Power Supply (for Arduino clone and other G2 electronics)

I don't quite understand how all the Generation 2 boards are powered.

I have a classic ATX power supply that I want to use, and I'm building my Arduino from this [itp.nyu.edu].

The doc [reprap.org] says "You generally only need to connect a wire from black (ground) and yellow (+12V) to the RepRap."

This doesn't quite make sense to me, as the wiring for the power (Step 1 in the PhysComp tutorial) has a 5VDC regulator (7805, I'm using LM78L05ACZX), which (from my limited electronics knowledge) would take the 12VDC power supply and output 5VDC. As far as I know, this would waste energy as heat (perhaps requiring a heatsink and/or fan).

Why wouldn't I just use the red +5V wires for power instead of the yellow +12V wires? In fact, why wouldn't I just power the board from the USB connection? That does appear to be how the board is wired, so why do I need the +12V yellow wires at all (and thus why do I need to connect the power supply [other than directly to the motor boards, etc])?

I'm sure there's a simple explanation, and I just haven't found it yet. What is it?

Quote

I'm sure there's a simple explanation, and I just haven't found it yet. What is it?

I think its a plan to allow the reprap to be powered from a 12v only source, like a car battery. There's nothing to stop you making use of existing 5v lines if you have them though.

Ru Wrote:
-------------------------------------------------------
>
> I think its a plan to allow the reprap to be
> powered from a 12v only source, like a car
> battery.
>
Yes, it is. The notion is that Repraps in the third world ought to be able to be powered by standard automobile batteries.

Ru Wrote:
-------------------------------------------------------
> I think its a plan to allow the reprap to be
> powered from a 12v only source, like a car
> battery.
Oh. How... thoughtful?
> There's nothing to stop you making use of
> existing 5v lines if you have them though.
Right. The ATX 20 pin Molex connector has 4 +5VDC wires/pins (see the diagram here [reprap.org]), and the +5VDC rail(?) on my power supply provides twice the current as the +12VDC rail(?) (about 83% the power). I have to ground pin 14 (the green on-switch wire) anyway, so it could be cool to yank a male 20 pin Molex connector (female housing) from an old motherboard, slap it on my Arduino, and then wire it from there.

Although, I do wonder if I can just power the entire Arduino board from the USB power line. Would there be anything wrong with totally skipping Step 1 in the tutorial [itp.nyu.edu] (though I might add the power indicator LED)? They do specifically say
"These capacitors and power regulator aren't strictly necessary if you're only ever going to power the board from the USB port."
but I want to make sure this applies to our (RepRap) use (though I don't see any reason why it wouldn't).

Now that I think about it, could wiring up a power supply to the board cause harm? Based on the pictures in the tutorial, the described setup is already USB-powered. If I added the power from the power supply (and thus if I used both power sources at once), could it fry something? Or would it make no difference?

Quote

If I added the power from the power supply (and thus if I used both power sources at once), could it fry something? Or would it make no difference?

The arduino diecimila has a jumper to select USB or other power source. I power mine solely from USB though; the alternate supply is only really useful if you want the system to run whilst disconnected from a PC. Which isn't practical for a reprap right now.

But when you say you want to wire the arduino directly to an ATX supply, what would you gain? You still have to run power lines from your controller to the peripheral boards, only now you've added an extra step with all the possibilities for error and breakage that implies. Have I missed something important here?

Ru Wrote:
-------------------------------------------------------
> The arduino diecimila has a jumper to select USB
> or other power source. I power mine solely from
> USB though; the alternate supply is only really
> useful if you want the system to run whilst
> disconnected from a PC. Which isn't practical for
> a reprap right now.
Alright, good to know.

> But when you say you want to wire the arduino
> directly to an ATX supply, what would you gain?
> You still have to run power lines from your
> controller to the peripheral boards, only now
> you've added an extra step with all the
> possibilities for error and breakage that implies.
> Have I missed something important here?
Actually I would prefer to run it off the USB. I'm guessing you're asking about the 20 pin molex connected to the Arduino board. I would probably not use the pins/wires on this main connector to power the other boards/motors/etc, I would probably use the 4 pin peripheral connectors (or something).

Anyway, using a computer power supply wasn't my idea, I just thought it was the standard RepRap power source.

Quote

Anyway, using a computer power supply wasn't my idea, I just thought it was the standard RepRap power source.

Oh, it is. I just wondered why you'd want to attach a big powerlead to the arduino. Which you apparently don't want to do anyway. So my confusion is resolved

I'm powering my Arduino off USB for now, and it works well. One thing I noticed though, different computers have different noise levels on their USB 5V. Since my thermocouple chip is powered from the Arduino, I get a couple more deg of temp jitter from certain laptops. Not a real problem though.

A problem I have noticed with laptops is that the mains adaptor is connected by a two pin plug and a length of 2 core cable. At the PSU end of the cable the 0V is connected to mains earth for safety reasons. The current drawn by the laptop drops voltage over the cable so the 0V of the laptop is noisy and is not at mains earth potential. It causes problems if you have a USB oscilloscope connected to a laptop!

They should really have a cable with three wires, one for 0V and another for earth, connected together at the laptop end. A twin screened cable would be ideal.

The 5V feed on the USB connector is intended for powering 3.3V systems with an LDO regulator. It is not really meant for powering 5V systems because if you took the maximum 0.5A and used the maximum cable length of 5M you would not have 5V at the far end.

---

[www.hydraraptor.blogspot.com]

Well,
considering what I have seen from old dissassembled laptops I used to have laying around...
All ground_routes are connected to the Metalparts of the case.
Connect that to some earth near and you should reduce noises a lot.

Oh You can (if you have) screw the cable to the serial or parallelport screwterminals, or you can simply use the kensington-lock if you have one (who hasn't)
Or, If you don't have that too, you may even unscrew the original screw of the RAM or ADDON Cover and Put your earthline there (Just one old HP Omnibook has threaded inserts that are not connected to the ground of the mainboard)
But you may find some other metalparts there, that will work.

'sid

Yes but whatever you connect it to earth with will carry some of the current powering the laptop and hence drop voltage. This is several amps and it is very noisy.

You would need to earth the case and then disconnect the mains earth to the power adapter to prevent the PSU current flowing in the earth connection. I would not advise that on safety grounds. It should be done by the PSU cable having ground and 0V separate. Analogous to mains leads which always have neutral and earth separate.

---

[www.hydraraptor.blogspot.com]

Wow, when I hooked my scope today (an ancient analog freebie) I saw a 48 V 60 Hz sine wave signal between my Arduino and the scope, which drowned out the 5V pwm signal I was trying to read. I have not been using the ground connections on the old scope, as I found it was causing problems with the Arduino, and now I just figured out why - that's a pretty serious signal there.

This is on an Asus Eeepc, with a 2 prong power plug. Interestingly, it works fine when I have the Eeepc connected to an external monitor, which is how I've often been using it, and why I was able to use the scope without problems up to now. The monitor cable appears to ground the Eeepc chassis. The Kensington lock hole doubles as the screw terminal for the monitor cable, so Sid's idea would be similar.

So yeah, a proper grounding system would be a good thing, especially if you're using a laptop with an Arduino.

Wade

nophead, that's dropping voltage by connecting it to an external monitor too;
honestly I don't think that's the case.

Next time I hook up my laptop to an external monitor, I'll measure the voltage of the usb port before and after

'sid

Sid,
Wade is using a two pin mains adapter, so what I said does not apply in that case. The output of the adapter is floating so earthing the Laptop end will solve the problem.

It is adapters with 3 pin mains plugs that connect mains earth to 0V at the adapter end that are the big problem. Earthing the laptop case will reduce the noise, but to eliminate it would have to be a very low impedance connection compared to the Laptop's power cable.

When this sort of adapter is used with an earthed monitor you can get some of the ground current flowing in the monitor cable leading to patterning on the screen.

---

[www.hydraraptor.blogspot.com]