Adjustable Current Supply

Inspiration I always want to challenge my analog circuit knowledge, as I learnt a lot but not applied in much designs yet. I want to challenge myself to design something also involving power management. Building an adjustable power supply requires heat management, circuit precision, etc. I love the challenge!

What it does

My project outputs from 0-1A current, and max now is 2.5W due to lack of high power resistors. When you want for instance 0.5A output, then you connect a resistive load to the output, for instance 1 ohm, then the voltage difference on the resistor is 0.5V. If you connect a 0.3 ohm one to the output, the voltage difference on the resistor is 0.15V.

How we built it

First, I find the desired output range for my project. Second, I draw the block diagram. Third, I design the schematic. (Throughout the process, I tried different designs, and I choose the one that works the best and fit my current component shortage situation). Forth, I build the actual circuit. Last, I test with oscilloscope to output analog voltage and check if the logic works. I also troubleshoot.

Challenges we ran into

1. Originally, I wanted to use ESP32 WROVER E's internal DAC to control the current supply. However, the MCU is not operating after around 1.5 hours of debugging. Then, I figured out that I need to switch method. I then used the oscilloscope to output analog voltage to control.
2. Originally, when I was thinking about using ESP32's internal DAC, I then find the output range do not match the MOSFET VDD. (ESP32 is 3.3V, MOSFET VDD is 5V), and I wanted to use PMOS to control, which makes the logic necessary. Then, I re-constructed my design to use NMOS logic. The trade-off is VGS on NMOS. NMOS works not too well as power switch when connect to high power, and PMOS works best there. However, to address the situation, I switched design and accepted the trade-off.
3. Power consumption was high. Even though now the max power consumption is 2.5W, but all of my resistors are either 1/4W or 1W. Then, I tried to connect resistors in parallel to distribute the 2.5W consumption and make individual resistors capable of operating.
4. I lack precise testing equipment. My oscilloscope is cheap, my multimeter is cheap, all my components are not very good, especially originally I wanted to design current mirror with lots of MOSFETs, but all my MOSFET are individual DIP package ones, which have large difference in Vth and W/L. Then, I cannot control the precision of current mirror. Finally, I accepted this situation and I just do one channel with power MOSFET.
5. Lack of output power. Like I mentioned, I lack power resistors above 1W, and even though I can connect resistors in parallel to distribute power consumption, I only have little 1W resistors. For each value, I only have 10. I cannot raise the output power to higher than 2.5W. I accepted this because I physically lack components.
6. I brought 4 power supplies, and 3 of them were broken once I arrived here. I then solved this by using ST Link to provide power.
7. My op amps are not rail to rail op amps. (rail to rail means the op amp can output well even though the output reaches either + or - VCC.) I addressed this by connecting one USB power gnd to the ST Link 5V to get ST Link gnd -5V. Then, this extends the power supply rail.

Accomplishments that we're proud of

1. Even though I lack lots of components, and my 3 power supplies were broken. I force myself to think and I figured out ways to address this problem.
2. Even though I lack components, I still achieved 0-1A adjustable current supply, and it is working well! I got no expensive ICs, only basic components and op amps!!! ## What we learned
3. I learnt that in reality, not every scenario will work the best for me, and I cannot expect everything to go well. I need to be more prepared. Or, if I really have no way to address the problem but regular methods, then accept the failure and find other ways to address.

What's next for Adjustable Current Supply

Built With

• circuit