Getting a steady 5v from a PSU... Molex??

Soldato
Joined
11 Oct 2011
Posts
3,334
Location
UK
Hi all,

As above, to power an Arduino Yun microcontroller for an electonics/programming project I would like to get a micro USB cable connected to a PSU

Or a direct cable coming from the PSU (with ground) that supplies a regulated 5V (must be regulated if I'm not going down the micro USB route)

What are your ideas please?

Is the molex power cable (5V one, not the 12V) usually regulated at 5V, or is it 'around' 5V? Or does this all depend on PSU? I was intending on using a cheap one to keep costs down

Thanks!

Extra info: If it seems odd to be using a computer power supply to be powering this... The reason is a PSU will already be involved for powering PC case fans for the project, so ideally with this solution the whole thing can run from one 3 pin mains outlet.
 
Last edited:
Sorry I should have explained

There won't be a motherboard as I'm only using a computer PSU for powering some case fans for the project (which doesn't involve a PC at all)

Just using the PSU for 12v and 5v supply for the electronics project
 
What sort of tolerances do you need on the voltages? I'd say on generic power supplies, you'd be looking at 5% tolerances and tighter still on more expensive power supplies.
 
Probably no tighter than you'd get from a standard 3 pin to USB adaptor or from a USB port

I'm not sure how much variance the arduino will tolerate

I think I'm correct in saying that a PSU more tightly regulates the 5v if it has a bit of load on it? Someone suggested a 10k ohm 10 watt resistor to add a bit of load and keep the PSU running (as some require a load)
 
10k ohm at 5v is next to no load at all, just 0.5mA. I'm not sure what a PSU requires to go into decent regulation though but expect something like 0.5mA to be a tad low. Also a 10W resistor would be overkill for a 5v potential. w
 
Any ideas on how I can check the regulation?

I mean, with just a multimeter attached the voltage is 100% steady, it doesn't change at all.... That's with no load at all though, not sure if it will change under load
 
An oscilloscope would be handy.

I would expect a PSU to be pretty good at regulation but if I were concerned I'd use a 500ohm and a led across the 5v. It won't be efficient but it should give 5v.

If it is regulating at no external load then I wouldn't bother messing further. If the supply is anywhere decent it'll stay within a few % of 5v with a minimum of ripple as demand varies. By the sound of it your demand will be low so you should not see issues.
 
Thanks for the response

As above, and as you've confirmed, I think I'm going to go with it as it is... If it isn't broken, don't try and fix it :D

It seems to be steady as I said without any load at all. The figures I'm getting are:

12.26v on the 12v line
5.12v on the 5v line
3.4v on the 3.3v line

Is 5.12v instead of 5v likely to cause any damage to the arduino microcontroller board? It says:

There is no built in 5V regulator. If you power your Yún with more than 5V, you will likely damage it. If you are not powering the Yún from the micro-USB connection, you can apply power to the VIN and 5V pins on the board. If using the VIN pin, you cannot provide more than 5V, it will damage your board. It is recommended to power the Yún from the USB connection whenever possible.

I'm guessing when it says 'more than 5v', it means considerably more... I don't think 5.12v will cause any damage, compared to 5v

Do you agree?

Thanks! :)
 
Most electronic components have a tolerance of 5% usually 10%. You can get basic 5V regulators pretty cheap or get a buck board i.e. something based on LM2596 (with an onboard voltmeter to reduce hassle) which will let you power it off either 12V or 5V rails with fairly tight regulation.

EDIT: Probably a bit pointless for 5V actually as the LM2596 wouldn't be providing any real advantage over the PSU's 5V regulation in terms of ripple, etc.
 
Last edited:
I edited my post above as I don't think it will provide any tangible benefits over using the 5V output from the PSU - fed 5V from the PSU the output will be pretty much the same albeit with maybe a tiny bit more fine control over the voltage.
 
Ahh sorry I only just realised that

I kind of like the idea of being able to precisely control it. And it makes my project look that little bit cooler ;) haha

Are there any downsides to using the LM2596?

EDIT: Ahh just saw this, LM2596 might have no effect then:

Output Voltage: 1.25V~37V (continuously adjustable, the input voltage must be 1V higher than the output voltage)
 
So even the 5.12v will be better than the 12v brought down to 5v with the lm2596 then?

Just confirming

Thanks for all the help :) appreciate it
 
Better in the respect that the PC PSU will likely have a lot less ripple and be massively more efficient provided you have some decent load going through it. From the datasheet the LM2596 has an efficiency of around 72%, this would go on top of your PC PSU's efficiency at whatever load you have on it.

One possible benefit you might see on 5.12v vs 5.00v would be if you are using the Arduino's analogue outputs to measure something and you need the accuracy to be super high, otherwise who cares about 0.1v.

What sort of load on the 12v and 5v are you planning anyway?

If I needed 5v or 12v near my PC and it wasn't an audio circuit, then I would use my PC power supply. PC SMPS are fantastic compared with a crappy LM2596 circuit from China.
 
Back
Top Bottom