R730xd won't power on after shutdown

edit: it seems the statement about power on by idrac is wrong. i just tried this again and it doesnt work. it worked once but just directly after shutdown.

edit 2: in the system event log i receive the following event: "The system board DIMM PG voltage is outside of range." a few seconds after shutdown. This event triggers the "electrical failure" indicator to blink and prevent power on.

This happens also a few seconds after i connect to power supply cables.

What is the meaning of the message? How to remediate?

I found similar events in this forum but there is a reference to a specific DIMM module. Here there is just PG ?!

The same message appears randomly when the machine is on and disappears immediately without affecting the system stability. Any ideas?

TomekHar,

I would start with the following

1- switch OFF the server
2- remove both power supply cables
3- press and hold the power button for 30 seconds

4 - Take the server to its Minimum to Post configuration (see below)

Minimum to post for the r730XD is removing everything internally and externally from the server, but the following;

- System board
- One power supply unit
- One processor (CPU) in socket CPU1
- One memory module (DIMM) installed in socket A1
- Left control panel (for power button functionality)
- One backplane card

See if the server will post under those conditions, if not then try swapping the power supplies and try testing individual dimms in A1.

If that works then you can try individually adding removed devices back until the issue reappears, pointing to the device causing the issue.

If the server still fails under this configuraiton, and you don't have known good parts we can test with, then you may want to consider that the issue is being caused by the systemboard.

Let me know what you see.

Dear DELL-Chris H ,

thank you for the reply. I have tried as you suggested but the issue persists.

The electrical error indicator starts blinking approx. 1 minute after shutdown.

Regards, T

TomekHar,

Did you try both power supplies individually in PS1 slot during the minimum to post configuration?

Also, is the server connected to a UPS or APC, if so can we disconnect from them and test it going straight to outlet?

If you have another R730xd, we could try testing with some of its known good parts to further isolate the issue, but if it all still fails then it may be a case of the motherboard needing to be replaced. 

You can also try this following test.

In the minimum to post configuration, disconnect the control panel from the systemboard, and while keeping the lid off of the server, hold in the iDRAC button on the rear for 15 seconds. If powers on then the issue could be the  control panel or the systemboard.

Hi DELL-Chris H,

i reply to both your recent posts.

The server is connected directly to power outlets, in a kind of lab setup. I also dont have another R730xd yet.

Yes i tried with both PSU in single setup as well as hot standby setup.

Since last weeks wednesday the server was in operation and i didnt see the voltage error in the idrac logs. So i shutdown the server today and after some time the error appeared again (in powered-off state).

I reset the idrac using the press idrac button for 15s method. After that all front panel indicators went off.

After a few seconds (during which i couldnt power on the server) however they came back (incl. electrical failure).

The only way to power it on was to remove and return electricity and power on during the moments before the electrical failure indicator goes on again.

I noticed that when the server is off, there is no PSU status available in the idrac web-interface. Is that normal?

However, i think we conclude that something is wrong with the board. I have checked for board replacements and came up with 3 options. H21J3, WCJNT and 72T6D. Where can i find information about the differences? Can i use any of these or should i go with the same that is in my server?

Best regards

T

TomekHar,

The different part numbers are different renditions of the motherboard. Now to find out which one, you can reply to private message I sent you with the svc tag and I can confirm the specific board installed, whereas I think you may also be able to by entering the svc tag on the support page.