Well I tried the industrial Noctua, but the fan hub got so warm it actually made my temps worse. Trying the EVGA fan now, hope it's not the winner as it's a real banshee on max speed.
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
S4 Mini w/ Ryzen 1700 and 1080 TI
- Thread starter Reglar
- Start date
Ok, so for anyone else who buys the Biostar X370GTN, you need to be aware that you'll need to dial in your fan settings, don't use the default ones. I had been doing that, and when I popped in the EVGA fan, it went to max quickly and kept the max temp at 67C. Now, it went to max fan around 60C, so clearly the curve needs to be optimized.
There is no actual tool to let you create a fan curve for this board (that I have found, speedfan won't work, and the Biostar fan control software only lets you switch between 100% on and a default curve. Even after I ran the calibration tool in BIOS on the fan, noise to performance was horrible (great performance, way too loud).
I'm dialing in my own profile now. It's finicky though, last round I thought I was doing well, was around 60% fan speed and temp was holding at 70C, it ticked up to 71C and my CPU fan jumped to 100% speed. Ugh.
Once I actually get a profile I like I'll include this in my report on each fan. Note that I'm not going back to the 2 Noctua's, they couldn't keep the temp below 75C at max.
There is no actual tool to let you create a fan curve for this board (that I have found, speedfan won't work, and the Biostar fan control software only lets you switch between 100% on and a default curve. Even after I ran the calibration tool in BIOS on the fan, noise to performance was horrible (great performance, way too loud).
I'm dialing in my own profile now. It's finicky though, last round I thought I was doing well, was around 60% fan speed and temp was holding at 70C, it ticked up to 71C and my CPU fan jumped to 100% speed. Ugh.
Once I actually get a profile I like I'll include this in my report on each fan. Note that I'm not going back to the 2 Noctua's, they couldn't keep the temp below 75C at max.
Last edited:
That's why i rely on corsair commander mini (and now corsair commander pro) to handle my fans. Even on my asus crosshair hero vi (asus is well known for fan control), i was not satisfied (software that taking too much ressources, insufficient bios setup, etc...)
I love my commander mini...
Well I got through 1 fan today. I know that doesn't seem like much but I'm doing a variety of tests with data graphs and sound recorded.
- Raw synthetics
- Heaven runs
- 3d Mark Time Spy benchmark
- 3d Mark Fire Strike Ultra benchmark
- Metro Last Light benchmark
- Superposition benchmark
Sorry I meant that the duration of the benchmark is sufficient to generate max heat; the temperatures on both CPU and GPU from the first run to the last run are all pretty consistent. Helped validate that using it as a data point is pretty safe for what I am trying to do.
How do you make those amazing graphs???Well I got through 1 fan today. I know that doesn't seem like much but I'm doing a variety of tests with data graphs and sound recorded.
One of my tests is 10 runs of Metro Last LIght, to see if extended use changes the performance. Short story is it didn't. Here are the CPU and GPU graphs. What's interesting is that even though I got great average FPS (103) with pretty reasonable noise levels.
- Raw synthetics
- Heaven runs
- 3d Mark Time Spy benchmark
- 3d Mark Fire Strike Ultra benchmark
- Metro Last Light benchmark
- Superposition benchmark
The data itself is via HWInfo logging, it takes snapshot of the values every second or so and throws it into a Comma Separated Values file. I then open it in Excel, and graph it.
Here's another update. Completed the Kingwin PWM fan today, and received the 5 new fans. Did not receive the PWM to Voltage do-dad yet so will mod the 5 fans Tuesday after work and start testing those. Of the 2 fans tested so far (Corsair, Kingwin) , the Kingwin was much easier on the ears and was pretty much just as effective as the corsair.
I thought I'd post the stats on the new 5 fans, so I updated the spreadsheet data. The new ones are the last 5 on the bottom.
I thought I'd post the stats on the new 5 fans, so I updated the spreadsheet data. The new ones are the last 5 on the bottom.
Whoops, yes that FN123 fan is definitly NOT 5db. Man that would be something. I'll fix it and fix the picture.
Also, minor correction, I said i had tested Corsair and Kingwin, I meant EVGA and Kingwin, with the EVGA definitely being louder.
I'm HOPING that the be quiet! fan will perform well when around 75% fan speed, this fan will be what I start testing tonight.
Also, minor correction, I said i had tested Corsair and Kingwin, I meant EVGA and Kingwin, with the EVGA definitely being louder.
I'm HOPING that the be quiet! fan will perform well when around 75% fan speed, this fan will be what I start testing tonight.
Mid week update, have now completed 5 fans (EVGA, Kingwin, be quiet!, Phanteks, Cougar).
Right now I think Phanteks is in the lead, man that was one quiet fan.
I received the PWM to voltage gizmo but it's not working like I expected with my motherboard, so I think non-PWM fans are out of contention (Corsair and the dark side GTs).
I'll continue testing this week, with the goal to be able to move onto testing the fans on the GPU this weekend.
Right now I think Phanteks is in the lead, man that was one quiet fan.
I received the PWM to voltage gizmo but it's not working like I expected with my motherboard, so I think non-PWM fans are out of contention (Corsair and the dark side GTs).
I'll continue testing this week, with the goal to be able to move onto testing the fans on the GPU this weekend.
Similar threads
Power Supply
PSU for Ryzen 7 1700 S4 Mini Build
