Adding a 5v rail to a PSU that doesn't have one

from what I’ve read, the 5v rails on PSUs are iffy because they have limited current (1.5A according to the datasheet for lm7805?). still, it seems a little wasteful to me to use a lm7805 voltage regulator in every digital module I make, because of the amount of energy lost as heat into multiple voltage regulators.

If I plan to make modules exclusive to my own system, can I use a single 5v rail, if I can stay within the 1.5A?

Also, if this works and I have more modules than allowable current, could I just make a second 5v rail using another lm7805? thanks!

Holy crap I thought Funque Mod was some kind of professional b

Yep. There’s no reason why that wouldn’t work.

Yes, as long as the two 5V outputs don’t get shorted together. You’ll need to tie the ground pins together, and you should be able to run them off the same 12V input.

Oh, also you’ll want some heatsinks (possibly with active cooling, they can get rather toasty) on those if you plan to draw the full 1.5A from them.


Awesome thank you for that! Time to look into some active cooling heatsinks.

1 Like

I would start by using heatsinks, then if the 7805s start to go into thermal shutdown you can add some fans to cool said heatsinks.


1.5 A is a lot. But a lot of PSUs do have much lower +5 V limits. The Frequency Central FC power for instance is specced at ~500mA at +/-12V, ~100mA at 5V. It uses 7812 and 7912 (rated at 1.5 A) for the ±12 V but 78L05 (100 mA) for +5 V, hence the feeble 5 V rail.

For comparison, my MIDI to CV module, which uses a Nano powered off the 12 V rail, draws 54 mA on +12 V and 17 mA on -12 V. Assuming about 40 mA is for the Nano then in principle an FC Power could power two such modules if they relied on the 5V rail for the Nano. Using the 12 V rail to power the Nano, an FC Power could handle about eight such modules.

So I use a 5 V regulator in any module that needs one, of which I have few. If you have a lot of modules that will want 5 V then one alternate approach is to use a PSU and distribution board with a 5 V rail, for instance a Eurorack style ±12 V, +5 V supply and 16 pin power headers, but making sure the PSU can supply enough current on the 5 V rail for your modules’ needs, perhaps using something like a 7805 rather than a 78L05. I don’t know how many such supplies you’ll find out there. An ATX supply would give lots of current on 5 V and +12 V but very much less on -12 V; it’ll also probably be noisy and not so well regulated. Another possibility would be to use a separate supply for the 5 V rail. There probably are quiet, well regulated, high current 5 V supplies out there, but I don’t know where.

And I guess you could build a supply along the lines of an FC power but using a 7805 instead of 78L05, or a 7805-based 5V supply to run off your +12 V supply. Start from the datasheet designs, I guess.

Added: Keep in mind nothing’s free — if you’re drawing 500 mA off a 5 V regulator that’s powered by the +12 V rail then that’s 500 mA less available on the +12 V rail. Actually, maybe more like 1 A less, taking into account regulator efficiency. That’s why I struck out that paragraph above, it doesn’t make much sense to use a 7812 to power a 7805. Better to design a supply around a 3-output transformer, or use an entirely separate 5 V supply.


And as soon as I said that, this came up:

A distribution board with a 7805-based 5 V supply powered off the +12 V rail. There are smaller versions that were available already, I hadn’t seen them.

Maybe you can make a case for this. Presumably you at least can get more than 100 mA — though not many times that — on the 5 V rail while still getting some hundreds of mA on the +12 V. Or of course the upstream ±12 V supply might be something that’ll provide more than 1.5 A. (But built as shown there’s no heat sink for the 7805, so I’d think it’d still be pretty limited.)


Thank you for all the help! I predict making a handful of digital modules and understand that there’s a lot of power wasted when using a 5v voltage regulator off 12v power. That’s my primary concern…

But is the energy wasted by having a few modules with their own regulator negligible in comparison with one that all modules are drawing from? If that’s the case I’d feel a little less hesitant about using one per digital module and eschewing the whole 5v rail idea.

I fully intend on reading the Art of Electronics when the time is available to answer these beginner questions myself, otherwise I’ll find some sort of hand holdy interactive online resource haha

I’m kind of guessing here… but wouldn’t it be a wash? The “energy wasted” is the difference between the amount of electrical power going into the 5 V regulator and the amount coming out: If you have an amp at 12 V going in and an amp at 5 V going out (in fact the current out will be slightly less than the current in, that’s the current the regulator itself uses) then that’s 12 W power in and 5 W out, so 7 W is lost (dissipated as heat). Whether that happens in one regulator supplying 1 A or five supplying 200 mA it seems to me would make little difference. Maybe the five regulators would use more total current than the one, but I don’t think that’s a huge factor.

On the other hand if the 5 V regulator were running not off a 12 V rail, but off a separate transformer coil supplying say 5 VAC, then that would be much more efficient.


Easiest solution. lol.


can’t lie, I was watching something very similar last night.

however as someone who still closes my eyes tight when plugging in a new DIY module, the YouTuber’s fork-in-socket logo gives pause.


10 µF on input and output?! The datasheet calls for 330 nF and 100 nF.

1 Like

And even that isn’t really necessary:

It is not necessary to bypass the output, although this does improve transient response. Input bypassing is needed only if the regulator is located far from the filter capacitor of the power supply