[SOLVED] TTX3 sudden shutdown after a while, even in BIOS.

[Rafo]

I have recently aquired a TTX3 Multi System.

From cold boot, after a while (hard to tell how long as it varies : the shortest was 10 minutes, the longest about half an hour), whether I'm playing a game or just letting the Attract Mode run on its own, the system suddenly shuts down, then restarts automatically, then shuts down again very quickly. I have to let it rest for a while before making it run again for a longer time.

First I thought it was because or heat, but the HW monitor in BIOS tells me the CPU is barely over 50°C and the "System" is about 35°C so I don't think it comes from here.
Inside of the unit is very clean, the previous owner tool very good care of it.

Anyway, I disabled the FAN monitoring, 'cos the "Target temperature" for CPU was pretty low (choices are DISABLED, 40°C and 45°C, 45 being the default setting) and first I thought it was because the fan was trying to cool it down to less than 45°C.
After disabling FAN monitoring I believed it was better but in fact it only lasted longer but I think it's just a coincidence.

Anyway I'm trying to figure out what is happening, why is the system suddenly and brutally shutting down (after 2 ou 3 shutdowns one after another, it switches to Windows check & repair procedure, that's why I can tell it's not a clean shutdown).

I also thought about instability from the PSU but in the HW monitoring section I could check the voltages and they look really stable to me.

Is there any other thing I should check on the system ? Any program I could use that could reveal a failure somewhere ?
I checked in the Windows Event Viewer to try and get data from any (critical ?) error, but to no avail.

I repeat, this happens even in the BIOS, so it cannot be a software/HDD issue.

Any advice is welcome and I'm ready to provide with any additional required information to pinpoint the source of these shutdowns.

Do you have any ideas ?

 

[djsheep]

Even though your temps are good, replace the thermal paste.

Have also had funny things happen with bad memory sticks, so if you have some spare RAM swap it out.

 

[Rafo]

Even though your temps are good, replace the thermal paste.

Have also had funny things happen with bad memory sticks, so if you have some spare RAM swap it out.

Thermal paste is brand new : changed it yesterday. No change (as I thought).

About RAM, I'm not sure it can do that.
Before replacing it (! don't have some compatible RAM with me right now), I'll run a memtest to check if anything's wrong.
I can also try with only one ram stick at a time, in case it would change something, in the BIOS at least.

The fact it only occurs after a certain amoint of time (never happens earlty after a cold boot) makes me thing something's heating up.
UI don't have a thermal camera to check it though.

I'll try to time the moment the system shutsdown, to check if the time between power on and shutdown is consistent or purely random.
I'll also check in BIOS without the GFX card inserted (can it even get to the bios with the onboard video ? Just in case it could come from the gfx board itself.
I'll also try with another PSU.
I'll keep investigating, hoping it's not the mainboard itself.

Meanwhile, any report would be appreciated from anyone experiecing this kind of symptom in the past.

 

[DavidMcGhan]

I'd be checking the RAM, then i'd be checking the PSU, capacitors could be going bad?

 

[streetfighterken]

Just because voltage measures good doesn't mean it's stable. Sounds like classic bad PSU.

 

[Rafo]

BREAKING NEWS !! I found out what's going on... and we all got it wrong.

This little guy here is responsible for this :

Now let me explain how I found out : I removed components one after another : gfx card, using onboard gfx, then one memory module, then putting it back and removing the other, each time after a while the unit shuts down, even in BIOS.

Then, I removed this board... Then the unit did not start when the power word was plugged in.

I discovered then that it's this board that triggers power on by shunting the PWSW+ and PWSW- pins on the mainboard thru the FP1 cable, just like a power button would do on any standard PC.

So I connected a simple wire between these two pins (orange wire to red wire) on the other side of the FP1 cable (just like the PCIE card would do)

and, the unit started....

It's been MEMTESTing for about an hour now without any shutdown. Usually it shuts down after a few minutes.

Si my theory is that this board decides for some reason these two pins should get connected at some time and you know what happens when you push the power button of your tower PC for a few seconds... right, it shuts down directly.

But the thing is after powering off, the board re-powers it on again after about 10 seconds.... then reshuts-down quite quicker. That's what was driving me crazy.

Anyway it looks like this happens always after a few minutes. But I cannot figure whether it's because this board has a some kind of defective component that does it by itslef, or if the board does it voluntarily because maybe something triggers it to.

Maybe someone can tell me how this board works, and in what circumstances it can decide to shut down the main board thru the "power pins" after a while.
Should I track something down ? Maybe the board does it out of security need because something went wrong on it ? Maybe this can be captured thru serial port used for debug ?
Right now I have no I/O board connected anywhere. (I'm using standard USB sticks connected to the on-board USB). Does it have anything to do with this ? Maybe it triggers shutdown after a while biy itself if it does not find any connection to a FastIO or JVS ?

Any suggestions are welcome.

Meanwhile, I'm gonne use a separate power button instead of this board to power on the system and test it extensively. Let's not close this topic now but I'll say there is a pretty good progress,

[UPDATE] The systme works perfectly without the FP1 connection : just had to set "power on" after power loss and everything's fine.

Set as solved.