20V3A constant current and constant voltage linear regulated experimental power supply

Tutorial posted on Digital Home  https://www.mydigit.cn/forum.php?mod=viewthread&tid=147224

Thermal stability test: Adjust the heat gun to 150 degrees and blow it on the board. It is found that the most sensitive thing to temperature is the voltage divider circuit, current sensing resistor and operational amplifier at the reference position. The changes are smaller than that of the reference position. After heating for 1 minute, the constant current output value goes from 1.017A increased to 1.024A, a change of 7mA, and Δt was visually greater than 70 degrees. It is definitely not that high in general use. After heating for 1 minute, the constant voltage output value increased from 10.96V to 11.02V, a change of 6mV. The thermal stability is pretty good, right? DC Stabilized power supply is an indispensable basic instrument in any electronic circuit test, and can be seen in basically all electrical-related laboratories. For an electronics enthusiast, a DC regulated power supply is also essential.

This project is a linear series stabilized experimental power supply that can output a voltage of 0-20V and the current limit value is adjustable from 0-3A. Because of the inherent disadvantage of low linear voltage stabilization efficiency, the power is relatively small. It is recommended to use 25v voltage input. The CCCV is set independently, and the constant voltage value and current limit value are set through two potentiometers. The temperature drift is low and the noise is extremely small (compared to the switching regulator). This design is simple and easy to install, with high practicality and stability. Well, the main purpose is to select simple materials, only use one DC power supply, and require constant current and constant voltage functions. Therefore, two reference voltage sources are used to avoid the troublesome problem of current detection. It is easy to make, does not require debugging, and can work if it is welded correctly.

Selection guide: The components of op amp, R6, R7, R10, R1, potentiometer, and TL431 directly affect the thermal stability of current and voltage. The low-temperature drift components in the BOM are ±50PPM. If the requirements are not high, ordinary grade components can be used.

Circuit simulation: The core schematic diagram of some peripheral circuits is omitted, and the simplest and most reliable three components of op amp + reference + NPN transistor are used to achieve closed-loop voltage stabilization. It is slightly worth mentioning that two benchmarks with different potentials are used, which simplifies the current detection circuit design and uses less expensive components while ensuring that the current detection and voltage detection do not affect each other. The disadvantage is that the input and output cannot share the same ground, otherwise it cannot Normal current limit.

Welding guide: The output current and voltage calculation formulas are included in the schematic diagram. You can calculate the resistance value based on the potentiometer you have on hand. The default output of the schematic diagram is 27.5V, 3.4A, but this product means that the power should not exceed 50w, because The power tube can't handle it if it's too big, unless you add input gear switching. The two pads on the right side of the power tube on the PCB are used to connect the pointer ammeter. This position is before voltage feedback sampling, so the internal resistance of the ammeter will not affect the output voltage. There is nothing much to say about soldering. First solder the reference and current-sense resistor to see if the output is normal 2.5v, and then continue to solder the others. If welded correctly, no debugging is required and it can be used right after welding. The finished product is as shown in the picture. It is best to fix the power tube directly on the radiator. Do not add silicone pads or mica sheets because the heat generation is huge and the thermal resistance needs to be reduced as much as possible.

Test photos: When short circuit output is allowed, the protection current and voltage can be adjusted from zero. The calculated value of the maximum voltage is consistent with the measured value, 18v. This can be freely changed and the feedback voltage divider should be as much as needed. When the input is 20v, the voltage can be stabilized normally. 18v no-load input and 19v output 17.72v. Thanks to the Rail-to-Rai output op amp, the minimum voltage difference can be reduced to 1.28v. It will be larger when loaded, but it is still within 1.5v when the transformer is used as the front stage. To measure the output ripple, I found an Heirloom Netgear transformer and added a rectifier filter to the current level 15vac to turn it into a pulsating DC load. The load is a 10 ohm cement resistor. At the output 1a current input end, you can see obvious 100hz power frequency voltage fluctuations, vpp reaches 1.11. The v output terminal is as quiet as the deep sea vpp13.2mv, with an effective value of several mv and a high PSRR. Compared with the ripple of a switching power supply that can easily reach hundreds of mv, it can be said to be very good and very suitable for testing power supplies for various precision sensors.

Thermal stability test: Adjust the heat gun to 150 degrees and blow it on the board. It is found that the most sensitive thing to temperature is the voltage divider circuit, current sensing resistor and operational amplifier at the reference position. The changes are smaller than that of the reference position. After heating for 1 minute, the constant current output value goes from 1.017A increased to 1.024A, a change of 7mA, Δt was visually greater than 70 degrees, which is definitely not that high for general use. The constant voltage output value increased from 10.96V to 11.02V after heating for 1 minute, a change of 6mV. The thermal stability is pretty good, right?

The attachment is a multisim14 simulation file project, which contains a single small board and a 2x2 pcb layout, which can be directly prototyped.