OK, just entering gpufancnv.pyw in terminal just got me "Command not found"

Entering ./gpufancnv.pyw executed the expected script output in the terminal window AND opened up the GUI control window.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

OK, just entering gpufancnv.pyw in terminal just got me "Command not found"

Entering ./gpufancnv.pyw executed the expected script output in the terminal window AND opened up the GUI control window.

Fantastic! That's as it should be, then.

When entering the command for the app in Startup Applications, just use the full path name, without "python". For instance, on this machine mine is entered just as /home/jeff/Programs/gpufancnv/gpufancnv.pyw. (I created the "Programs" directory just to keep some things separate.) It looks like yours would probably be /home/keith//Desktop/gpufancnv/gpufancnv.pyw

Hi again Jeff. Yes, that is exactly the entry I have for the program in Startup Applications. I must have the program set to hide the window from the Desktop somewhere. I know the program opens because it gets the window carat on the side of the icon in the launcher. Clicking on the icon in the Launcher opens the window as expected. It just doesn't show up in initial bootup as an open window.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Oh, that's interesting. Yeah, must be something external because the program itself has no "start minimized" option. I thought about building in such an option, but I'm concerned that it might be too easy to mistake a startup failure for a successful, but minimized, launch. That happens to me periodically with Precision X in Windows, when Windows apparently gets too busy at startup and simply bypasses some of the things it's supposed to start. Since I don't often keep a close eye on my crunch-only machines when they restart each weekday evening, it's usually a disturbing surprise the next day to find that the GPUs have been running at 80C or so all night! :^(

Yes, I'm going to have to look around in the Desktop and UI settings for something that might be causing the problem. There aren't that many things starting at boot, not like my daily driver in Windows.

I was surprised this morning when I gave a look at the Ryzen Win 10 system and noticed the glowing red pump LED. That shouldn't happen. I then noticed the fans on the radiator were only running 900 rpm instead of their normal 1800. For some reason SIV had lost my custom fan profiles and the pump had fallen back to quiet settings. I wasn't happy that the cpu was at 66 degrees C. and the radiator temp was 39 degrees C.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I wasn't happy that the cpu was at 66 degrees C

70°c is about as cool as my i7's CPU gets.
Liquid cooling is affordable now, but that system is getting rather long in the tooth, and it's got a few issues. So I keep saving the $. With Threadripper coming out soon, and Intel's multi core responses, hopefully prices will get a bit of a shake up & i'll get a new system later this year or early next.

---

Grant
Darwin NT

Well the system has a 1700X in it so it has the +20° offset. So the 66° C. was in reality considered to be 86° C. by the system. At 95° C. the CPU shuts itself down. I have that happen every time the system black screens on me and I reboot. The pump for some reason and against all facts only runs at 400 rpm instead of the normal 2850 rpm. The pump is directly powered by a SATA power connector and should run at fixed constant rpm at all times. This allows the system to overheat and shut down. I have to let the system cool down enough to allow me to fully get booted into Windows and have enough time for the pump and fans to run at normal speeds for the temps to fall enough below the shutoff temp. I have not been able to discover why this happens or even CAN happen with the SATA powered pump. All I know is that it does this every time I blue or black screen and have an abnormal Windows shutdown. If Windows or the system is powered down normally, the problem doesn't manifest.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Keith, describe what cables you have on that pump? or give me a name of the pump so I can google it.

You said it's powered by the SATA? Is there also a PWM 4 pin that connected somewhere?

It's a Corsair H110i. The pump is powered by a SATA power connector. The only other connector from the pump to the motherboard is a single wire that supplies the pump speed readout to the CPU fan header. Or in my case to the AIO pump header on my ASUS Prime X370 Pro motherboard. The pump also has a double headed fan connector for controlling the radiator fans. The pump does have a USB Link connector connected to a USB 2.0 port on the motherboard. Corsair Link control is by SIV, NOT by any Corsair Link software.

The problem has me dumbfounded. I have searched the Corsair forums and Googled the problem now for 4 months and have not ever found anything remotely similar.

I have not contacted Corsair since I am and have not been impressed by their support and knowledge of their products in the Corsair forums.

The problem is caused by the pump only running at 400 rpm as reported in the BIOS Monitor pages after a black screen or blue screen or abnormal Windows shutdown. The pump moves very little coolant in normal operation and only has 150 ml of volume to begin with. That is why I moved the radiator to the front of the case because of an in-depth tech report on the AIO that said it has a difficult time in dealing with as little as 12" of gravity head. With it only running at 1/7 of normal rpm after a system failure, it is no wonder the system overheats quickly. I had no idea these AIO systems were so anemic. I can definitely see the appeal of a custom loop with a proper pump.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Ok, it is what I thought it was. But I have to wonder if there is an issue with the pump itself, since it's only running at 1/7th of normal. Was it running normal before you install linux? or was the first time you tried it was when you built your system?

Do you know if another cooler will work with your computer?

I've never cared for the link drive on mine. I prefer the older coolers. These are harder to come now a days, but there are still a few out there. Would it work if you decide to try a different cooler?

https://www.newegg.com/Product/Product.aspx?Item=9SIA6ZP3K32295&cm_re=Hydro_Series_H110_CPU_Cooler-_-35-181-035-_-Product

The pump does have a USB Link connector connected to a USB 2.0 port on the motherboard.

Have you tried leaving that disconnected & rebooting & seeing if the pump then runs at full speed by default?

---

Grant
Darwin NT

That sounds like my SuperMicro board which occasionally gets some bad readings on CPUTIN, PECI Agent, and PCH Chips inputs raging from -116-127C.
It got a lot better since I upgraded my overloaded 750W PSU, but it still happens. My AIO is currently 63C @ 690RPM (should be 52C). But man is it quiet :D It was downright dangerous before on air when it would go to 99C !!!

I know what causes the CPU cooler (and other fans) to slow down, I just don't know what causes it to happen.
All I can say it to try and monitor/log your board sensors.

I wondered if I would confuse people with the post in the Linux thread. This is about the Ryzen Win 10 system. The new linux system has the thrice RMA'd H100iV2 AIO in it. It just runs I assume in balanced mode out of the box since there is no link connected or software to control it in linux in the first place. That system keep the FX-8300@4Ghz roughly running at between 50-54° C. when fully loaded. That is normal for the FX cpus. I have no issue with the temps. And no issues with the pump since it is running without any configuration. I have no idea what speeds the radiator fans are running at since only the pump can report those speeds unless I hook them to a motherboard header and I have no spares.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Ok, since this is your windows computer. I really have to think it's the pump then. I can't think of a reason why the pump wouldn't run normal if it's in it's native environment.

The pump does have a USB Link connector connected to a USB 2.0 port on the motherboard.

Have you tried leaving that disconnected & rebooting & seeing if the pump then runs at full speed by default?

I haven't. The pump runs at a fixed 2850 rpm because there is no control of the pump speed. At the +12V provided I assume by the SATA connector, 2850 rpm is what you get. It never varies under normal operation. I haven't tried to measure the SATA connector voltages under failure condition. I would need to google the pinout on the connector and rig up a safe multimeter connection to see what the voltage is when the BIOS is reporting 400 rpm pump speed.

As I hope I have conferred, the pump always runs at a fixed 2850 rpm under normal boot and Windows conditions. It is only when I have black screened the system because of incorrect BIOS settings for memory or cpu clocks that can cause the failure condition. In case anyone doesn't know what 'black-screened' means, you certainly will if you build a Ryzen system. A black screen is an abnormal Windows shutdown that leaves no traces of itself unlike a 'blue-screen' where Windows can do normal mini-dump logging and you can at least investigate the reason for the abnormal exit. Black screen events are almost always caused by insufficient cpu voltage for the desired overclock.

What I keep wondering and searching for is any reference that the Link enabled Corsair power supplies, in my case the H1000i can not reset the output voltages to normal if the system abnormally exits. I wonder if an abnormal exit causes the link enabled devices to enter an unusual state that limits the power supplies outputs and thus a low SATA +12V voltage output which causes the low 400 rpm pump speed.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

I bought the H110i specifically on the recommendation of Ray Hinchcliffe, the developer of SIV. It is based on the CooliT AIOs and NOT on the Asetek coolers that have mostly taken over the Corsair AIO line. Ray has the opinion that any Asetek branded cooler is a dumbass kludge that has no reason to exist. I agree with him based on my experience with the Asetek originned Corsair H100iV2 AIO on the linux system. The whole Corsair AIO branding and naming scheme is a morass. The H110i GTX for example is Asetek sourced. My H110i is what it used to be known as the H110GT. But the old H100GT that used to be CooliT sourced is now known as the H100iV2 and is in fact Asetek sourced.

The link from Newegg you provided is the H110i GTX and is Asetek sourced. You can always tell an Asetek AIO by the round pump body. My H110i has a square pump body.

I don't think there is an issue with the pump unless it is getting upset somehow by the error condition and enters an invalid state or something. Thankfully that doesn't happen often anymore since I have mostly just settled on my current overclock settings which are stable. It is only when I play around with something new that allows the system to become unstable and exit uncleanly.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

If there is no speed control of the pump, then the only way it can run slower is if it's power supply voltage is reduced. That would only happen if there is a fault with the power supply.
However if there is speed control for the pump, then it would be done by that USB2 connection (as there aren't any other connections that would make it possible).
The issue could be in the pump, if (as it appears) there is a speed control mechanism other than control of the supplied rail (not going to happen with it connected directly to the PSU). However if that were the case I would expect the pump to either run flat out, or not at all.

---

Grant
Darwin NT

What I have done today is to change the default PSU fan speed from quiet @ 535 rpm to performance @ 1460 rpm. My thinking is that it is possible that the PSU is throttling the SATA voltages after system reboot and abnormal Windows exit because for some reason it thinks it is or really is overheated. If I run the PSU fan at full speed and have another black screen event and the pump runs at normal rpm in the BIOS, I have figured out the culprit is the power supply. I really need to jury rig up a DVM to SATA power connector setup to be fully certain.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

If there is no speed control of the pump, then the only way it can run slower is if it's power supply voltage is reduced. That would only happen if there is a fault with the power supply.
However if there is speed control for the pump, then it would be done by that USB2 connection (as there aren't any other connections that would make it possible).
The issue could be in the pump, if (as it appears) there is a speed control mechanism other than control of the supplied rail (not going to happen with it connected directly to the PSU). However if that were the case I would expect the pump to either run flat out, or not at all.

There is no possible control of pump speed. It either runs flat out or not at all.

[Edit] I think I will backtrack on the last statement. I just went back into Link Fans control and it does in fact have a parameter input box for pump speed control. I have it set for State 0 or Fixed PWM with a value of 255 or max. I then tried the other States 1-5 and they correspond to the usual fan control setting or fan speeds such as Default, Fixed RPM, Quiet, Balanced and Performance. So it is possible that the pump speed can be upset or possibly be put into an invalid state. The other test to perform would be as suggested to pull the USB 2.0 link cable. I'm not sure what that would do to radiator fans speeds. I would think they stay in their last controlled state based on the writings in the linux forums about no link control.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

My thinking is that it is possible that the PSU is throttling the SATA voltages after system reboot and abnormal Windows exit because for some reason it thinks it is or really is overheated.

Even multi rail PSUs wouldn't control their rail outputs that way- the only control the motherboard has over the PSU is to turn it on or off. If there's an excessive load, the output of the rail would be current limited, so the voltage would drop. And whatever was causing that excessive load would start smoking...

---

Grant
Darwin NT