I want to make a Bench power supply unit based on:

  • For initial AC to DC conversion - ATX PSU (Enermax EG365AX-VE 350W) capable of providing 12V 26A

  • Internally Step up module for powering the below - one of below:

  • PSU output adjustable and provided by RD DPS5020 (input 6-60V, output 0-50V 0-20A)

For up to 10V application I would like to omit the step up module and power the DPS5020 directly from ATX. For higher voltages I want to manually switch to use step up module.

My first idea was to use 3P3T switch, but I could not find one that can safely handle 12V DC and 26A. Then I wanted to use the switch that would control 2 relays, but high current relays are very expensive. The third idea was to use high side switching with two Power MOSFETs.

N-channel MOSFET configuration: N-channel MOSFET configuration Link to the editable source

or alternatively P-channel MOSFET configuration: P-channel MOSFET configuration Link to the editable source

What MOSFETs should I choose? Are the following criteria safe:

Vds = min 75V 24V (12*2) (or should it be much higher?)

Id= min 30A (what if the mosfet would become warmer)

Vgs = +/-20V (I will ensure that the mosfet is fully open with Vgs=12V) - However as I have the ATX psu, I could easily use 5V or even -12V, - 5V.

To minimize the power loss (heat) the Rdson should be low.

Each mosfet won't switch frequently (and most probably not under load), as it would switch its state only when SW1 is flipped. Can I disregard then the switching power losses, so I should care only about conduction losses, witch would be P=IIR_dson.

How about:

  • IRLB8314PbF (30V, 130A, 2.4mOhm, 125W)

  • IRFB3607PBF (75V, 80A, 9mOhm, 140W)

  • IRFB4410ZPBF (100V 97A, 9mOhm, 230W)

How about the power dissipation and safe operation area? The conduction losses even for 26A would be max 6-7W.

To make sure that when then M2 is providing the 12V and the Step up converter is off there is no voltage on the + output pin I added the D1 diode, is it sufficient protection?

  • \$\begingroup\$ What load impedances? \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jul 10 '18 at 1:11
  • 1
    \$\begingroup\$ @TonyEErocketscientist As it should be the general use Bench PSU - some ICs, LiPo charger, etc. We should assume that there would be some load impedance. I lack experience how to anticipate it. Also the step up and step down components have capacitors and coil, so they may introduce and propagate some impedance on the MOSFETs too. \$\endgroup\$ – rafalmag Jul 10 '18 at 6:39
  • 1
    \$\begingroup\$ I would follow DIY Instructables, maybe no need to switch outputs , just have 12V , 5V, 3.3, -12V and booster \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jul 10 '18 at 10:30
  • 1
    \$\begingroup\$ @TonyEErocketscientist I plan to have 12V, 5V, 3.3V, -12V too, but I am really looking forward for voltage and current regulated output of DPS5020. \$\endgroup\$ – rafalmag Jul 11 '18 at 22:46
  • 1
    \$\begingroup\$ The problem with the n-channel setup: The MOSFET switches on if its VGS is high enough (12V if it was off before), but think what happens when the MOSFET turns on, there would now be up to 12V on the source pin, what does this do to VGS? If the p-channel MOSFETs are too weak, you can try use a n-channel LOW SIDE circuit or a n-channel high-side circuit with a gate driver with floating ground to solve the above problem. \$\endgroup\$ – Marco Zollinger Jul 12 '18 at 6:25

Your Answer

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

Browse other questions tagged or ask your own question.