So I literally just woke up from a nap to the sound of the fault alarm on my motherboard (Supermicro X9DAE) going off. God only knows how long it was going off, i was asleep for about 3 hours. I checked the front panel of my Aquaero 6XT and noticed that my CPUs were at 178F and all the fans were all turned off! Figuring it was a temperature alarm, I logged in and saw Open Hardware Monitor was still running (and CPUs clocked themselves down to 1.2Ghz) and responsive so I killed the Aquaero process and relaunched it. It was taking a while (>15 seconds) to open, so I decided to open the side panel on my case and smelled that tell tale smell of baking electronics, and a few seconds later the Aquaero software finally to launch, and the fans all turned on at full speed and everything returned to normal.

At this point i've lost a huge amount of faith in this product and am wondering if I should ever leave this computer unattended again with the Aquaero controlling the fans. Does anyone have any clue what happened?!? Are there debugging logs or something else I can look at to do perform root cause analysis?

The Aquaero has two fans connected to each channel, and the most powerful of those fans only draw 0.35A each with the software fuses set to 1A. Since the fans all turned on once the Aquaero software loaded it wasnt due to an over current condition. The front panel was responsive so its not like the microcontroller on it crashed. And Open Hardware Monitor was responsive. And although I should certainly decrease the default sensor time out from 5 minutes (who came up with that value?), since the front panel of the unit was reading 178F, a sensor timeout shouldnt be the cause either.




And on a side note, i've had some problems with this bug ridden software, it crashes whenever I try to create a new controller object as can been seen below. I get ridiculous values whenever I try to enter a value in to the fallback temperature field; Its like it is doubling whatever number I enter. And aparently it can't remember the value of the curve shape for the controller objects.

Zitat von »Jakusonfire«

Don't use software sensors for anything that is safety related.

This doesn't make any sense at all. The fallback temperature was enabled, and set to such a high value that the controller would put the fans on max speed if it lost communication with the PC.

Zitat von »Shoggy«

You should go to the system tab and download the event log from the device in the event log box. It is likely that your software fuse setting has caused the problem because when a fan spins up it draws much more power than it is usually rated. The rated values quite often only tell the regular power consumption but not the peak values.

For the crashes in the software it would be helpful to provide the error logs from the Windows event viewer. Each crash creates two reports but we only need the extended one:
[attach]5481[/attach] [attach]5480[/attach]

I'm aware of inrush current, but as I mentioned, the software fuse couldn't be the cause as according to the manual, you have to remove power to the Aquaero in order for the software fuse to be reset. Or is the manual wrong? The machine (and Aquaero) was never power cycled, yet the fans turned back on after restarting the software on the PC. I'll try to post the event log later today.

Shoggy, i'm sorry but i've had some other obligations and still havent had a chance to get you those logs. Well it happened again, I snapped a couple photos before powering off the system:

Zitat von »Shoggy«

It is likely that your software fuse setting has caused the problem because when a fan spins up it draws much more power than it is usually rated. The rated values quite often only tell the regular power consumption but not the peak values.

I am aware of inrush current. However these fans were likely already spun up when they went off, therefor inrush current would likely not be the cause from that alone. In addition because the manual clearly states that the Aquaero has to be power cycled to reset the software fuses, and the first time this happened the fans restarted after I relaunched the Aquaero Windows application, I have more difficulty believing software fuses were at fault.

I have a Fluke 117 multimeter (along with a oscilloscope and logic analyzer if you want me to break those out) and took some peak current measurements as can be seen below. Although 3 pairs (GPU & PSU, CPU Front and CPU Rear) inrush current exceeds the software fuse's rating, on all previous occasions when I have powered up the computer, they did not trip the software fuse when they spun up. Furthermore the CPU Middle (the Nidecs) pair's inrush current is below the software fuse's 1A rating. Even if the CPU temps both managed to drop below 100F (which would rarely ever happen), and the controller's logic turned off the fans, and then later spun them back up, it would have only triped the fuses on 3 of the 4 channels. The GPU's temp NEVER would have dropped below 100F as they're sandwiched together and one of them pretty much runs at 140F idle, and the controller's logic is configured to not shut them off till they both drop below 100F, therefor if software fuses were at fault, only 3 of 4 channels should have turned off if that was the root cause. I've been up all night and am desperate for sleep, i'll try to get you those logs when I wake up. There was a single software fuse trip when I first installed it and had left the stock rear CPU fans (Sanyo Denki SanAce 9G0812P1F09) which draw 0.58A each (non-inrush), and it did require that I shutdown and then restart the system for the fuse to reset. Just a note but both times I've experience this quad channel failure, those fans were not installed.

Sanyo Denki SanAce 9GA0812P2M0011 peak current from 0 RPMs to 6000 RPMs


Nidec UltraFlo T92T12MS3A7-57 peak current from 0 RPMs to 3800 RPMs

Zitat von »Shoggy«

You should go to the system tab and download the event log from the device in the event log box. It is likely that your software fuse setting has caused the problem because when a fan spins up it draws much more power than it is usually rated. The rated values quite often only tell the regular power consumption but not the peak values.

For the crashes in the software it would be helpful to provide the error logs from the Windows event viewer. Each crash creates two reports but we only need the extended one:

[attach]5481[/attach] [attach]5480[/attach]

So I finally had a chance to check those logs for you. The Aquaero's event logs do not show that software fuses were tripped, except that one time I mentioned previously when I first installed it and forgot that the stock rear fans (Sanyo Denki SanAce 9G0812P1F09) drew over 0.5A each. Just to try to put the software fuses (or other over current conditions) theory to rest i've pulled half the fans out of this system, and each fan is now on its own independent channel. Also oddly enough the Aquaero app is now again allowing me to create controllers without crashing. Do you have any other theories as to what could have caused this quad channel failure?

Windows Event Log - .NET Runtime Stack Trace


Windows Event Log - Application Error


Aquaero Event Log