Mini Lab Power Supply

[blueskull]

Inspired by Dave's mini lab power supply project, I designed my own one.
It is going to be an open source project.
Since I'm not ready to prepare and publish all my design files, and the software part has not started yet, I posted this design here, rather than open source hardware section.

The design philosophy is simple, I just want to have something to convert USB 5V to any voltage from 0V to 15V.
Maximum output power is 5W, while maximum output current is 2A.
The design should be as small as possible, so that it can be used as a programmable lab power MODULE that fits in any projects.

Unlike Dave's LED display interface, I want it to be an OLED and USB dual mode device. It should be able to be controlled by CapSense Keys with OLED, or simply PC software.
Also, full isolation is required. Ideally it should be designed to meet UL/IEC60950-1 spec, but only functional isolation is acceptable.

My current design is a 3 stage converter. The first stage is a flyback isolator, and the second stage is a synchronous buck.
The third stage is a LDO, to provide a very clean output. This buck+LDO topology is both high efficiency and low ripple, if control loop is properly designed.

To make the design process more challenging, I designed my own primary stage gate driver and I decided to use full digital controller on all SMPS.
The first stage uses a voltage mode and current mode hybrid control algorithm, the second stage uses simple voltage mode algorithm, and the third stage is controlled by an opamp.

The attached is the schematic, in PDF and SchDoc generated by AD14.3.16. I'm now working on its PCB layout. Please give me some suggestions on this design.

[prasimix]

If power supply is going to be miniature that does not assume that schematic also has to be "mini" and packed on the single sheet . Even if that is a limitation of application which you choose for design I'd like to suggest to break it up into more modules/sheets that other people could follow your ideas more easily.

[Marco]

That schematic is a work of art (which is not necessarily the same thing as pretty of course).

[Marco]

Why did you decide on a PNP output stage? It's always going to give poorer regulation than a NPN one, because of the flyback giving you a higher voltage to use you have the headroom to control a NPN output stage (or a N-MOSFET for that matter for even lower drop out from the buck).

[Psi]

If you want it to be really small and be a 3 stage design youre going to need to consider efficiency. Even if each stage is 90% its going to add up to 72% for the entre system. Luckily your only dealing with usb levels of power so it shouldnt be too bad. It also depends how small you mean.

[rob77]

interesting project, i like it 
but you really need to learn drawing schematics the hand drawn ones are OK - i can "see" them, but the PDF one is.... well... i was starring at it few minutes till i realized what am i looking at... it's ok and actually preferred to draw connectors separately and name the signals,but it's not good to draw a transformer separately as a rectangle without drawing the windings and just name the signals....