3
\$\begingroup\$

TL;DR: Go to the bulleted list below.

As a weekend project I tried to design a simple power supply (without sofisticated switching mode ICs, since they are very hard to find where I live). So after some calculation and study, I came up with the following circuit, it is supposed to be a 7V 4A power supply for small projects, enough to power some microncontrollers (5/3.3V 100mA), small DC motors (5V 1A), and charge a li-ion battery :

enter image description here

I am aware of the nonidealities of these components, so extreme precision is not a important role here, since I don't intend to replace my good ol' bench supply. I will wire a digital voltmeter/ammeter to the output for calibration.

That said, I would like to ask some questions:

  • About the resistors R3 and R6, is it correct to pull up a MOSFET gate for such operations? Should I pull it down instead? Or remove the resistors? I am concerned about LM324 output sink/source capabilities with those resistors.

  • Also, in my output circuit, should I add other components to improve safety and stability?

  • And last but not least: is it safe to charge a li-ion battery with this circuit? (Considering a 2200mAh battery, with 4.2V max voltage) If not, how could I improve this circuit to achieve this?

Any help/advice will be very apprecieated!

Datasheets:

IRF540 - https://www.vishay.com/docs/91021/91021.pdf

LM324 - http://www.ti.com/lit/ds/snosc16d/snosc16d.pdf

1N4732A & 1N4737A - http://pdf.datasheetcatalog.com/datasheet_pdf/bkc-international-electronics/1N4728_to_1N973B-1.pdf

\$\endgroup\$
  • 1
    \$\begingroup\$ Why are all your schematics drawn sideways? Convention for legibility is positive supply at the top and negative (and usually ground) at the bottom. All your ground symbols are sideways. \$\endgroup\$ – Transistor Jan 13 '18 at 18:36
  • \$\begingroup\$ Well, the original circuit was following the convention. I just moved the blocks to fit them in a rectangle with a aspect ratio of around 1:9; so it would be proportional to most screens when there is a taskbar or title bar, and fit in the entire view. But will rotate it in just a sec! \$\endgroup\$ – KawaungaXDG Jan 13 '18 at 18:54
  • \$\begingroup\$ What are R1 and R4 for? They will dissipate lots of power at 4A \$\endgroup\$ – Sean87 Jan 13 '18 at 19:44
  • \$\begingroup\$ I inserted your rotated images. They aren't any larger than the original sideways jobbies, and they are a damned sight easier to read. \$\endgroup\$ – JRE Jan 13 '18 at 20:05
  • 1
    \$\begingroup\$ Some thoughts (I may change this to an answer). The pullups are not necessary (but pulldowns of the order of 100k might be useful for startup performance); the feedback loops will definitely need compensation (or at least analysis). \$\endgroup\$ – Peter Smith Jan 13 '18 at 20:13
1
\$\begingroup\$

About the resistors R3 and R6, is it correct to pull up a MOSFET gate for such operations? Should I pull it down instead? Or remove the resistors? I am concerned about LM324 output sink/source capabilities with those resistors.

R3 and R6 have no effect, except maybe during startup. I would make R6 a pulldown to help start the output at 0V.

Also, in my output circuit, should I add other components to improve safety and stability?

As drawn, the current control is probably unstable because it has gain but no compensation. Add a resistor (10k?) between the shunt and the inverting input, and a small capacitor (100p?) between the output and inverting input of U1B. This slows down the current control by making fast changes to its output appear immediately at its inverting output and will prevent it from oscillating.

Also, the output resistance will be 1 ohm, because the voltage control doesn't consider R4. Overall I don't think this is a good setup. You should look at some other bench power supply schematics and see what they do.

\$\endgroup\$
  • \$\begingroup\$ Oh thanks for the answer! I will add your improvements to the circuit! So, after these modifications, will it be okay to charge a battery in such circuit? As I pointed out in the last item. PS.: I've looked at some circuits and I just tried to simplify then, trying to keep the same functionality. Now I am not sure if I did it right '-' \$\endgroup\$ – KawaungaXDG Jan 13 '18 at 20:40
  • 1
    \$\begingroup\$ In case of failure of the feedback, R3 and R6 could turn their MOS on permanently causing short circuits and burnout. (if I understand the schematic correctly) \$\endgroup\$ – Fredled Jan 13 '18 at 23:04
  • 1
    \$\begingroup\$ @Fredled if by failure of the feedback you mean U1 catching fire, yes. Otherwise U1's output dominates. \$\endgroup\$ – τεκ Jan 13 '18 at 23:20
  • 1
    \$\begingroup\$ Maybe I'm wrong: R3 and R6 will turn their MOS on permanently and immediately causing 12V to load without any regulation, (only R4 would put some limitation). It's like connecting OUTB to 12V directly. U1 and U2 have no effect. \$\endgroup\$ – Fredled Jan 14 '18 at 0:20
  • 1
    \$\begingroup\$ @Fredled U1's output impedance is what, 100 ohms open loop? \$\endgroup\$ – τεκ Jan 14 '18 at 1:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.