Using an OEM power supply

Question

I am trying to use an 12V OEM type power supply to supply 12V to stepper motors through an L293 h bridge in an Arduino powered circuit but my current setup starts burning all the negative wires and almost killed my pc when I plugged the Arduino in the USB port. I am not sure what could be causing this. I have connected my Live, Neutral and Ground wires correctly and I am then using the "adj. +V" pins as my motor voltage supply and I have connected the "-V" to the common rail/ground of my circuit. The power supply does seem to work fine and if I put an LED with a resistor between "adj. +V" and "-V" nothing strange happens and the LED goes on just as normal. I should also mentioned that the circuit where the ground wires fried has been used hundreds of times with a PC power supply and gave no trouble at all.

I am guessing that the OEM setup is probably the problem here. I have found this page http://reprap.org/wiki/RepRapPro_Mendel_power_supply indicating that I should connect the ground on mains to the -V as to earth my circuit. This could possibly explain my trouble as it means that once I plug in the Arduino in it seeks earth through my PC, or is that just absurd? Even if that is the case, following the above tutorial I have come across an issue, in the testing segment they state that there should be no resistance between ground and -V and this is indeed the case once I have done as stated in the beginning of the tutorial and connected ground and -V. The troubling thing is though that there is also almost no resistance between +V and the ground pin, even without me making the aforementioned connection. I know they don't state this as something to test or watch out for but my head tells me this means that +V and -V are then connected and making a short and that that is something very dangerous, I am therefore to afraid to test the power supply after connecting ground and -V.

Do any of you have experience with these power supplies and know if this is normal or have had the issue before and might know of a possible solution?

Here is photo of the terminals without a connection between GND and -V:

EDIT: I have now removed all the wires and there is still a connection between +V and earth.

EDIT: The circuit basically entails that firstly all the ground pins meaning the ground on the Arduino and the -V on the power supply and the ground pins of the h bridge are all connected on a common rail. The VCC on the h bridges are then connected to the 5V on the Arduino and the +V on the PSU is only connected to the motor voltage pin on the h bridge. But as I previously said, this circuit works fine with a pc power supply. Also, for some reason only the black wires connecting the ground pins were burned, the 12V and 5V wires are all fine.

Answer 1

If your PSU looks like this 12 Volt (single voltage output), 30 amp power supply:

• The adj label probably refers to an adjustment screw, not to any of the terminals.
• The -V label does not mean -12V, it is equivalent to COM on the PSU in your reprap link.
• I would not expect the ⏚ terminal to have a low resistance to the +V terminal with all cables disconnected.

Answer 2

The following earthing configurations are quite common in DC power supplies:

• Positive output (i.e. +12V): low-side connected to earth
• Negative output (i.e -54V): high-side connected to earth

Note that in both of these cases it's the reference rail that's earthed. A +12V supply is +12V with regard to the earthed rail; a -54V supply is -54V with regard to the earth rail. In both cases, the 'zero' rail is the earthed rail.

Most(if not all) ATX PC power supplies earth the reference rail, so when you connect your PC peripheral, if the positive rail of the OEM power supply is indeed earthed, you are shorting the OEM power supply rail via earth, and the connections going back to the PC are the ones that smoke (the wires aren't rated for heavy current and burn).

You shouldn't fear a short circuit on a safety-approved power supply. Part of the safety testing done on the design will prove that it will protect itself without a hazard - units will either latch off safely or hiccup in a low-energy mode indefinitely, or if something does fail, it will not cause a shock hazard or spew debris everywhere.

You really should test the power supply per the wiki entry (with V- earthed) since the worst that could happen is the power supply will protect itself, and if the power supply does deliver output, your observation about earthing is incorrect.

Answer 3

Ok so the reason that there is low resistance between +V and ground is merely because there is a capacitor between them and my multimeter was charging it. The reason for only the negative wires frying then was probably because the ground on the arduino/pc had a much lower potential than of the PSU and a very large amount of current then then went from PSU GND to PC/Arduino gnd which was at earth level.