ATX PSU for Mendel90, I need help!

Im trying to build a Mendel90 from Nophead's newest revision. His latest revision uses 500w ATX PSU. On the pdf instructions from [github.com] , I see that he uses Tyco THS 15 10R and THS 15 4R7 Aluminium clad resistors.

I have no idea what the numbers means as I don't have an engineer background. I found this: [passives.te.com], from Tyco's website but it wasn't that much useful to me.

Other reprap models like Pursa i3 that uses ATX PSUs don't seem to use resistors, is it necessary to even use a resistor for a 500w ATX PSU?

Can I use something in place of a resistor like a 15w light bulb or fan or something? Also, don't we just use the green wire to short a black wire on the ATX power connector without a resistor like this: ?

Here is a picture of the PSU setup from Nophead's instruction.
I have tried to search online for the resisters, but it is hard when you don't have an engineering/electronic background.

Some power supplies require a resistor on the 5V line to get stable 12V output. Not all do. The few ATX supplies that I've used didn't require it, only requiring green and black wires be connected. Try wiring up your power supply without the 5V rail loaded, and check the 12V line while your printer is running to make sure that it maintains 12V.

The power resistors that he uses are 15 Watt rated 10 Ohm and 4.7 Ohm, like this:[www.digikey.com] and [www.digikey.com]

Edited 2 time(s). Last edit at 12/26/2012 11:34PM by NewPerfection.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

Thanks NewPerfection, you've saved me from some confusion. I actually don't have the 500w ATX yet, but rather I have a spare 400w ATX. Do you think I can use the 400w instead? From my understanding, I just need something that can deliver at least 18 amps to the electronics right?

mtvan85

Power and current are related by the equation P=IV, so if you're trying to see if 400W is enough to provide 18A at 12V, you simply multiply 18 by 12 = 216W

So yes, a 400W supply would be more than sufficient.

Edited 1 time(s). Last edit at 12/27/2012 03:46AM by marcosscriven.

@marcosscriven

That's not that simple, stated power is for whole power supply and there are usually further limitations for given rails. For example 300W cheap PSU that I have currently at hand is rated only 15A at 12V rail (180W).

I can't imagine a 400W PSU not being able to provide 200W on the 12V rail?

Modern ones have a lot of the power allocated to the 12V rail. Old ones had a lot on the 5V rail.

---

[www.hydraraptor.blogspot.com]

marcosscriven Wrote:
-------------------------------------------------------
> I can't imagine a 400W PSU not being able to
> provide 200W on the 12V rail?

It might seem crazy, but PC PSUs are designed to split output power over several voltages, so it would not be a surprise to see half the power on the 12V.

Here is a spec taken at random:

Technical Specification
Model: iCute 400
DC Output
  +3.3V   +5V    +12V     -12V    +5VSB
    23A     35A　　18A     0.5A    1.5A　
Wattage max   400W

So that is 216W on 12V. Tests of cheaper power supplies show that they can rarely meet their own spec under all conditions, some of them don't meet their own spec under any conditions! Pushing a PSU to its limit only works reliably on the more expensive PSUs. Quality varies a lot though, some cheap PSUs may cope.

Edited 1 time(s). Last edit at 12/27/2012 07:00AM by bobc.

Yes its amazing how PC PSU get away with fake ratings. To get CE approval it must pass the EN60950 safety standard. That requires a rating label and the test house will load it to the values stated on the label and measure the temperature of important components. Most PC PSUs blow up if you load them to half their ratings, despite being CE marked, UL, etc. I can only assume they all have fraudulent approvals.

I have known PSUs fail EN60950 even when the component temperatures stay within spec but not have enough margin for the measurement uncertainty of the test house. That is how picky a genuine test house is. No way would they pass 99% of PC PSUs.

---

[www.hydraraptor.blogspot.com]

Quote
mtvan85
Other reprap models like Pursa i3 that uses ATX PSUs don't seem to use resistors, is it necessary to even use a resistor for a 500w ATX PSU?

I am interested in this too. Here a power supply the seller says he uses on Prusa has resistors: Power supply

Quote
mtvan85
I have tried to search online for the resisters, but it is hard when you don't have an engineering/electronic background.

I'm in the same boat as you. Zero electronic background so am learning a lot.

I found these resistors on eBay:
4.7 Ohm 4.7R 25W Watt Power Metal Shell Case Wirewound Resistor
10 Ohm 10R 25W Watt Power Metal Shell Case Wirewound Resistor

These are 25 WATT rather that 15 WATT. (Funny I could not easily find similar resistors in 15 WATT.) Does that mean they are or are not useable for the PSU mod? If they are usable is there any worry about quality of these eBay resistors?

The power rating on a resistor tells you how much power it can dissipate under normal conditions without overheating and failing. More is better.

Its easy to calculate the power rating we actually need - 5V across a 10Ohm resistor gives you 0.5A of current. 0.5A @ 5V gives you 2.5W. So 15W is overkill, 25W is overkill, but it'll work, and the larger resistors dissipate heat more effectively which means the dummy load will be cooler, a 2.5W resistor would get hot enough to burn you.

To tell if you need resistors at all .. build it without, if you get 12V out of it while the PSU is loaded, then you don't need resistors.