NF4 Expert will not cold boot- psu issue?

[scottR]

READ and UNDERSTAND The Important Info Here Before Continuing With The Thread!

 

Some time has elapsed since this issue was first seen. It has been determined that the Expert board does indeed have a higher current draw than previous boards. At the same time, it has been determined that many new nF4 motherboards have the same high current draw which means the issue is not just with the DFI Expert board.

 

Regardless of the current draw by the Expert board, the same issue has been experienced on dozens of other boards. Units from ABit, ASUS, MSI and Tyan all experience the same issue if you have them running on the newer "high power" PSUs.

 

Replacing the PSU resolves the issue for some users.

 

The big issues are below...

 

First - Having the 3.3V, 5V and 12V rails balanced. It has been shown in testing that if you don't have a proper load on 12V rail you can't reach maximum current on the 3.3V or 5V rails. Additionally, if you don't balance the load across the 12V rails you can also cause the PSU to fail on startup.

 

Second - These new "high power" units have only been available for two years or less. There are still some growing pains involved.

 

Third - The higher current draw motherboards seem to be ahead of what the power supply manufacturer can handle.

 

Conclusions...

 

The Expert board can draw more than the rated 2.0Amps on the 5VSB line. But! If the Expert board was the only point of failure, no power supply would run it properly.

 

We've got hundreds of Expert board members in this forum that are not having a single issue with their rig.

 

I've built dozens of Expert boards and only had one unit that wouldn't start properly with a PSU. I simply put another PSU in place and the rig has been running fine for weeks.

 

Part Two...

 

It has been determined that there are issues with certain power supplies failing at startup that relate to "switch bounce". When a PSU starts a sequence of events must happen in a timed progression.

 

Normally, the PSU senses a voltage on the start circuit. When you push the start button on your case or motherboard, the voltage is drawn down to 0.0V. This condition must last for at least 250ms

 

"Switch bounce" occurs when the PSU has trouble latching into the PS_ON# state as dictated by the ATX 12V 2.X specification. The internal circuitry should allow for a certain amount of oscillation in the "start circuit". If this isn't accounted for, the PSU will attempt to start then shutdown.

 

It has been determined that the addition of a 0.22uF capacitor across the start circuit can smooth out the oscillation and allow the PSU to start properly.

 

The capacitor can be added anywhere on the start circuit from the case front panel all the way to the inside of the PSU.

 

I now return you to the original post.

 

ExRoadie

-----------------------------------------------

UPDATE 6/4/2006

 

If you do not read and understand this message and post a similar complaint that has already been covered, your post will be deleted without comment.

 

Once again for clarity. READ AND UNDERSTAND THE ABOVE INFORMATION BEFORE POSTING.

 

Do not waste your time or our time repeating the same issues.

 

ExRoadie

 

 

 

Hey guys, great forum here!

 

Interesting problem with my new NF4 expert setup- see my sig for specs. Initially the system powered up and I was able to enter the bios on the first try. I did some bios adjustments (date, time, usb mouse support etc.), rebooted then manually powered off to hook up peripherals and get ready to load an OS. When I attempted to turn it back on the system powered up briefly, just long enough for the diagnostic leds to light up and fans to spin for about 1-2 seconds, then it powered off on its own.

 

The system will continue to do this unless I disconnect the power cord from the psu, then turn the rocker switch off and back on. When I pull the cord and hit the switch I can hear a faint "leaking" sound (high pitched lasting about 2 seconds). Put the plug back in and it powers up just fine.

 

The system seems stable once it is up- I haven't tried gaming or otherwise power intensive use yet. Any suggestions on how to troubleshoot this problem? Do I have a bad psu?

[wevsspot]

If you have a digital multimeter you can use it to check your voltage rails. They should be within the 5% tolerance for each rail. Check voltages at idle and at load. Since you have an OCZ Powerstream 520W I'm assuming the idiot lights on the back of the PSU are all green at all times?

[scottR]

Yes, all psu leds are green when the unit is powered on. Unfortunately my multimeter lcd is messed up- I need to get another one to go further.

 

Any other ideas?

[wevsspot]

Are you trying to boot with any USB devices attached? Are you booting with minimum hardware installed?

 

Try a boot with only the following;

 

1. CPU (of course)

2. One stick of RAM

3. Video Card

4. PS2 mouse and keyboard

5. One optical drive

 

Once you are down to a minimum of hardware, boot and see if you power up and get into BIOS. If you get that far load optimized defaults, save and exit. Let me know how that goes. If you're still no go, try full CMOS clear and try again.

 

Be sure to disable all USB support in your BIOS for now.

[ILuv3ggs]

don't take the CMOs battery out, use the jumper ... i learnt that the hard way >

[scottR]

OK, tried booting with only ps2 keyboard/ mouse, 1 stick of ram and video card. Nothing else attached, loaded optimal settings in BIOS. When I power off the system with the case switch it will not cold boot again unless I flip the psu switch to off for a few seconds.

 

This all seems to be related to how the system powers down- if I shut down from windows and try to cold boot, same problem and solution. My guess is that either the psu has issues and needs to discharge (leaking sound when switched off manually) before recharging to power up or the board needs to lose power before allowing a cold start.

 

Board or psu?

[Cyberparasite]

i would say it's the board... you are not the first person to have this very same issue.

[scottR]

Does this mean RMA the board or is this a bios fix?

[Undesirable]

I had the exact same problem and solved it by purchasing a Hiper Type R HPU-4K580 power supply unit.

[scottR]

Hmmm, board, power supply, ram?

 

I've done some reading on this issue- cold boot bug seems to be related to ram voltage? Do I need to try different ram voltage settings or should I try to flash the bios to 12/07 and see if things change (and void my warranty).

 

Maybe I should rma the power supply, but surely the highly "recommended" OCZ power supply is sufficient. I really do not want to drop another $100 plus on another power supply just to see...

 

Undesirable, which psu were you using before?

[Cyberparasite]

actually i did see it mentioned by bigtoe that certain memory is very sensitive to voltages and just a little high or low either way would cause errors.

[Undesirable]

The PSU I was using before was the Enermax Liberty 620W. It was the only thing I replaced, and the only thing that fixed the problem. The rest of my specs are in my sig.