constant reboots, no post

[dark41]

Here's one for the brainiacs out there, and new to me.

I shut my system down normally the other night. It sat for a few hours before I tried to start it again. PSU fires up, fans spin up, CPU cooler lights up and spins a couple times, then the system shuts down. Nothing on the monitor, no beep codes, then the system starts again, same, thing, over and over. I let it run for 30 mins, but no go. Reset CMOS, tried again, same thing. I start ripping out components 1 by one, USB devices, all hard drives, DVD-RW, floppy, and Vid card - until its just MB, CPU, RAM and case switches. Same thing, no post, no beep codes.

Now I figures its the motherboard. So pull that out and replace it with a brand new P35-DS3P. Same thing. I replace the CPU with a new CPU (both E6750s). Same thing. I replace the Crucial Ballistix DDR2-800 with generic DDR2-800 and the system starts right up. I figure its licked. Set up Vista Ultimate on the system and do several reboots with no problems. Then I shut this down and go to bed. Next morning, same thing. No post, constant reboots, but after 10 or 20 it eventually starts. Reboots seem fine, but complete shut downs make the system constantly reboot until it eventually starts.

At this point I'm ready to pull what's left of my hair out. I put the old CPU in another system and it boots right up a couple dozen times, ruling out the CPU.

I put the old Crucial Ballistix in another system with an Intel board and got 3 beeps, which points to the memory. Problem is that this memory has never run at less than 2.1v and the Intel board has no setting for memory voltage. So not sure that really proves anything about the Crucial memory, but I find it hard to believe the brand new generic memory would be bad in the same way in the new P35-DS3P motherboard.

I even plug the entire system into an extension cord run into the garage to make sure we're not running too much from a single outlet (we have 3 computers and an all-in-one printer running from one outlet in the computer room normally).

Anyone ever seen a PSU go bad in this manner? I did try the DQ6 on an Antec PSU on the test bench with only CPU, RAM, and Vid card (no hard drive or optical drives) and same thing, just kept rebooting without post. But I spose the MB could now be shot from whatever caused it to go in the first place.

The only other thing I can think of would be the case (Thermaltake Kandalf) power switch itself, somehow shorting out or under-volting the boards? Anyone heard of such a thing before?

Any and all ideas are appreciated.

Original system:
Kandalf (18 mos old)
E6750 @ 3.6GHz (18 mos old)
Crucial Ballistix DDR2-800 @ 900MHz 4-4-4-12, 2.1v (6 mos old)
OCZ Powerstream 600w (18 mos old)
965P-DQ6 (18 mos old)
Leadtek PX7900GTX TDH Extreme (18 mos old)
6 hard drives, 2 x 2 HDs on RAID 0 arrays, 2 just by themselves, all on ICH9R controller (various ages from 1 to 3 yrs old)

Current system:
Kandalf (18 mos old)
E6750 @ stock settings (brand new)
Generic RAM (brand new)
OCZ Powerstream 600w (18 mos old)
P35-DS3P (brand new)
Leadtek PX7900GTX TDH Extreme (18 mos old)
Single hard drive with Vista (brand new)

 

[bobbknight]

Rule out the case power switch remove the case power switch connector from the motherboard header and then to turn on the computer touch the 2 contacts briefly with something like a screwdriver, if it powers right up it's the power sw. This kind of thing does happen.

Also look for bulging or leaking capacitors on the motherboard, if it's not the power sw this is the most likely culprit, with the first board.

I'm guessing that your system load is about 450 watts, for the first build, but:
Due to component aging a PSU will decrease in power capacity over time, 18 months is is just about the time for one to start having troubles.
I would try another PSU, of good quality and at the same or higher wattage.
If you can go into the bios and read the voltages see if they are within 15% of +3.3v, +5v, and +12v.

If you have bad caps on the motherboard I hope it's got a 2 year warranty, and if it does RMA it fast.
The same for the PSU.
As to the case it might be hard to get a replacement power sw.



Rattus Viola: Sino non they quisnam operor non have scientia futurus vestri rector.

 

[jsc]

The simplest way to rule out the case power switch is to swap it with the reset switch. That way, you are not poking in around the motherboard with a conductive metal object.

 

[dark41]

Jsc, never thought of that. Thanx.

MrsBytch, The first thing I did when I ran into problems was clear CMOS. That sets everything to stock speeds and voltages. Unfortunately the Ballistix won't post at stock settings (requires 2.1v), but even generic RAM wasn't working with stock settings.

Bobbknight, All Gigabyte boards have a 3 year warranty from the manufacturer. My supplier is testing the DQ6 at their shop. If it's faulty, I'll have a new board in 2 weeks or so. However from my latest tests I'm thinking the board is probably fine. Also, both of these boards have solid capacitors. That's why I prefer Gigabyte.

I've now put the DS3P into a new case with Speed 500w PSU, the same CPU, CPU cooler, DVDRW, single SATA HD, and 1 stick of generic RAM. Booted right up.

I upped the dram voltage and replaced generic with 1 stick of Crucial, also booted right up. I added the 2nd stick of Crucial and it rebooted several times before posting and booted up. Not sure what this proves as either the RAM or controller could be stuffed, or it could be a one off. So I'll be testing the RAM extensively, but since even the brand new generic RAM was doing this I still think something else is contributing to the problem.

After testing the RAM, and if either the Crucial or generic are found to be stable, I'll start installing one component at a time (fans, cables, HDs, RAID array, PSU, etc.) to see if anything else causes the problem. Fun fun.

Thanx for the replies.

 

[ZOldDude]

1) Case PWR switch is shorted.
2) OCZ PSU is bad.

Next time get a real PSU...PC Power & Cooling (by dirrect from them for lowest price).

 

[jeremyrailton]

same as what they all said. One additional thing to check: make sure the boards are seated in the case so that nothing can be shorting them out. maybe even try starting up outside the case. I had an asus board that was very sensitive to this type of thing once, had to be VERY careful about it touching anything.

 

[dark41]

Put a Corsair HX620w PSU, same problem. We can now safely rule out PSU. Seasonic for the win.

I posted initially "I did try the DQ6 on an Antec PSU on the test bench with only CPU, RAM, and Vid card (no hard drive or optical drives) and same thing, just kept rebooting without post." So grounding to the case is not the issue.

 

[rainbringer]

Hi,

Was the problem solved? What was the solution?
I've the same problem, have done most of the steps you've been through, except of changing CPU and PSU.
I've checked that no metallic pin shortens the mobo (which I've just bought after thinking the problem is in mobo), disconnected the reset switch, power switch etc... nothing helps.
I'm desperate! I'm not in a financial condition now to buy all new PC

Alex.

 

[dark41]

Hi,

Was the problem solved? What was the solution?
I've the same problem, have done most of the steps you've been through, except of changing CPU and PSU.
I've checked that no metallic pin shortens the mobo (which I've just bought after thinking the problem is in mobo), disconnected the reset switch, power switch etc... nothing helps.
I'm desperate! I'm not in a financial condition now to buy all new PC

Alex.

Hi Alex,
It was resolved. This was the 1st of many issues that I had with Crucial Ballistix memory and our Gigabyte motherboard systems. We're resellers for Crucial, and their support is 2nd to none (had all bad kits replaced within 5 days every (4 or 5) time), but unfortunately their memory doesn't work well with my systems. All my problems were resolved by switching to Kingston HyperX memory.