by Power Factor PF of the load? e.g. Could you use Triac dimmed loads on the output?
13. (Although I admit to not reading all of the contents of this thread) the efforts, I have seen show thought into
design criteria, test plan, test results. even a simple table. Bad form, weak supervision.
I digress, I know you may be just students, but the internet has so much info more than my day where I had to rely on trade magazines in the library. So I expect more. or at least as much as we accomplished back in the 70's in Canada as students or young graduates. But I too remember being a Rookie, so I have patience. It takes time, patience, simplification to block diagrams then attention to details in each block and how that affects the over-all design.
Hobbiests? no problem have fun, watch TV.