Update 10/12/15
Unfortunately, many of the originally ordered parts have not arrived, and it is starting to make the project a little problematic. However, I have placed my second order for parts and will have the LCD screen, new amplifiers, coil selection relays, sense selection reed relays, heatsinks for rectifier diodes and MOSFETs, the high flow fan, thermistors for temperature monitoring, the precision voltage divider network, and the reservoir capacitor drainage resistor. I have also received the SMD converted boards I had manufactured and I am finally ready to mount the MCU and begin testing. I am also having the parts of the prototype thermal tunnel cut on the laser cutter this week so that the heatsinks can be mounted and properly cooled.
Thermal design
SMD Converter Boards
On Friday, I started the construction of the thermal tunnel. The parts were laser cut out of acrylic and assembled. This design is merely a prototype and I already know of one revision that must be made for the final version.
On Sunday I began redoing some tests of the control and regulation circuitry using the actual parts that have been selected for this power supply. The rotary encoder was tested with an oscilloscope and a quick program that allows the microcontroller to understand the distinction between counterclockwise and clockwise rotation was created. The precision voltage divider (an 0805 package) was mounted onto a SOT-23 SMD converter board and tested for accuracy. The relative accuracy of the two resistors was .009% and I will likely not include a trimmer pot between the two resistors. I tested the voltage reference, at it was very accurate with a 5.0002 V output. This reference will likely be trimmed.
Finally, I retested the voltage regulator circuit and was encountered with a lot of trouble getting the system to work. I realized that the voltage comparator must be referenced to the negative side of the supply rail, but it also needs to be referenced to the high side to drive the gate. These ground points can be up to 40 volts off, so I will need to use another isolated winding to power a gate driver. Using an optocoupler between the output of the op amp (comparator) and gate of the MOSFET should work and a suitable pull down resistor must be selected on the gate line.