Z800 Boot Issues - Common Troubleshooting Isn't Working

[ember1205]

I have a few of these machines that I use as home lab servers, and I recently bought another one to add to the farm. When it arrived, it had a Quadro K600 graphics card, a single 2.4GHz CPU (X5620), and two 4GB RAM sticks (installed into CPU0 DIMM3 and DIMM5 slots). The 1TB hard drive has a Windows 10 install on it that I have no use for.

I pulled the hard drive, connected a VGA only capable monitor via a VGA/DVI adapter. I got no POST beeps from the machine and zero video output. I shut it down and slid the hard drive back in, tried again. Same results except that it did proceed on to load Windows 10 and I could see video at that point.

Shut it down and switched to a known good graphics card that would output standard VGA and I was able to see the HP Splash Screen and enter the BIOS and such. Pulled the hard drive back out and started to modify the system for my use.

I installed a SSD cage that supports up for four drives. I moved the on-board HD cables to the bottom-most row of connectors (SATA) and set up the on-board Serial-Attached SCSI / SATA connectors as follows:

- 0 - DVD ROM

- 1 - SSD drive 0

- 2 - SSD drive 1

I added a second X5620 CPU, swapped out the RAM for 12 8GB DIMM's of PCL3-10600 ECC. I swapped out the original heatsink for a high-performance one and added a high performance one to the second CPU. I proceeded to move two SSD's from a system that experienced some sort of failure and booted it up. This machine was now running ESXi off of SSD0 with SSD1 as a local data store for the VM disks.

On top of the base OS, I was running a Windows 10 Pro workstation and had just built out a linux machine to be the head-end for a media server (content will be offered via attached NFS drives from another machine). All ran well for about a week.

The system board is a Rev 3 board but had an older BIOS. I bought a matched pair of X5680 CPU's and swapped them in. The machine not only wouldn't boot, but I got pretty much nothing from it in terms of LED flashes or beeps. Admittedly, I had forgotten to update the BIOS prior to the CPU swap, but the result should have been errors from the board indicating unsupported CPU's.

I've swapped EVERYTHING back to its original form and the system does absolutely nothing. I've noticed during its current state when powering on that the memory fans will initially spin, but stop almost immediately. I have swapped every part (memory, memory fans, CPU, heatsink, graphics, power supply) with known-good ones and even moved some of these parts (memory fans, memory, heatsinks) into other machines and found them to be fully operational.

During all of my spart-swapping, I did seemingly discover that one of the high-power heatsinks has some sort of issue. If I install the heat sink and don't connect the power, the [good] system will boot but indicate low-power heatsink detected (which is a misnomer because it isn't powered at all - so it SHOULD say high-power heatsink NOT detected). If I connect the power, that [good] system will give similar no-output results.

In the bad system, removing the memory does not result in BIOS beeps but removing the CPU does. Under no condition now do the memory fans spin.

I haven't any idea where to go from here as it is seeming like this is a system board issue but nothing that I did should have caused an issue. Thoughts?

[DGroves]

some HP "Z" systems detect the presence of a high power cooler by the fact that the fan connector has a "5th" jumper wire attached low power coolers lack this wire jumper but the z800/z820/840 systems do not do this , .......they do however check the power supply and will report a error  if you have a 130 watt cpu with the stock 850 watt supply as i recall, same for the 150 watt cpu's  which HP requires the liquid cooler option and the 1150 watt supply

only the Rev xxxxxxxxx-003 boards ofically support xeon 56xx cpu's 

https://h30434.www3.hp.com/t5/Business-PCs-Workstations-and-Point-of-Sale-Systems/HP-Z800-CPU-overvi...

https://andybrown.me.uk/2014/11/01/z800/

If the boot block date is 01/30/2009, the system only supports the Intel 55xx processors. If the boot block date is 11/10/2009, the system supports both Intel 55xx and 56xx processors.

try removing the cmos battery and the power cord from the pwr supply for 30 min then press the pwr button for 10 sec  on the problem board, reconnect cmos batt/pwr cord

does the front light do anything upon power on?

[ember1205]

Thanks for the response.

I'm aware of the Rev 3 boards (-003) supporting the higher powered chips and how the jumper is used to determine that it's a high-power heatsink. I've used the high-power heatsinks on low power CPU's with the 850W PS without any issues (guessing that it cares about high power PSU's only for driving the CPU's as the additional fan draw should be fairly negligible in the grand scheme). In fact, I've used THESE high power heatsinks with an 850W PSU without issue. It wasn't until I tried using the 130W X5680 that the issues cropped up.

It IS possible that there's a problem with the CPU, too. The chip that's installed in the CPU1 slot is where the heatsink issue appeared. I -also- am getting checksum errors on DIMM3 and DIMM4 for that CPU and this is PC3-10600 ECC memory (8G sticks) that I have used without issue for about three years with X5620 CPU's.

If memory serves me correctly, I've had zero issues with this system since replacing that CMOS battery about six months ago with a new one out of a package. I may try backing up the CMOS settings, resetting it, pulling the battery, and trying again to see what happens overall, though.

[DGroves]

rev 003 boards have NOTHING to do with high power cpu's ALL board rev's will run 130 watt 55xx cpu's

YOU REQUIRE A Rev 003 Board when using a 56xx model CPU

[ember1205]

I did a TON of hardware swapping between a working system and this "broken" one combined with various spare parts... Ultimately, I was able to chase this down to bent CPU pins.

I used a high quality phone camera with a bright flashlight to zoom in and see the misaligned pins. I then used a SIM tool (used to pop the SIM trays out of a smart phone) to deflect the pins correctly and straighten them out. It took a gentle touch and some time, but I was able to re-align the pins correctly and get the system to boot correctly again.

I have been able to get the BIOS upgraded to the most current revision but have not yet attempted to get the higher speed CPU's working in this system as of yet (nor have I tried to confirm the fault I seemingly found with a high-power heatsink).