1
\$\begingroup\$

I'm trying to power a stepper motor with the 5V rail of an ATX PSU and control it using arduino. To keep the PSU running, I have shorted the PS_ON pin to ground. Also, I have connected PSU ground to arduino ground. The arduino is powered over USB.

The problem I'm having is that the PSU turns off automatically when I upload a new sketch to arduino. I can reproduce the problem with these steps:

  1. Turn on power (PSU is running fine)
  2. Set one motor pin (e.g. digital pin 9) to HIGH
  3. Upload a new sketch to arduino (doesn't matter which one, e.g. the Blink example)

-> PSU turns itself off.

Why does this this happen and what can I do to prevent it? The PSU seems to perform well otherwise.

schematic

simulate this circuit – Schematic created using CircuitLab

Edit: I just found these alternative steps to reproduce:

  1. Power on
  2. Set one motor pin to HIGH
  3. Disconnect 5V rail of PSU from motor driver

-> PSU turns off

Further tests:

  1. Disconnect motor from driver
  2. Power on
  3. Set one motor pin to HIGH
  4. Upload new sketch

-> PSU stays on!

  1. Connect motor to driver
  2. Power on
  3. Set one motor pin to HIGH
  4. Disconnect motor from driver

-> PSU turns off

It seems to me that the PSU doesn't like losing load on 5V after it has seen load once.

\$\endgroup\$
12
  • \$\begingroup\$ Welcome to EE.SE! Inrush current to said motor being higher than your power supply can supply without tripping OCP? \$\endgroup\$
    – winny
    Commented May 24, 2020 at 11:40
  • \$\begingroup\$ Most modern ATX PSU's are 12V centric. Might require some dummy load on the 12V rail to keep it happy. \$\endgroup\$ Commented May 24, 2020 at 11:45
  • \$\begingroup\$ @winny Mhh, the stepper driver only draws 130mA with one pin HIGH. Multimeter doesn't show high current when uploading a sketch, but maybe too fast to measure? However, I just found new steps to reproduce when disconnecting the driver (see post), which seems to indicate against inrush current? \$\endgroup\$ Commented May 24, 2020 at 12:03
  • \$\begingroup\$ @Unimportant I might try that, although the PSU runs fine when no load is connected on power on. The problem only occurs when transitioning from load to no load. \$\endgroup\$ Commented May 24, 2020 at 12:05
  • 1
    \$\begingroup\$ @Reroute Not quite, during reset all IO ports go immediately to their default state which is floating - inputs with pull-up resistors disabled. \$\endgroup\$ Commented May 24, 2020 at 12:38

1 Answer 1

2
\$\begingroup\$

As multiple comments have said, most ATX supplies require a minimum load on the 5V line for proper regulation, many tutorials use ~10 Ohm resistor or smaller on the 5V rail.

This is due to the type of SMPS topology, typically an LLC resonant converter, that requires a minimum load for tuning of the resonant circuit within a given frequency range.

Solution: Try putting a constant load on the 5V line, and your issue may go away.

\$\endgroup\$
1
  • \$\begingroup\$ Yep! Added a 10 Ohm 25W resistor on another 5V rail and now the PSU stays on. \$\endgroup\$ Commented May 26, 2020 at 11:09

Not the answer you're looking for? Browse other questions tagged or ask your own question.