M1-ATX power on "feature" discovered

**JimmyFitz** · 01-08-2006, 01:58 PM

I have been running my carputer on the desk inside for a couple of weeks now using a 12DC power brick feeding to an M1-ATX. I have a switch wired in that setup to simulate the ignition power for the car to signal PC startup/shutdown and I also hooked up the power button on the front of the case and all works well from either switch. The M1 jumpers are set to the suggested 5sec/2hour mode. Yesterday, I started putting wiring in my car to run the system there. I connected the constant 12V feed but I haven't hooked up the ignition signal yet. I tried turning on the PC with the switch on the case and get nothing. I tested the wiring with a meter and everything seems fine so I bring the PC back to my desk. It works flawlessly there just as before. After further experimenting I found that I could duplicate the problem at my desk.

Here is my discovery: the FIRST time the PC is connected to the 12V source (either the car battery or the brick), the only way to power up is by signaling the M1-ATX turn on lead. The button on the case doesn't do anything until after the M1-ATX has been turned on at least once. After I have the PC on the first time, the button on the case functions normally. The manual doesn't provide any details about this "feature". I hope this discovery will help keep someone else from struggling with this.

**DarquePervert** · 01-08-2006, 03:05 PM

Good find. Someone should make this a sticky int he Power Supplies forum, perhaps with a more descriptive title.

**evandude** · 01-08-2006, 03:31 PM

when the power button on the case isn't working, most likely the power supply has not turned on the +5V standby line. Sounds like the M1 doesn't enable the line by default, until it has turned the PC on and off once by itself... sounds like a valid bug, I can't see any reason that would be desirable. given the nature of it (that it would only be noticed when you FIRST connected power to the PC, and even then only when you used the power button, which isn't used as often) I can see how the oversight in programming could slip by unnoticed...

it's a good thing that you had the patience to do some testing. I can imagine that if someone hooked up an M1 to their system for the first time and pushed the power button and it didn't turn on, they might freak out a bit... wonder if there's ever been any "defective" M1's unknowingly returned for that reason