CFA633 and ATX power not working

[PinkFreud]

I have an old CFA633 (v1.5a) that I've repurposed for a new project. It was configured to operate in ATX power mode for an older motherboard (Supermicro P4SCI), and it appeared to work.

Since upgrading the board, I've been unable to replicate this.

With the power and reset wires connected to the motherboard, the moment I connect power, the system powers up. I can hold down the X key until the module reads POWER OFF, but the system is never shut down, and the CFA633 stays in this state until a reset is issued by holding down the check button. The reset function resets both the CFA633 and the system, as expected.

Jumper jp8 is open and jp9 is closed, as per the data sheet. According to the output returned by command 30, the bits KEYPAD_POWER_OFF, KEYPAS_POWER_ON, KEYPAD_RESET, LCD_OFF_IF_HOST_IS_OFF, and POWER_INVERT are active (RESET_INVERT is not). I've tried changing the setting of POWER_INVERT to no avail - no matter what, I cannot power this system off without pulling the power cable. Even a shutdown from the OS (Linux) sees the power turn off briefly... and then come back on. When connected to a normal switch, powering off behaves as expected.

Any ideas on what I'm doing wrong?

Crystalfontz LCD Blog

Looking for additional LCD resources? Check out our LCD blog for the latest developments in LCD technology.

 

[PinkFreud]

Problem resolved!

The lack of power off was due to the POWER_INVERT (0x04) option being set incorrectly for this board. I did notice that once I disabled that, power off would work... except the machine would power on a moment later.
As it turns out, I also needed to choose a better power source for the +5V wire. I wound up connecting it to one of the +5V pins on an internal USB header, and enabling USB wakeup in the PC's BIOS (thus allowing the board to supply power to USB while powered off). Using +5V from the Speaker connector (on the front panel header) or the internal SPDIF connector caused the odd power bounce I was seeing.

 

[CF Support2]

Completely missed your post -- our apologies!

You did basically what I was going to walk you through. We've seen with some of the newer (z77 and newer) chipsets, in particular server system boards, is that the USB is very picky about what's connected to it and pulling the +5VSB from it causes the system to not shut down. Inverting the power pulse is dependent on the system board. Most SuperMicro boards tend to like the pulse inverted (held low, pulsed high).

Let me know if you have any questions.

Thank you.