0
\$\begingroup\$

I plan to build a bench supply using a used ATX desktop computer PSU, with 3.3v, 5v and 12v binding posts as well as a cheap voltmeter-ammeter combo module (DSN-VC288). I know it's going to be inaccurate and have terrible ripple, but it would still be a nice addition to the variable bench supply I already have.

Now, I would like to add a rotary band switch to toggle the voltmeter-ammeter module between the three outputs. However, passing the output current through the band switch is not an option, as the switch is only rated for 300 mA, and the supply should go up to at least 5A (will be adding fuses on each of the outputs).

I have a bunch of IRFZ44N MOSFETs lying around which can handle 35A continuous DC, so my first instinct is to use those to switch the currents when the rotary switch is turned. However, I wonder if I am overcomplicating it - or using the wrong tool for the job?

I am fairly new to electronics so it's very possible that there is a standard way to do this that I'm just not familiar with. Or is the MOSFET route a viable solution?

\$\endgroup\$
  • 1
    \$\begingroup\$ I'm guessing you want something robust for this application--perhaps go with a relay-based solution. Either way though, some ATX power supplies will refuse to turn on if there's no load on some of their outputs, (my understanding is that this is because they're designed to be used in a very specific situation and not powering up when only some outputs are plugged in protects things from damage) \$\endgroup\$ – Hearth Feb 4 at 21:22
  • 1
    \$\begingroup\$ do you need a common ground between all 3 when not being metered? If so, that seems to contra-indicate the NCHAN fet idea, unless you get a higher voltage mosfet driver, but then you're really getting complex for a switch replacement. I would consider stand-alone SPST switches that can "mux" the lines, even if you only use 1 at a time... \$\endgroup\$ – dandavis Feb 4 at 21:29
  • 1
    \$\begingroup\$ @Hearth: fair enough, I'm just saying a LOT of people waste power+heat on these types of home-brews. If there's a 3.3v sense line (brown), it shouldn't be superstitious about minimum load. 100% correlation in the couple dozen i've seen... \$\endgroup\$ – dandavis Feb 4 at 21:32
  • 1
    \$\begingroup\$ @JensRoland In light of what dandavis said, perhaps skip the resistor load if it works without it. I doubt the slower switching would be a problem. \$\endgroup\$ – Hearth Feb 4 at 21:34
  • 1
    \$\begingroup\$ I would guess that module requires an external current shunt. In which case you can dedicate one to each of the three outputs and switch the sense lines. \$\endgroup\$ – sstobbe Feb 5 at 3:41
2
\$\begingroup\$

Another option is to get a current meter that can use an external shunt, then place one shunt on each output and route the shunt connections through the rotary switch. This way, you are not going to be switching the current path at all, which not only simplifies the design, it also ensures that the supply to the load is not interrupted when switching the meter to a different output.

\$\endgroup\$
  • \$\begingroup\$ Thanks - this seems to be the right solution; I just need to determine if I should do the sensing on the high or on the low side, and whether on the low side I would have to switch both sense leads or if one would be enough (with common ground) \$\endgroup\$ – Jens Roland Feb 6 at 6:51
  • \$\begingroup\$ I think you're going to want to do the sensing on the high side. The power supply you're using probably has a common ground across all of the outputs, so measuring the current on the low side of your output terminals won't tell you much of anything useful if you have two outputs connected to the same circuit. Measuring on the high side will actually give you the correct current for each output. If the power supply channels were totally isolated from each other, then this would not be a problem and low side current sense would work fine. \$\endgroup\$ – alex.forencich Feb 6 at 7:30
1
\$\begingroup\$

The easiest solution would probably be to use relays. If you're concerned about the switching delays or even the sound produced during switching, use solid state relays. They are quite fast, and essentially employ FETs for switching purposes. So you're just replacing your MOSFETs for better packed versions of the same.

However, in my experience, I've found that both of them heat up quite often. This might provide a more detailed view-https://www.ia.omron.com/support/guide/18/overview.html

\$\endgroup\$

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.